Field Study of PID Parameter Tuning Investigation in Peristaltic Dosing Pump Control for Use in Automated Fertilizer Mixing System

Abstract

Proportional–integral–derivative (PID) controller is one of the linear control methods of closed-loop irrigation control. Today, it is mostly used in the industry to regulate industrial process variables such as pressure, flow rate, temperature, heat, and others. In precision farming, the use of technologies such as sensors and PID controllers is popular nowadays. The technologies help the farmer fasten the fertilizer's mixing process before using it directly to the crop. However, not all farms afford to use high technology, and using low-cost mixing systems is one way to reduce their cost. Therefore, this research investigates the accuracy of fertilizer mixture using low-cost methods, compares the results, and proposes improvements to ensure a better fertilizer distribution for precision farming. Different parameter tuning will affect the mixing process of the fertilizer. A series of experiments will be conducted based on the PID control parameter, the concentration of the salt solution, and the initial amount of nutrients in the mixing tank. The salt solution will be used as an alternative for the fertilizer because it contains some similar composition ingredients that the EC sensor can detect. The experiment starts with a preliminary test to determine the temperature effect on reading where room temperature will be the best for the next experiment. The best PID parameter that is used is K_P= 130 and K_I= 60 to prevent the phenomenon of overshoot occur during the mixing process. Using the same PID parameter for the next experiment had also found out that the higher the concentration of the salt solution, the higher the chance occur the phenomenon of overshoot. The initial amount of nutrients inside the mixing tank does not affect much for the mixing process because the best parameter for the experiment we had found from the previous experiment is used. The results extracted from the experiment that shows the mixing process from a low-cost automated fertilizer mixing system.

Application of PSO-BPNN-PID Controller in Nutrient Solution EC Precise Control System: Applied Research

“…To meet their requirements, the structure of the nutrient solution to control the secondary mixed fertilizer was designed, and a mathematical model of the dynamic process of the nutrient solution was established. At the same time, a PI control algorithm was also designed, and their testing verified that the steady-state time dimension of the system was 100 s, the overshoot was 3%, and the control performance was excellent [ 4 ].…”

“…However, the initial value of the weight of the traditional BP neural network is randomly selected according to experience. This method leads to slow convergence of the network, which makes it easy to fall into the local optimal solution, and the final result has a large degree of instability [ 4 ]. Because the particle swarm algorithm has memory, it can transfer the memory of the best position of the particle in the history of the group to other particles and has the advantages of having fewer parameters to be adjusted and a simple structure, so the particle swarm algorithm was selected to weight the BP neural network.…”