Liquid level control of water tank system based on improved polyclonal selection algorithm and RBF network

To realize the adaptive control of liquid level of a 2-container water tank system, a novel PID controller based on improved polyclonal selection algorithm and RBF network is put forward. Firstly, the initial PID parameters and their learning rates of controller are optimized using the improved polyclonal selection algorithm, which reduces the influence of control effect due to improper initial parameters. Secondly, the PID parameters of controller are adjusted on-line based on the RBF identification network during the liquid level control of water tank system. The simulation results show that the proposed controller is valid in the liquid level control of water tank system. Compared with other corresponding control algorithms, the proposed controller is characterized by quick response speed and small overshoot and error.