Article

An Intelligent Optical Dissolved Oxygen Measurement Method Based on a Fluorescent Quenching Mechanism Fengmei Li, Yaoguang Wei *, Yingyi Chen, Daoliang Li and Xu Zhang Received: 1 October 2015; Accepted: 1 December 2015; Published: 9 December 2015 Academic Editor: Frances S. Ligler College of Information and Electrical Engineering, China Agricultural University, 17 Tsinghua East Road, Beijing 100083, China; [email protected] (F.L.); [email protected] (Y.C.); [email protected] (D.L.); [email protected] (X.Z.) * Correspondence: [email protected]; Tel.: +86-10-6273-6764; Fax: +86-10-6273-7741

Abstract: Dissolved oxygen (DO) is a key factor that influences the healthy growth of fishes in aquaculture. The DO content changes with the aquatic environment and should therefore be monitored online. However, traditional measurement methods, such as iodometry and other chemical analysis methods, are not suitable for online monitoring. The Clark method is not stable enough for extended periods of monitoring. To solve these problems, this paper proposes an intelligent DO measurement method based on the fluorescence quenching mechanism. The measurement system is composed of fluorescent quenching detection, signal conditioning, intelligent processing, and power supply modules. The optical probe adopts the fluorescent quenching mechanism to detect the DO content and solves the problem, whereas traditional chemical methods are easily influenced by the environment. The optical probe contains a thermistor and dual excitation sources to isolate visible parasitic light and execute a compensation strategy. The intelligent processing module adopts the IEEE 1451.2 standard and realizes intelligent compensation. Experimental results show that the optical measurement method is stable, accurate, and suitable for online DO monitoring in aquaculture applications. Keywords: dissolved oxygen; fluorescent quenching; intelligent compensation

1. Introduction Dissolved oxygen (DO) refers to the oxygen molecules dissolved in water and is essential to maintain human and animal life. Oxygen is an important analyte because of its key role in the life science, biotechnology, medicine, and aquaculture industries. The DO content in water is an indication of water quality, and careful control of oxygen levels is important in the self-purification processes of wastewater [1,2]. Water quality is closely related to the contaminants present in water, such as H2 S, NO2 , NH4 + , and organic matter. Wastewater characteristics, including color, chemical oxygen demand (COD), and biological oxygen demand (BOD), specifically indicate the level of pollutants in industrial wastewater [3]. At the same time, DO plays a very important role in the health and growth of aquatic organisms [4,5]. A DO content of less than 2 mg/L for a certain number of hours causes the suffocation and death of aquatic organisms [6]. For humans, the DO content of drinking water should not be less than 6 mg/L. Consequently, the determination of oxygen concentrations is of high importance in the aquaculture industry and in daily life. However, monitoring the DO content with all its external influencing factors, such as temperature, pressure, and salinity, is difficult. To obtain an accurate DO content, the detection method should implement

Sensors 2015, 15, 30913–30926; doi:10.3390/s151229837

www.mdpi.com/journal/sensors

Sensors 2015, 15, 30913–30926

intelligent compensation. In general, three methods can be used to detect DO content: iodometric, electrochemical, and optical methods [7,8]. The iodometric method [9,10] is a popular and precise method of detecting the DO content in water. It is a fiducial method but has a complex detection process and cannot be used to detect the water quality online. This method is mainly used as a benchmark in the laboratory environment (off-line). The electrochemical method [11–14] uses electrodes to detect the current produced by redox reactions at the electrodes. This method can be classified as polarographic type or galvanic cell type based on the detection principle. The electrochemical method has a long history in the detection of DO content; the first so-called Clark polarographic method was designed by Clark of the YSI Company in 1956 [12]. In contrast to iodometry, the electrochemical method monitors DO content by the oxidation-reduction reaction that occurs between the electrode and DO molecules and consumes oxygen in the detection process. Given that instrumental drift is inevitable with the large number of factors that are involved in determining the detection result, electrochemical sensors require regular calibration and replacement. Optical DO sensors [15,16] are more attractive than the iodometry and electrochemical methods because they have a fast response time, do not consume oxygen, have a small drift over time, have the capability to withstand external disturbances, and require marginal calibration. The detection principle of optical DO sensors is based on fluorescent quenching, including fluorescent lifetime detection and fluorescent intensity detection. Intensity detection can be achieved through the photodiode, unlike lifetime, which should be detected based on the phase shift [17]. This study develops an intelligent optical measurement method based on the fluorescent quenching mechanism. The aforementioned methods have some advantages and disadvantages that make them unsuitable to the aquaculture industry in China. First, the detection of DO content is difficult for the aquarist who must deal with numerous factors which influence the aquaculture industry. A comparison of the three methods shows that the electrochemical method is not a good choice because of its weak anti-interference properties. Second, the DO content is not constant, and insufficient concentration in natural water leads to the death of fishes. Thus, detecting DO content in real time is very important. However, iodometry method water samples must be tested in the laboratory, rendering this method unsuitable for monitoring organisms in actual production for this reason. Finally, traditional optical sensors have several disadvantages, including susceptibility to changes in external temperature, pressure, and salinity and attenuation of light source and drift because of the degradation or leaching of the dye. The influence off all these factors can be decreased by adding intelligent processing modules. The conventional optical DO sensor introduced from abroad is expensive and does not have high accuracy when used in the aquaculture industry. Thus, designing and developing an inexpensive and intelligent dedicated optical DO sensor is necessary. This study proposes and develops an intelligent DO measurement method based on the fluorescent quenching mechanism. The sensor contains four modules: fluorescent quenching detection, signal conditioning, intelligent processing, and power supply modules. The sensor based on fluorescent quenching has several advantageous aspects: lower power consumption, smaller size, higher accuracy, and stronger anti-interference properties than iodometry or electrochemical sensors. 2. Materials and Methods 2.1. The Overall Design of Optical Dissolved Oxygen Sensor Given the presence of unstable influencing factors, the sensor adopts an optical probe based on the quenching of fluorescence. Compared with the traditional DO sensor, the intelligent optical DO sensor proposed in this study has an enhanced probe structure and an additional intelligent processing module. These calibration parameters are stored in the transducer electronic data sheet (TEDS) memory. Figure 1 shows that the fluorescent quenching detection, signal conditioning, intelligent processing, and power supply modules are included in the intelligent sensor.

