A Measurement System For Multiply Ion Concentration Of Nutrient Solution

In facility agriculture, the measurement of ion concentration in nutrient solution is prerequisite for optimal control of nutrient solution. The ion selective electrode is a new branch which appeared in the analytical chemistry field in 1970’s. Because of its special advantage, it will get more extensive application in soilless culture industry. However, the measurement accuracy is seriously interfered by interfering ion, especially the cross-sensitive phenomenon between similar ion in nutrient solution. Traditional approaches cannot gather consideration to both real- time and accuracy, with tedious operational procedure and temperature draft, which restricted the widely use of ion selective electrodes in soilless culture industry.Due to the above problems, this research based on ion selective electrodesfocuses on the ion concentration measurement of the nutrient solution in soilless culture. Firstly, the theory of ion selective electrodes and the approach of multiply ion detection is analyzed before arranging the measurement procedure and the experimental methods. Then after the principle analysis of ion selective electrodes’ cross-sensitive phenomenon, a new correction method for the cross-sensitive phenomenon based on the classification and regression trees(CART) is proposed. A new improvement on the judgment of global fitting error is applied on C ART with its advantages in solving multi- feature complex interacting data. The correction model of cross-sensitive phenomenon is established. The verification tests show the relative measuring error of the potassium ion selective electrode descends from 27.5% to 5.25%, revealing the effect of this new correction method. Finally, A measurement system for multiply ion concentration of nutrient solution is built withMSP430 MCU as control core, ion selective electrodes as main sensors, a touch screen as human-computer interaction interface. Also, DS18B20 is applied to measure temperature of the nutrient solution for auto temperature compensation. The experiments show the average relative error of sodium ion concentration is 5.44%; the average relative error of potassium ion concentration is 6.01%; and the average relative error of nitrate ion concentration is 5.17%. The influence of t he cross-sensitive phenomenon has been obviously improved, which will have a well prospect in facility agriculture.