Design and Performance Evaluation of an Injection-Type Online Liquid Fertilizer Blending System with WOA Optimized Fuzzy PID Control

Abstract

To address poor mixing uniformity, unstable control accuracy, and limited flexibility in concentration adjustment during liquid fertilizer deep application, this study developed an online liquid fertilizer blending system integrating a Venturi injector, a helical static mixer, an electrical conductivity (EC) detection module, and a whale optimization algorithm (WOA)-optimized fuzzy PID controller. EC calibration tests, Simulink control simulations, and bench tests were conducted to evaluate the accuracy of calibration-based indirect concentration estimation, dynamic response, blending stability, and outlet uniformity. The EC calibration model used to indirectly estimate the preparation concentration of the tested fertilizer solution had a coefficient of determination of 0.997, with validation errors within 7.23%. Under the specified delay-free nominal plant model, the simulated overshoot and settling time of the WOA-optimized fuzzy PID controller were approximately 6.5% and 3 s, respectively. Bench tests of the system implementing this controller showed a blending error of 7.94%, an average measured response time of approximately 4.5 s, and a maximum spatial coefficient of variation of 8.95%. These findings provide a basis for developing integrated concentration-monitoring and feedback-control systems for on-demand liquid fertilizer blending in deep-application equipment.