30914

Sensors 2015, 15, 30913–30926

The fluorescent quenching detection module contains a temperature probe and a DO probe. The temperature probe is responsible for collecting the water temperature signals, and the DO probe is responsible for collecting the DO signals. The original input signal can be converted into 0–2.5 V Sensors 2015, 15, voltage signal bypage–page the signal conditioning circuits. The MSP430 microcontroller, which is the core of the intelligent processing module, is connected to the signal conditioning circuits, TEDS memory, Sensors 2015, 15, page–page the intelligent processing module, is connected to the signal conditioning circuits, TEDS memory, and serial interface [18]. The collected data are fused through multi-probe data fusion technology, and serial interface [18]. The collected data are fused through multi-probe data fusion technology, the intelligent processing is connected tocompatible the signal conditioning circuits, TEDS memory, and the DO value obtained ismodule, transferred through compatible RS485 interface after being processed and the DO value obtained is transferred through RS485 interface after being processed and serial interface [18]. The collected data are fused through multi-probe data fusion technology, and analyzed by the microcontroller. allowsthe the microcontroller to communicate and analyzed by the microcontroller.The TheRS485 RS485 interface interface allows microcontroller to communicate and upper the value obtained is through compatible RS485 interface after being processed with with the PC. The sensor is transferred powered an on-off powersupply, supply, which is also controlled by the the DO upper PC. The sensor is poweredby by an on-off power which is also controlled by the and analyzed by the microcontroller. The RS485 interface allows the microcontroller to communicate MSP430 microcontroller. MSP430 microcontroller. with the upper PC. The sensor is powered by an on-off power supply, which is also controlled by the MSP430 microcontroller.

Figure 1. Overall design of the optical DO sensor.

Figure 1. Overall design of the optical DO sensor. Figure 1. Overall design of the optical DO sensor. 2.2. Design of the Fluorescent Quenching Detection Module

2.2. Design of the Fluorescent Quenching Detection Module

The schematic of the fluorescent detection module is shown in Figure 2. The probe 2.2. Design of the Fluorescent Quenchingquenching Detection Module

The schematic of the fluorescent quenching detection module is shown in Figure 2. isThe has an approximate length of 16 cm and a diameter of 4 cm. The compact probe configuration usedprobe The schematic of the fluorescent quenching detection module is shown in Figure 2. The probe forapproximate compatibility with theofrequirements aquaculture industry. illustrated in Figure 2, the has an length 16 cm and of a the diameter of 4 cm. The As compact probe configuration is has an approximate length of 16 cm and a blue diameter ofsol-gel 4 cm. The compact probered configuration is used contains dual film, glass slide, optical filter, used DO for probe compatibility with high-brightness the requirements ofLEDs, the aquaculture industry. As illustrated in blue Figure 2, for compatibility with the requirements of theThis aquaculture industry. As illustrated in resistance Figure 2, the optical filter paper, and photodiode. modulesol-gel also includes a platinum the DO probe contains dualsilicon high-brightness blue LEDs, film, glass slide, red opticaltofilter, DO probe contains dual high-brightness blue LEDs, sol-gel film, glass slide, red optical filter, blue monitor ambient temperature during measurements. The fluorescent intensity and temperature are to blue optical filter paper, paper, and andsilicon siliconphotodiode. photodiode. This module includes a platinum resistance optical filter This module alsoalso includes a platinum resistance to processed in the software for temperature calibration. monitor ambient temperature during Thefluorescent fluorescent intensity temperature monitor ambient temperature duringmeasurements. measurements. The intensity andand temperature are are processed in the forfor temperature processed in software the software temperaturecalibration. calibration. PCB

OF1

PCB Temperature Probe

OF1 Reference Blue LED Blue Reference LED Blue Excitation LED Blue Excitation LED OF2

Silicon Photodiode Temperature Probe Silicon Photodiode OF3 OF3 Metal Case

OF2 Sol-Gel Film

Metal Case Glass Slide

Glass Slide Sol-Gel Film Figure 2. Schematic of the fluorescence quenching detection module.

Figure 2. Schematic fluorescence quenching detection module. Figure 2. Schematic ofthe the(LA470-02) fluorescenceare quenching detection module. The dual high-brightness blue of LEDs modulated at the same frequency, so that the reference LED can be used to compensate the excitation LED because the light intensity loss of The dual high-brightness blue LEDs (LA470-02) are modulated at the same frequency, so that the blue are exactly theblue same. At the same frequency, the photodiode cansame reduce background The dualLEDs high-brightness LEDs (LA470-02) are modulated at the frequency, so that the reference LED can be used to compensate the excitation LED because the light intensity loss of radiation due to the ambient light in the measurement environment and avoid the excitation of any the reference LED can be used to compensate the excitation LED because the light intensity loss of the blue LEDs are exactly the same. At the same frequency, the photodiode can reduce background fluorescent material. In the addition, the intensity of frequency, the LEDs are to a low level at which the the blue LEDsdue are exactly same. Atthe themeasurement same thereduced photodiode reduce background radiation to the ambient light in avoid can the[19]. excitation of any photobleaching phenomenon of the dye has a smallenvironment probability and of occurring The central radiation due tomaterial. the ambient light inthe the measurement environment and aavoid the excitation of any fluorescent In excitation addition, intensity thenm, LEDs are reduced low level at whichfilter the wavelength of the blue LED is aboutof465 which is filteredtoby a blue bandpass fluorescent material.phenomenon In addition, the intensity the probability LEDs are reduced to a[19]. low The levelcentral at which photobleaching thewavelengths. dye has aofsmall of occurring paper to screen out the light of of other The experimental results show that the blue light wavelength of the blue excitation LED is about 465 nm, which is filtered by a blue bandpass filter paper to screen out the light of other wavelengths. The experimental results show that the blue light 30915 3

3

Sensors 2015, 15, 30913–30926

the photobleaching phenomenon of the dye has a small probability of occurring [19]. The central wavelength of the blue excitation LED is about 465 nm, which is filtered by a blue bandpass filter paper to screen out the light of other wavelengths. The experimental results show that the blue light can induce the sensitive membrane to emit fluorescence at 650 nm. To reduce the influence of parasitic light, the probe is equipped with blue bandpass filter papers (OF1 and OF2) in front of the LEDs and red high-pass filter (OF3) in front of the silicon photodiode. A silicon photodiode (OPT 301) is used to receive the fluorescence emitted from the sol-gel film and blue light from the reference LED. The blue excitation LED and blue reference LED are separated on different sides of the red high-pass filter, which is beneficial in cutting off parasitic light and guaranteeing the accuracy of the optical signal detection. The fluorescent sensing film is the most important part of the DO sensor and its performance significantly influences the accuracy, efficiency, and stability of the sensor. Researchers have conducted several studies on fluorescence indicators [20–22] and found that the most common fluorescent indicators contain metal porphyrin complexes, organic polycyclic aromatic hydrocarbons, and transition metal complexes [23]. Ru(bpy)3 Cl2 has been chosen as the fluorescence indicator in this study because of its highly emissive metal-to-ligand charge-transfer state, long lifetime, and strong absorption in the blue-green region of the spectrum, which is compatible with the high-brightness blue LED [20]. The dye is entrapped in a porous and hydrophobic sol-gel film that is approximately 0.04 mm. The sol-gel film is mounted on the glass slide surface, which should be transparent for the excitation and luminescence to penetrate. The film should also be in the shape of an arc and maintain a stable size; the arc surface is designed to increase the area of contact and avoid surface bubble. The operating principle of the sensor is based on the fluorescent quenching mechanism. The fluorescent quenching process is described by the Stern-Volmer equation [24–26]: F0 “ 1 ` Ksv rQs F

