›› 2016, Vol. 29 ›› Issue (6): 26-.

• Articles • Previous Articles     Next Articles

Simulition of Multivariable Control Systems Based on Improved PID Neural Network

SONG Shuiquan   

  1. (Training Center, Huizhou Engineering Technical School, Huizhou 516001, China)
  • Online:2016-06-15 Published:2016-06-22

Abstract:

PID neural network has dynamic characteristics, and it can achieve better results than traditional PID control method in system control, but its learning algorithm is a gradient learning algorithm. The initial weights are random obtained. In order to improve the speed and system response time of the control system, we introduce the particle swarm algorithm to optimize the initial weights, and use Matlab software to simulate the improved PID neural network algorithm. The result of simulation reveals that the proposed method is better than traditional PID neural network control performance.

Key words: PID, neural network, multivariable control systems, swarm optimization algorithm

CLC Number: 

  • TP183