배액의 다량원소 이온농도 제어를 통한 양액 조제 시스템 개발

Abstract

An automated nutrient solution management system by macronutrient concentration control was developed for a closed hydroponic system. The system measures the concentrations of K, Ca, and NO3 ions with ISEs in the drain solution and supplement the drain solution with concentrated stock solution through an adaptive mixing algorithm. The unmeasured concentrations of other ions such as Mg, SO4, and PO4 were predicted through the correlation between the concentrations of macronutrients measured with ion chromatography in the drain solution. It is crucial to measure the ion concentrations to reuse the drain solution, but how to accurately supply an insufficient amount is also essential. The concentrated stock solutions were used to compensate for the inadequate concentration, and the relationship between the nutrient concentrations was investigated through a lettuce cultivation experiment to determine the unmeasured nutrient concentrations in the recovered drain solution. In the case of PO4, the relationship between the concentrations of measurable nutrients with ISEs showed a high correlations with the coefficient of determination of K = 0.91, NO3 = 0.96, and Ca = 0.87. Among them, NO3 with the highest coefficient of determination was used to predict the concentration of PO4 for the management algorithm. In the case of Mg, the closest relationship with Ca showed the coefficient of determination of 0.99. So the concentration of Ca was used to predict the concentration of Mg for the management algorithm. In the case of SO4, the closest relationship with NO3 showed the coefficient of determination of 0.97. However, the concentration of Ca was used to predict the concentration of SO4 for the management algorithm because the fertilizer’s oversupply occurred when NO3 was used to predict the concentration of SO4. As a result of the supply performance by the algorithm, the overall replenishment error was within 10% for all ions. Still, the error for NO3 and SO4 always occurred inevitably because the algorithm was programmed to calculate the volume of fertilizer according to the input amount’s priority and to minimize excessive supplement. To evaluate the performance of the developed system, the concentrations of K, NO3, and Ca in the controlled solution mixed with the drain solution and stock solution were measured with ISEs in lettuce cultivation test. The replenishment error of K, Ca, and NO3 were within ±10% on average, showing worse results than expected. The performance test showed that the developed system was applicable to a closed hydroponic system.