(1)

where: F0 denotes the fluorescence signal intensity of anaerobic water; F denotes the fluorescence signal intensity of the water samples; KSV denotes the Stern-Volmer constant; and [Q] denotes the concentration of quencher. As the equation shows, the degree of fluorescent quenching F0 /F has a straight-line relationship with the concentration of the quencher [Q]. In this system, the quencher is the DO, so the DO content can be determined by using Equation (1). 2.3. Design of the Signal Conditioning Module The block diagram of the intensity measurement system is shown in Figure 3. The signal conditioning module is responsible for collecting and processing the information. The dual LEDs are modulated (MOD) into on-off at a frequency of 20 kHz. LED1 induces the sol-gel film to emit red fluorescence. At the same time, LED2 emits blue light as a contrast. The luminescence signal from the sensor dyes arrives time-delayed at the single photodiode (because of the luminescence lifetime of the dye), so that the reference and luminescence signals are detected in different time windows without interference. The fluorescence signal from the photodiode is converted to a voltage signal via an I/V switching circuit. The signal is then amplified (AMP) to obtain a voltage signal of 0–2.5 V and eliminate the higher harmonics of the signal. The voltage signal is filtered by a low pass (LP) filter and provides a voltage signal proportional to the measured intensity shift. Finally, the output voltage is divided by the voltage value based on the intensity in the anaerobic water to obtain F0 /F, which has a linear correlation with the DO content. As presented in the Stern-Volmer equation, the DO content can be determined. 30916

Sensors 2015, 15, 30913–30926 Sensors 2015, 15, page–page

Pt

LED2

Sensors 2015, 15, page–page RS485

OF1

MOD MCU Pt

Sensitive membrane

LED2 LED1

PD

TEDS OF2 OF1

RS485 MOD

TEDS

MCU LP

amp

OF3 Sensitive membrane

LED1 I/V OF2

Figure 3. Schematic the signalconditioning conditioning module. Figure 3. Schematic of of the signal module. LP 2.4. Design of the Intelligent Processing Module

PD

OF3

I/V

amp

2.4. Design of the Intelligent Processing Module

With the rapid development of computers, communications, and networks, the DO sensor

Figure 3.of Schematic of the signal conditioning module. With theberapid computers, communications, and thetraditional DO sensor should more development intelligent, networked, and integrated [20]. To solve thenetworks, problems of shouldoptical be more intelligent, networked, andsensor integrated [20]. Tostudy solveisthe problems traditional sensors, the intelligent optical DO proposed in this developed withof automatic 2.4. Design of the Intelligent Processing Module which are optical suitable DO for the online detectionin of this aquaculture long periods. opticalcompensations, sensors, the intelligent sensor proposed study isover developed with automatic With the rapidare development computers, communications, and networks, theperiods. DO sensor compensations, which suitable forofthe online detection of aquaculture over long

IEEE1451.2 Standard should be more intelligent, networked, and integrated [20]. To solve the problems of traditional optical sensors, the intelligent optical DO sensor proposed in this study is developed with automatic IEEE1451.2The Standard data transfer is implemented using the protocols described in IEEE 1451.2 [27,28]. The compensations, which are suitable for the online detection of aquaculture over long periods. intelligent sensor is includes a smart using transducer interface module (STIM) and 1451.2 network capableThe The data transfer implemented the protocols described in IEEE [27,28]. application processor (NCAP), and the sensor provides a general interface standard and a TEDS intelligent sensor includes a smart transducer interface module (STIM) and network capable IEEE1451.2 Standard standard format for the intelligent sensor and the field bus. NCAP is connected to the STIM through application processor (NCAP), and the sensor provides a general interface standard and a TEDS The data transfer is implemented using the protocols described in IEEE 1451.2 [27,28]. The an RS485 interface, as shown in Figure 4. standard format for theincludes intelligent sensor and the bus. NCAP is(STIM) connected STIM through intelligent sensor a smart interface module and to network capable The calibration parameters aretransducer stored in field the TEDS as shown in Figure 4. the Through these an RS485 interface, as shown in Figure 4. application processor (NCAP), and the sensor provides a general interface standard and a TEDS parameters, the original signals can been converted into corrected signals, including DO, The calibration are stored in the TEDS as isshown in toFigure 4. through Through standard format forparameters the intelligent sensor and the field bus. NCAP connected the STIM temperature, pressure, salinity, and luminous flux signals. RS485 interface, shown in Figure 4.(2) been these an parameters, the as original signals converted into corrected signals, including DO, D (1) can D D( n) p The pressure, calibration are stored in C the Figure f (salinity, Xparameters X n ) luminous ...flux [ X1  Has1 ]i [shown X 2  H 2in ] j ...[ X n  H4. (2) temperature, signals.    1 , X 2 ,...and i , j ... pTEDS n ] Through these i 0 j 0 p 0 parameters, the original signals can been converted into corrected signals, including DO, Dÿ pand 1qthe Dÿ pluminous 2sensor q Dÿ pnoutput q flux signals. temperature, pressure,Xsalinity, In the equation, variables; Hn denotes the revised output variables; n denotes i

j

f pX X2 , ...X q “ variables; ... and s rX D (1) D (2) D ( n )C n ´ Hn s Dn denotes the ofnoutput Ci,j...p i,j,…,prX denotes polynomial coefficient. 2 ´ H2 s ...rX 1 , order 1 ´ H1the i j p f ( X1 , X 2 ,... X n i)“0  0 Ci , j ... p [ X 1  H1 ] [ X 2  H 2 ] ...[ X n  H n ] j“0 p...“ i 0

j 0

p 0

p

(2)

(2)

In theInequation, Xn denotes the denotes revised output variables; the equation, Xn denotes thesensor sensoroutput output variables; variables; HHn ndenotes thethe revised output variables; Dn denotes the order of output variables; denotes polynomial coefficient. Dn denotes the order of output variables;and andC Ci,j,... i,j,…,p,p denotes thethe polynomial coefficient.

Figure 4. Model of data correction.

The calibration standard equation is shown in Figure 4. The relationship between the output variables and revised output variables, such as DO, temperature, pressure, salinity, and luminous Figure4.4.Model Model of of5data data correction. Figure correction.

