Dynamic modeling, control and verification for citrus variable-rate technology (VRT) fertilization

It is essential to conserve diminishing natural resources. Hence, precision agriculture practices such as yield monitoring and variable-rate fertilization are widely being implemented in order to minimize the potential negative effects of agriculture on the environment.; The first objective was to benchmark the performance of the commercial variable-rate controller systems with various possible configurations of hydraulic and mechanical components such as flow control valves and encoders. The second objective was to empirically model the physical components of the variable-rate spreader such as flow control valves, hydraulic motor and encoders. The third objective was to develop a PID control algorithm based on these empirical models of the components to control the response of the applicator to the fertilizer requirements of individual trees.; From this study it was determined that the present commercial controllers were not customized for citrus VRT fertilization. Features such as real-time sensor offset compensations for speed variations in the field were not currently available. A common test procedure to benchmark the VRT systems' performance in both GPS and real-time trigger mode for all possible configurations of the commercial VRT systems and also for the model-based PID controllers was developed. Two performance criteria, the "Total Application Error" (TAE) and the "Total Single Tree-Zone Application Error" (TSAE) were proposed. The best performing commercial VRT systems' configuration was determined.; Tests were conducted to determine the dynamic characteristics of the hydraulic components of the VRT system. Based on these characteristics a model-based tuning rule was used to determine the gains for the PID controller. The benchmarking tests proved that the model-based PID controller's performance, for TAE and TSAE criteria, was 62% and 82% better than the best performing commercial VRT controller with the same system components. Additional features such as the delay time and speed compensation which greatly enhance the performance of the applicator were presented and experimentally proven for commercial scale implementation with the existing technologies.