LabVIEW를 이용한 CTD용 해양물리센서의 정밀도 검정 시스템 개발에 관한 연구

Abstract

In this paper, “Sensor Verification System(SVS)“ which is used to check the status of Temperature, Conductivity, Oxygen sensors using with CTD(Conductivity Temperature Depth) system has been designed and performance test was done, CTD is the most famous equipment in the world using universally in ocean exploration. SVS has been designed with CTD sensors and "LabVIEW" software, NI PXI system, DAQ board, and more. To share the measured data of CTD system, local network system has been installed with CTD and SVS. Also, Optical oxygen sensor has been installed to compare with original electrochemical oxygen sensor using in the CTD. The Software of the SVS was designed three types programs for test each sensors of CTD system. The result of test for each sensors using SVS is as follows.

Temperature sensor and Conductivity sensor use magnitude of analog output frequency as the source of sensors data. The performance test of SVS is implemented using 500kHz frequency which is maximum sampling frequency of NI DAQ board. The standard deviation of temperature sensor was 0.1~0.2℃, this is 10 times greater than CTD system. In the case of Conductivity sensor, standard deviation was 0.01~0.02 S/m, this is 50 times greater than CTD system.

Assuming that the CTD data is standard data of this test, the average data error of temperature sensor from SVS was -0.0176℃ and average of the absolute error value was 0.1473. The average data error of conductivity sensor from SVS was -0.0011 S/m and average of the absolute error value was 0.0133 S/m.

Temperature sensor and Conductivity sensor have high accuracy in the test of average data error, but, when we check the average of the absolute error value, this value is similar with standard deviation and it appears that, the precision is slightly lowered. this means, SVS has a lower stability compared to the CTD system. To solve this problem, we can make the system to produce a output data by averaging acquired data for 1~10 minutes, then we can have high accuracy and reduce the data errors. But, radical method to improve the accuracy of data is that, upgrade NI DAQ board to the better one which has more high quality of sampling frequency than we can reduce standard deviation of the data.

In the case of test SVS with oxygen sensor, we have to make test environment because, electrochemical oxygen sensor measure the data with consuming real oxygen so, we have to supply oxygen at the hall where the membrane is located continuously on the speed higher than certain level. So that, performance test of SVS was implemented in the air first, and second test was implemented in the seawater located in the plastic water tank.

The two electrochemical oxygen sensor was installed on the SVS system. In the first test, electrochemical oxygen sensor doesn't use frequency as a resource of oxygen data but, DC output voltage so, we can measure the high accuracy data even though SVS doesn't have high frequency of sampling data.

The standard deviation of each sensor data of the CTD was 0.0013 ml/l, and in the case of SVS, first oxygen sensor was 0.0018 ml/l, second was 0.0017 ml/l. this looks, Difference of Standard deviation between CTD and SVS is very small. And the standard deviation of data acquired from optical oxygen sensor was 0.0043 ml/l, this was 3 times greater than CTD system.

Assuming that the CTD data was standard data of this test, average of data error was 0.0581 ml/l, 0.016 ml/l. optical oxygen sensor was 0.2834 ml/l. Average of the absolute error value was same with average of data error. The tendency of this test was that, the stability of the SVS system was similar with CTD system but, there are some data errors which is not small. The main cause of this errors are follows, first is measuring environment of oxygen sensor is not flat. and the oxygen of the air consist not equally. second is correction error of own sensor.

Second test is implemented in the seawater. Average of error data from SVS was 0.0909 ml/l, -0.0223 ml/l, Optical oxygen sensor was 2.9503 ml/l. Average of the absolute error value from SVS was 0.0912 ml/l, 0.0444 ml/l. Optical oxygen sensor was 2.9503 ml/l. The data of optical oxygen sensor was 2 times greater than data of CTD, SVS. So, If we acquire the oxygen sensor data in the seawater, it is impossible to compare oxygen data with optical oxygen sensor, in this case we have to prepare the special equipment which supply equal oxygen to compare the data of oxygen sensor more precisely.