The calibration standard equation is shown in Figure 4. The relationship between the output variables and revised output variables, such as DO, temperature, pressure, salinity, and luminous 5

30917

Sensors 2015, 15, 30913–30926

The calibration standard equation is shown in Figure 4. The relationship between the output Sensors 2015, 15, page–page variables and revised output variables, such as DO, temperature, pressure, salinity, and luminous flux flux signals are converted a standard The sensor can directly determine accurate signals are converted into ainto standard form. form. The sensor can directly determine accurate valuesvalues when when the calibration parameters in the TEDS, not onlyautomatic facilitatesrecognition automatic it uses it theuses calibration parameters stored instored the TEDS, which not which only facilitates recognition but the alsosensor solvesautomatic the sensorcorrection automaticproblem. correctionInproblem. In theoptical intelligent but also solves the intelligent DO optical sensor, DO the sensor, the total, channel, and TEDS calibration must In bethis defined. this study, the of total total, channel, and calibration must TEDS be defined. study,Inthe definition of definition total and channel and channel TEDS makes the intelligent optical sensor moreand standardized, and that of the TEDS makes the intelligent optical DO sensor moreDO standardized, that of the calibration TEDS calibration may correct original at any time. These havefor significance for the may correctTEDS the original data the at any time.data These functions havefunctions significance the maintenance maintenance diagnosis the sensors [29,30]. Thus, theoptical intelligent DO sensor is called and diagnosisand of the sensorsof[29,30]. Thus, the intelligent DO optical sensor is called a “plug anda “plug device. and play” device. play” 3. Experimentsand andDiscussions Discussions 3. Experiments 3.1. Calibration and 3.1. Calibration and Validation Validation 3.1.1. Temperature Compensation of the Optical DO Sensor 3.1.1. Temperature Compensation of the Optical DO Sensor The temperature dependence of the oxygen sensor has a number of contributions, including the The temperature dependence of the oxygen sensor has a number of contributions, including the temperature dependence of the fluorescence sensing film, light source, and the diffusion coefficient temperature dependence of the fluorescence sensing film, light source, and the diffusion coefficient in the water. As discussed in Section 2.2, the light intensity related to temperature is measured by in the water. As discussed in Section 2.2, the light intensity related to temperature is measured by the the probe, and the temperature effect is measured by the sensor based on the thermistor [31]. This probe, and the temperature effect is measured by the sensor based on the thermistor [31]. This study study designs a correction model based on the temperature applicable to online DO monitoring. This designs a correction model based on the temperature applicable to online DO monitoring. This model can perform temperature compensation to reduce the influence of temperature through the model can perform temperature compensation to reduce the influence of temperature through the curve fitting algorithm. curve fitting algorithm. To eliminate the influence of temperature on sensitive film and light, anaerobic water and To eliminate the influence of temperature on sensitive film and light, anaerobic water and oxygen saturated solutions are prepared as standard solutions, which are heated with an integrated oxygen saturated solutions are prepared as standard solutions, which are heated with an integrated thermostatic magnetic blender (HWCL-1) from 0 ˝ C to 40 ˝ C in an off-light environment. The values thermostatic magnetic blender (HWCL-1) from 0 °C to 40 °C in an off-light environment. The values are measured by a commercial DO meter (YSI 6150) in steps of 5 ˝ C (0–40 ˝ C), and the data between are measured by a commercial DO meter (YSI 6150) in steps of 5 °C (0–40 °C), and the data between output signal and temperature obtained are displayed in the plot shown in Figure 5. output signal and temperature obtained are displayed in the plot shown in Figure 5.

Figure Line chart chart of of the Figure 5. 5. Line the DO DO content content and and output output voltage voltage signal. signal.

The relationship can can be be described described by by the the following following equations: equations:  DOt1  1   t1U t1 “ β 1 ` β t1 U DO DOt 2   2   t 2U DO (3)  t2 “ β 2 ` β t2 U (3) ..  ’ ’ ’  DO  .   U ’ % DO tn “ βn ` βtn U tn n tn The slope is an important parameter to achieve the DO content, which changes with temperature. The relationship between the slope and temperature can be obtained based on the 309186, the slope of the output voltage signal can be curve fit method in this study. As shown in Figure fitted by Equation (4):

$ ’ ’ ’ ’ &

6

Sensors 2015, 15, 30913–30926

The slope is an important parameter to achieve the DO content, which changes with temperature. The relationship between the slope and temperature can be obtained based on the curve fit method in this study. As shown in Figure 6, the slope of the output voltage signal can be fitted by Equation (4): Sensors 2015, 15, page–page

β t “ α0 ` α1 T ` α2 T 2 ` α3 T 3 ` α4 T 4 ` α5 T 5 (4) t  0  1T   2T 2  3T 3   4T 4  5T 5 (4) In the equation, α0 = 52.488, α1 = ´4.3436, α2 = 0.2157, α3 = ´0.0061, α4 = 0.00008, In the equation, α20 == 52.488, α1 seen = −4.3436, α2 =6,0.2157, 3 = −0.0061, α4 = 0.00008, α5 = −0.0000004, α5 = ´0.0000004, and R 0.9998, as in Figure which αdenotes a very good linear fitting. The DO 2 = 0.9998, as seen in Figure 6, which denotes a very good linear fitting. The DO content can be and R content can be obtained at any temperature, taking Equation (4) into Equation (3), and Equation (5) obtained atachieved. any temperature, taking Equation (4) into Equation (3), and Equation (5) can then can then be be achieved. 2 3 4 DOt “DO β `α0 U ` α4 UT ` α55 UT 5 2 ` α3 UT 3 4  `0Uα1UT 1UT`α2 UT t 2UT  3UT   4UT  5UT

(5) (5)

Figure 6. 6. Fitting Fitting curve curve of of the the slope. slope. Figure

The calibration standard Equation (2) in the TEDS calibration can be adapted as follows: The calibration standard Equation (2) in the TEDS calibration can be adapted as follows: 1

5

i j DOt 1  5 Cij [U  H1 ]  [T  H 2 ]

ÿi ÿ 0 j 0 DOt “ Cij rU ´ H1 si ¨ rT ´ H2 s j 0 j“ 0 be presented by Equation (7): If H1 = 0 and H2 = 0, then Equationi“ (6) can

(6) (6)

DOt Equation  C00  C01T C02Tbepresented C03T  C04by T Equation C05T If H1 = 0 and H2 = 0, then (6)can (7): 2

