Identification of PID parameters for system-specific nutrient mixing control for ISE-based hydroponic nutrient management

Hydroponic crop cultivation systems are becoming more popular, but automatic target nutrient solution preparation is still a critical issue. Nutrient mixing control parameters need to be identified for specific systems, as they are not universal for different systems with different components and design specifications. In this study, parameters (Kp, Ki and Kd) of the proportional-integral-derivative (PID) control method were identified for nutrient mixing of a small-scale ion-selective electrode (ISE)-based hydroponic nutrient management system through simulation and validation tests. The nutrient mixing test bench consisted of tanks for nutrients A, B, distilled water, and a nutrient mixing tank, K⁺, NO₃^-, and Ca²⁺ ISEs, pipes, nutrient mixing pumps, and a controller. A PID control method was implemented, and the parameters were identified through mathematical simulations using the Ziegler-Nichols method, and laboratory validation tests. The effects of different pump flow rates on the characteristics of the unit step response were also analyzed. The nutrient mixing performance was evaluated based on the control accuracy and response time. For the experimental test bench, PID control parameters were adjusted. The best performance was observed, when the Kp, Ki, and Kd were 5.85, 4.35, and 1.56, respectively. The settling time, oscillation duration, overshoot (%), and steady-state error for those values were 177 s, 64 s, 13.38%, and -0.95%, respectively. Low pump flow rate (0.03 L min^‑1) showed the minimum steady-state error (-0.40%) but took a long time to reach steady (364 s). In contrast, the rising time for the high flow rate (0.1 L min^‑1) was short (89 s), but the steady-state error was high (-1.36%). The result indicates that the pump flow rate needs to be adjusted system-specifically along with the PID parameters. The PID-based nutrient mixing model and identification process of the parameters would be feasible for preparing the target hydroponic solution with minimum error. However, the flow rates of the input solutions need to be adjusted properly for large scale applications.