3

4

5

(7) C10U  C11UT  C12UT 2 2 C13UT 3 3 C14UT 4 4 C15UT 5 5 DOt “ C00 ` C01 T ` C02 T ` C03 T ` C04 T ` C05 T (7) 3 ` C UT 4 ` C 5 The calibration values with and output voltage. The DO `C10 Uare ` Cconsistent UT 2the ` Cactual 11 UT ` C12 13 UT temperature 14 15 UT value must be corrected according to the real-time temperature. Taking C00 = β, C00 − C05 = 0,C10 = α0, calibration temperature and output The and DO C11 =The α1, C 12 = α2, C13 values = α3, Care 14 = consistent α4, and C15with = α5 the intoactual Equation (7) and detecting the voltage. temperature value be corrected according to can the be real-time temperature. Taking C00 = β, the C00range ´ C05of=the 0, outputmust voltage, the accurate DO value determined in the 0–40 °C range (i.e., C = α , C = α , C = α , C = α , C = α , and C = α into Equation (7) and detecting the 10 0The11temperature 1 12 2 13 14 4 track 15 5 sensor). probe in this3 sensor can the temperature response of the DO sensor temperature outputachieving voltage, the accurate DO value can beindetermined in thewhere 0–40 ˝frequent C range in real time,and thereby temperature compensation an application (i.e., the range of the sensor). The temperature probe in this sensor can track the temperature short-term changes in temperature occur. The calibration curve based on intensity measurement response of the DO in real time, thereby temperature inindicator, an application changes because ofsensor the impactsof typical driftachieving sources, including thecompensation leaching of the LED where short-term in temperature occur. Therequires calibration curve based on aintensity output,frequent and ageing of the changes sol-gel matrix. Thus, the sensor recalibration after certain measurement because of the impactsof typical than drift the sources, including the leaching of the period, but thechanges frequency achieves much more progress electrochemical sensor. In addition, indicator, LED output, and ageing of the sol-gel matrix. Thus, the sensor requires recalibration after the temperature compensation should be processed in the shade to avoid the influence of daylight aand certain period, but the frequency achieves much more progress than the electrochemical sensor. In background fluorescence. addition, the temperature compensation should be processed in the shade to avoid the influence of daylight and background fluorescence. 3.1.2. Pressure Compensation of the Optical DO Sensor Given that the DO value increases when the sol-gel film is moderately affected by pressure, the pressure affects the accuracy of the measurement. Pressure compensation is needed in the 30919 [32] is as follows: measurement of oxygen, and the correction formula

DO  CS *

P 101.3

(8)

Sensors 2015, 15, 30913–30926

3.1.2. Pressure Compensation of the Optical DO Sensor Given that the DO value increases when the sol-gel film is moderately affected by pressure, the pressure affects the accuracy of the measurement. Pressure compensation is needed in the measurement of oxygen, and the correction formula [32] is as follows: Sensors 2015, 15, page–page

DO1 “ CS ˚

P 101.3

(8)

In the equation, DO’ denotes the DO in the pressure of P; CS denotes the normal atmospheric In the equation, DO1 denotes the DO in the pressure of P; CS denotes the normal atmospheric pressure value; and P denotes the actual atmospheric pressure value. pressure value; and P denotes the actual atmospheric pressure value. CS can through the tableatatdifferent different temperatures. At different CS be canobtained be obtained through thenational national standard standard table temperatures. At different temperature, the the aqueous solutions to different differentpressures. pressures. mapped in Figure temperature, aqueous solutionsare aresubjected subjected to AsAs mapped in Figure 7, the7, the experiment shows thatthat thethe optical inthis thisstudy study can effectively eliminate experiment shows opticalDO DOsensor sensor introduced introduced in can effectively eliminate the the effecteffect of pressure andand that the pressure hasa agood good effect. of pressure that the pressurecompensation compensation has effect.

Figure test. Figure7.7.Pressure Pressure compensation compensation test.

3.1.3.3.1.3. Salinity Compensation ofof the Salinity Compensation theOptical OpticalDO DO Sensor Sensor Given solubility of of oxygen decreases as the increases, the DO must Given thatthat thethe solubility oxygenininwater water decreases assalt thecontent salt content increases, the DO be compensated according to the salinity. When the salinity (i.e., expressed as the total salinity) is must be compensated according to the salinity. When the salinity (i.e., expressed as the total salinity) below ppt,the thecorrection correction formula given by:by: is below 3535ppt, formula[32] [32]is is given

DO nC DO2  “C CSS ´n∆C S S

(9) (9)

In the DO”DO” denotes the the DODO value at aatsalinity of n; thethe DODO value in in pure water; Inequation, the equation, denotes value a salinity of C n;S denotes CS denotes value pure n denotes salinitythe value; andvalue; ΔCS denotes reduction of DO according to salinity (1 ppt).(1 ppt). water;the n denotes salinity and ∆CSthe denotes the reduction of DO according to salinity CS and be obtained throughthe thenational national standard different temperatures. The The CS and ΔCS∆C can be obtained through standardtable tableat at different temperatures. S can ++, and CO 2´ 2− aqueous solution is measured using KCl, NH as ions in the experiment. The ion density aqueous solution is measured using KCl, NH44 , and CO3 3 as ions in the experiment. The ion density of the measurementsolutions solutions is 0 and 40 ppt. values separatelyseparately are shown are of the measurement isbetween between 0 and 40 The ppt.DO The DO measured values measured on the basis of the experimental data in Figure 8. The analysis of the data shows that the DO values shown on the basis of the experimental data in Figure 8. The analysis of the data shows that the DO have certain effects when ionic strength regulators are added. The experiment shows that the optical values have certain effects when ionic strength regulators are added. The experiment shows that the sensor satisfies the salinity compensation in the aquatic environment. The salinity compensation optical sensor satisfies the salinity compensation in the aquatic environment. The salinity parameters are also stored in the calibration TEDS. compensation parameters are also stored in the calibration TEDS.

30920

aqueous solution is measured using KCl, NH4+, and CO32− as ions in the experiment. The ion density of the measurement solutions is between 0 and 40 ppt. The DO values measured separately are shown on the basis of the experimental data in Figure 8. The analysis of the data shows that the DO values have certain effects when ionic strength regulators are added. The experiment shows that the optical2015, sensor satisfies the salinity compensation in the aquatic environment. The salinity Sensors 15, 30913–30926 compensation parameters are also stored in the calibration TEDS.

Figure 8. 8. Salinity test. Figure Salinity compensation compensation test.

3.1.4. Excitation Compensation 3.1.4. Excitation LED LED Compensation The experimental experimental results results show show that of the the excitation excitation LED The that different different intensities intensities of LED have have aa large large influence on the fluorescence intensity of the sensitive film. The intelligent processing module also influence on the fluorescence intensity of the sensitive film. The intelligent processing module also allows for the adjustment of the calibration parameters to correct the excitation blue LED at the time allows for the adjustment of the calibration parameters to correct the excitation blue LED at the time of measurement (i.e., if this differs from the excitation at the time of calibration). The luminous flux of measurement (i.e., if this differs from the excitation at the time of calibration). The luminous flux 8 exponential form at 100 ˝ C. When the service attenuation curve of the blue LED shows a similar time reaches about 10,000 h, the luminous flux of the blue LED (L470-02) drops to about 70% of the original luminous flux. Thus, when the LEDs work beyond 10,000 h, the sensor probe must be replaced. Before replacement, illuminant compensation is required to guarantee the accuracy of the DO sensor. The experimental results show that this sensor can effectively compensate the effect of the light source caused by the instability or aging of the light source. The compensation formula is shown in the following equation, and the calibration parameters are also stored in the TEDS. When the intensity of the reference blue LED changes, Equation (8) is used to correct the intensity value of fluorescence quenched by oxygen: Eδ “ Eγ ¨

eβ eα p0.7 ď ď 1q eβ eα

(10)

In the equation, eα denotes the original intensity value of the reference LED; eβ denotes the measured intensity value of the reference LED; Eγ denotes the measured intensity value of the quenched fluorescence; and Eδ denotes the corrected intensity value of the quenched fluorescence. The long-term stability of the sensor in aquaculture tank water has been continuously studied for over 12 months. With the passage of time, the sol-gel matrix is prone to losing its original features, which may cause drifts in calibration, leaching, and bleaching of the dye. As described above, if the intensity of the fluorescence declines (i.e., after about 12 months), then the sensor probe should be replaced rather than the whole sensor device with the separate structure. 3.2. Analysis of the Performance of the Sensor To verify the performance of the sensor, the optical DO sensor is tested in terms of our aspects: accuracy, stability, precision, repeatability, rapidity. Its superior performance proves that the intelligent optical DO sensor can be used to monitor the DO content in the aquaculture industry. The experimental processes are shown below. 3.2.1. Accuracy Test In the laboratory, the room temperature is about 25 ˝ C, and three groups of standard aqueous solutions are measured: 5.02, 9.98, and 18.03 mg/L. The defined oxygen contents are adjusted by mixing the oxygen and nitrogen with mass flow controllers. The sample contents are checked using a commercial DO meter. Compared with reference values, the experimental data in Table 1 show that 30921

Sensors 2015, 15, 30913–30926

the relative error is less than ˘2%, the measurement error is less than 0.2 mg/L, and the resolution of the optical DO sensor is 0.01 mg/L within the range of 0–20 mg/L, which conforms to the accuracy requirements. At the same time, the temperature error is less than 0.5 ˝ C, which shows that the temperature probe is accurate. Table 1. Accuracy test. Samples 5.02 9.98 18.03

Measurements 4 5

1

2

3

4.86 10.12 17.85

4.79 10.09 17.83

4.89 10.21 17.91

4.93 9.89 17.87

4.89 10.06 17.96

Absolute Relative Error Error

AVG

6

7

8

4.91 9.91 17.87

5.09 9.89 17.93

4.98 10.13 18.03

4.92 10.04 17.91

0.1 0.06 0.12

1.99% 0.60% 0.67%

3.2.2. Stability Test The response of the optical sensor appears to be stable within the measurement error, which meets the specified stability requirement of 0.2 mg/L. In the laboratory, the DO content is measured Sensors 2015, within 12 h15,inpage–page steps of 1 h. The standard aqueous solution is adjusted by mixing the oxygen and nitrogen with mass flow controllers. The experimental results are 2.51, 5.58, 7.22, 10.01, 14.56, 18.1218.12 mg/L. As shown in Figure 9, the largest error the and mg/L. As shown in Figure 9, the largest errorisisless lessthan than 0.02 0.02 mg/L, mg/L, which which meets meets the requirements of the optical dissolved oxygen sensor. requirements of the optical dissolved oxygen sensor.

9. Stability test. Figure 9.

3.2.3. Precision Precision Test Test 3.2.3. Precision refers refers to to the the discrete discrete degree degree of of the the values values that that are are repeatedly repeatedly measured measured by by the the same same Precision method under the same experimental conditions for the same sample. Generally, the relative method under the same experimental conditions for the same sample. Generally, the relative standard standard deviation (RSD) used tothe explain sensor precision. RSD indicates that the deviation (RSD) is used to isexplain sensorthe precision. The RSD The indicates that the results areresults more are more concentrated. The standard solution contains Na 2 SO 3 , and this sample is continuously concentrated. The standard solution contains Na2 SO3 , and this sample is continuously measured ten measured times. The results measurement results are shown Table 2. The micro-controller calculates times. The ten measurement are shown in Table 2. Thein micro-controller calculates the RSD every the measurements, RSD every ten indicating measurements, indicating that the precision of this group of data is the veryRSD good ten that the precision of this group of data is very good when is when the RSD is less than 2%. The micro-controller inputs these data into the internal memory. As less than 2%. The micro-controller inputs these data into the internal memory. As shown in Figure 10, shown in Figure 10, the sensor has a high precision. the results show that theresults sensorshow has a that highthe precision. Table 2. Precision test. Sample

1

2

3

4

5

6

7

8

9

10

AVG Average average

RSD

A

5.56

5.62

5.60

5.44

5.49

5.51

5.50

5.82

5.71

5.60

5.59

1.97%

B

9.96

9.91

9.89

10.22

10.15

10.04

10.00

9.98

10.02

10.15

10.03

1.10%

C

18.90

18.85

18.86

18.74

18.69

30922 18.56

18.74

18.63

18.60

18.59

18.73

0.64%

measured ten times. The measurement results are shown in Table 2. The micro-controller calculates the RSD every ten measurements, indicating that the precision of this group of data is very good when the RSD is less than 2%. The micro-controller inputs these data into the internal memory. As shown in Figure 10, the results show that the sensor has a high precision. Sensors 2015, 15, 30913–30926

Table 2. Precision test. Table 2. Precision test. Sample A B C

1

2

Sample 1 5.62 2 5.56

3

4

5

6

7

8

9

3 5.60

45.44

5 5.49 6

7 5.51

85.50

9 5.8210

10

AVG Average 5.71 average 5.60

A 5.60 5.44 5.51 10.04 5.50 5.82 5.60 10.02 5.59 9.96 5.56 9.91 5.62 9.89 10.22 5.4910.15 10.00 5.71 9.98 10.15 B 9.96 9.91 9.89 10.22 10.15 10.04 10.00 9.98 10.02 10.15 10.03 18.90 18.8518.8518.86 18.74 18.69 18.69 18.74 18.60 18.63 18.59 C 18.90 18.86 18.74 18.56 18.56 18.74 18.63 18.59 18.60 18.73

AVG Average average

RSD

RSD 5.59

1.97%

1.97% 10.03 1.10% 18.73 0.64%

1.10% 0.64%

Figure 10. 10. Precision Precision test. test. Figure

3.2.4. Repeatability Test Na2 SO3 is added into the aqueous solutions 10 to achieve the sample solutions (i.e., 2.51, 9.24, and 12.57 mg/L). The sensor is used to measure the DO values of the sample solutions. The data are recorded every hour and continuously measured ten times. Table 3 shows that the relative error is less than 2%. The result of the data analysis indicate that the repeatability meets the requirements of the optical DO sensor. Table 3. Stability test. Samples 2.51 9.24 12.51

1

2

3

2.54 9.22 12.49

2.5 9.19 12.44

2.44 9.16 12.45

Measurement Results 4 5 6 7 2.56 9.13 12.44

2.62 9.33 12.67

2.48 9.25 12.71

2.63 9.23 12.58

8

9

10

2.55 9.22 12.55

2.59 9.16 12.49

2.63 9.13 12.51

Relative Error 1.59% 0.43% 0.16%

3.2.5. Speed Test Two cups of water are prepared in laboratory. 5% Na2 SO3 and a suitable amount of CoCl2 are added to one of them. They are placed in the air for two days to ensure that both temperatures are same, which can eliminate the influence of temperature on the measurement results. Series A indicates the change of DO content from tap water to anaerobic water, and the Series B line indicates the change from anaerobic water to tap water. From Figure 11, we can see that the response time is less than 3 min. After 1 min, the Series A is nearly 0 mg/L. However, the response time of Series B is about 3 min. The reason of this phenomenon is that the process of fluorescence quenching must achieve dynamic balance. Series B should take some time to react, but it doesn’t need time in anaerobic water. The data indicates that detection response speed meets the requirements of the optical DO sensor.

30923

indicates the change from anaerobic water to tap water. From Figure 11, we can see that the response time is less than 3 min. After 1 min, the Series A is nearly 0 mg/L. However, the response time of Series B is about 3 min. The reason of this phenomenon is that the process of fluorescence quenching must achieve dynamic balance. Series B should take some time to react, but it doesn’t need time anaerobic water. The data indicates that detection response speed meets the Sensors 2015,in 15, 30913–30926 requirements of the optical DO sensor.

Figure 11. Speed Test. Figure 11. Speed Test.

3.2.6. On-Site Verification Test 3.2.6. On-Site Verification Test Figure 12 is the real-time graph of dissolved oxygen changes in aquaculture ponds. After Figure 12 is the real-time graph of dissolved oxygen changes in aquaculture ponds. After sunrise, sunrise, phytoplankton produces photosynthesis, which releases a large amount of oxygen. The phytoplankton produces photosynthesis, which releases a large amount of oxygen. The increase of increase of oxygen is more than the consumption, so the DO content gradually increases. Over a oxygen is more than the consumption, so the DO content gradually increases. Over a period of time, period of time, at the peak of photosynthesis, DO content is at a maximum. After sunset, the at the peak of photosynthesis, DO content is at a maximum. After sunset, the phytoplankton undergo phytoplankton undergo respiration which consumes oxygen instead of photosynthesis. In addition, respiration which consumes oxygen instead of photosynthesis. In addition, with the respiration with the respiration of aquatic animals, DO content decreases rapidly. After all night, the of aquatic animals, DO content decreases rapidly. After all night, the accumulation is reduced to Sensors 2015, 15, page–page accumulation is reduced to a minimum at dawn. These changes perfectly accord with the test data, a minimum at dawn. These changes perfectly accord with the test data, which proved that the which proved that the developed DO sensor has a reliable performance. In contrast to change rates of developed oxygen consumed byhas organisms, temperature littletoinfluence on DO contentconsumed in the DO sensor a reliablethe performance. In has contrast change rates of oxygen 11 on-site test. by organisms, the temperature has little influence on DO content in the on-site test.

Figure 12. 12. On-site Test. Figure On-site Test.

4. Conclusions 4. Conclusions To detect the dissolved oxygen content exactly in aquaculture industry applications, an To detect the dissolved oxygen content exactly in aquaculture industry applications, an intelligent optical dissolved oxygen measurement method based on a fluorescent quenching intelligent optical dissolved oxygen measurement method based on a fluorescent quenching mechanism has been designed in this paper. The advantages of the unique sensor head based on mechanism has been designed in this paper. The advantages of the unique sensor head based fluorescent quenching have been highlighted and intensity measuring circuits and intelligent on fluorescent quenching have been highlighted and intensity measuring circuits and intelligent processing have been described. With the optical dissolved oxygen sensor probe in the paper, the processing have been described. With the optical dissolved oxygen sensor probe in the paper, the detection of dissolved oxygen became more accurate and interference-free. In view of the influence detection of dissolved oxygen became more accurate and interference-free. In view of the influence of temperature, pressure, salinity and blue excitation LED on measurement, this paper has designed of temperature, pressure, salinity and blue excitation LED on measurement, this paper has designed an intelligent processing module. The accurate DO content can be achieved by the intelligent optical DO sensor with the correction coefficients stored in the calibration TEDS. The experiments have shown that the sensor performance can meet the30924 specifications for the monitoring DO content in aquaculture industry, such as low-power consumption, fast response, ability to detect online, high accuracy and strong stability, which will enable reliable operation for periods of months at the very

Sensors 2015, 15, 30913–30926

an intelligent processing module. The accurate DO content can be achieved by the intelligent optical DO sensor with the correction coefficients stored in the calibration TEDS. The experiments have shown that the sensor performance can meet the specifications for the monitoring DO content in aquaculture industry, such as low-power consumption, fast response, ability to detect online, high accuracy and strong stability, which will enable reliable operation for periods of months at the very least. In addition, the use of an optical measurement method and the intelligent calibration technique could be applicable in other medical and environmental monitoring areas. Acknowledgments: Funds for this research was supported by “Special Fund for Agro-scientific Research in the Public Interest”, Modern Fishery Digitization and Integration & Demonstration of Internet of Things Technology (201203017). Author Contributions: Study concept and design: Yaoguang Wei, Daoliang Li and Yingyi Chen; data acquisition: Fengmei Li and Xu Zhang; statistical analysis: Xu Zhang and Fengmei Li; writing of the manuscript: Fengmei Li and Xu Zhang; study supervision: Daoliang Li, Yaoguang Wei and Yingyi Chen. Conflicts of Interest: The authors declare no conflict of interest.

References 1.

2.

3. 4.

5.

6.

7. 8. 9.

10.

11. 12. 13. 14.

Gao, M.; Zhang, F.; Tian, J. Dissolved oxygen online acquisition terminal based on wireless sensor network. In Proceedings of the 2008 4th International Conference on Wireless Communications, Networking and Mobile Computing (WiCOM), Dalian, China, 12–14 October 2008; pp. 1–4. Niu, C.; Zhang, Y.; Zhou, Y.; Shi, K.; Liu, X.; Qin, B. The Potential Applications of Real-Time Monitoring of Water Quality in a Large Shallow Lake (Lake Taihu, China) Using a Chromophoric Dissolved Organic Matter Fluorescence Sensor. Sensors 2014, 14, 11580–11594. [CrossRef] [PubMed] Chong, S.; Aziz, A.; Harun, S. Fibre Optic Sensors for Selected Wastewater Characteristics. Sensors 2013, 13, 8640–8668. [CrossRef] [PubMed] Wilhelm Filho, D.; Torres, M.A.; Zaniboni-Filho, E.; Pedrosa, R.C. Effect of different oxygen tensions on weight gain, feed conversion, and antioxidant status in piapara, Leporinus elongatus (Valenciennes, 1847). Aquaculture 2005, 244, 349–357. [CrossRef] Foss, A.; Vollen, T.; Øiestad, V. Growth and oxygen consumption in normal and O2 supersaturated water, and interactive effects of O2 saturation and ammonia on growth in spotted wolffish (Anarhichas minor Olafsen). Aquaculture 2003, 224, 105–116. [CrossRef] Wang, J.; Bian, C.; Tong, J.; Sun, J.; Li, Y.; Hong, W.; Xia, S. Modification of Graphene on Ultramicroelectrode Array and Its Application in Detection of Dissolved Oxygen. Sensors 2015, 15, 382–393. [CrossRef] [PubMed] Ramamoorthy, R.; Dutta, P.K.; Akbar, S.A. Oxygen sensors: Materials, methods, designs and applications. J. Mater. Sci. 2003, 38, 4271–4282. [CrossRef] Dai, W.Y.; Sun, L. The Measurement Methods of Dissolved Oxygen in Water. Anhui Agri. Sci. Bull. 2007, 13, 77–79. Novic, M.; Pihlar, B.; Dular, M. Use of flow injection analysis based on iodometry for automation of dissolved oxygen (Winkler method) and chemical oxygen demand (dichromate method) determinations. Fresenius Z. Anal. Chem. 1988, 332, 750–755. [CrossRef] Zhao, Y.; Sun, L.; Li, F.M. The research about detection of dissolved oxygen in water based on C8051F040. In Proceedings of the 2009 International Conference on Information Engineering and Computer Science, ICIECS 2009, Wuhan, China, 19–20 December 2009; pp. 1–4. Landman, M.J.; van den Heuvel, M.R. An improved system for the control of dissolved oxygen in freshwater aquaria. Water Res. 2003, 37, 4337–4342. [CrossRef] Jalukse, L.; Leito, I.; Mashirin, A.; Tenno, T. Estimation of uncertainty in electrochemical amperometric measurement of dissolved oxygen concentration. Accred. Qual. Assur. 2004, 9, 340–348. [CrossRef] Hasumoto, H.; Imazu, T.; Miura, T.; Kogure, K. Use of an optical oxygen sensor to measure dissolved oxygen in seawater. J. Oceanogr. 2006, 62, 99–103. [CrossRef] Helm, I.; Jalukse, L.; Vilbaste, M.; Leito, I. Micro-Winkler titration method for dissolved oxygen concentration measurement. Anal. Chim Acta 2009, 648, 167–173. [CrossRef] [PubMed]

30925

Sensors 2015, 15, 30913–30926

15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28.

29.

30. 31. 32.

Li, A.; Wang, Y.; Hu, Q.; Shieh, W. Few-mode fiber based optical sensors. Opt. Express 2015, 23, 1139–1150. [CrossRef] [PubMed] Deepa, N.; Balaji Ganesh, A. Sol-gel based portable optical sensor for simultaneous and minimal invasive measurement of pH and dissolved oxygen. Measurement 2015, 59, 337–343. [CrossRef] Eich, S.; Schmälzlin, E.; Löhmannsröben, H. Distributed Fiber Optical Sensing of Oxygen with Optical Time Domain Reflectometry. Sensors 2013, 13, 7170–7183. [CrossRef] [PubMed] Zheng, G.L.; Xu, Z.W. A new design of high-precision dissolved oxygen sensor. Transducer Microsyst. Technol. 2012, 112–114. McDonagh, C.; Kolle, C.; McEvoy, A.K.; Dowling, D.L.; Cafolla, A.A.; Cullen, S.J.M. Phase fluorometric dissolved oxygen sensor. Sens. Actuators B Chem. 2001, 74, 124–130. [CrossRef] Zhao, S.Y.; Harrison, B.S. Morphology impact on oxygen sensing ability of Ru(dpp)3 Cl2 containing biocompatible polymers. Mater. Sci. Eng. C 2015, 53, 280–285. [CrossRef] [PubMed] Lee, S.; Ibey, B.L.; Coté, G.L.; Pishko, M.V. Measurement of pH and dissolved oxygen within cell culture media using a hydrogel microarray sensor. Sens. Actuators B Chem. 2008, 128, 388–398. [CrossRef] Chu, C.; Chuang, C. Optical fiber sensor for dual sensing of dissolved oxygen and Cu2+ ions based on PdTFPP/CdSe embedded in sol-gel matrix. Sens. Actuators B Chem. 2015, 209, 94–99. [CrossRef] Hartmann, P. Photochemically Induced Energy-Transfer Effects on the Decay Times of Ruthenium Complexes in Polymers. Anal. Chem. 2000, 72, 2828–2834. [CrossRef] [PubMed] Feng, W.W.; Zhou, N.; Chen, L.X. An optical sensor for monitoring of dissolved oxygen based on phase detection. J. Opt. 2013, 15, 55502–55507. [CrossRef] Ding, Q.S.; Ma, D.K.; Li, D.L. Research and application on compensation and calibration methods for smart sensor of dissolved oxygen. J. Shandong Agric. Univ. 2012, 42, 567–571. Sakaue, H.; Ozaki, T.; Ishikawa, H. Global Oxygen Detection in Water Using Luminescent Probe on Anodized Aluminum. Sensors 2009, 9, 4151–4163. [CrossRef] [PubMed] Woods, S.P.; Lee, K.; Bryzek, J. An overview of the IEEE-P1451.2 smart transducer interface module. Analog Integr. Circuits Signal Process. 1997, 14, 165–177. [CrossRef] Shao, H.; Zhuang, J. Research into the Technology of Intelligent Sensors Based on IEEE1451 Standard. In Proceedings of the 2012 International Conference on Industrial Control and Electronics Engineering (ICICEE), Xi’an, China, 23–25 August 2012; pp. 1797–1799. Petru, P.; Mircea, B. Hardware implementation of a PIC18F448 based TIM for IEEE1451.2 compliant actuator control. In Proceedings of the 2010 9th International Symposium on Electronics and Telecommunications (ISETC 2010), Timisoara, Romania, 11–12 November 2010; pp. 119–122. Kumar, A.; Singh, I.P.; Sud, S.K. Energy Efficient and Low-Cost Indoor Environment Monitoring System Based on the IEEE 1451 Standard. IEEE Sens. J. 2011, 11, 2598–2610. [CrossRef] Ding, Q.S.; Tai, H.J.; Ma, D.K. Development of a Smart Dissolved Oxygen Sensor Based on IEEE1451.2. Sens. Lett. 2011, 9, 1049–1054. [CrossRef] Lv, B.; Lei, Z.; Liu, J.; Qu, J.L. Design of a PIC18 F2520 based polarographic dissolved oxygen sensor. Shandong Sci. 2012, 25, 73–77. © 2015 by the authors; licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons by Attribution (CC-BY) license (http://creativecommons.org/licenses/by/4.0/).

30926

An Intelligent Optical Dissolved Oxygen Measurement Method Based on a Fluorescent Quenching Mechanism.

Dissolved oxygen (DO) is a key factor that influences the healthy growth of fishes in aquaculture. The DO content changes with the aquatic environment...
NAN Sizes 0 Downloads 10 Views