Study on Neural Network ControlMetod for Proportional Valve Based Pneumatic Servo System

朱春波,包钢,程树康,王祖温
DOI: https://doi.org/10.3321/j.issn:1004-132x.2001.12.025
2001-01-01
Abstract:In this paper a neural network PID controller for proportional valve based pneumatic servo system is designed .A neural network identifier (NNI) is used to approach the nonlinear dynamic system and adjust control parameters on-line. Experiments and analysis indicate that through proper choosing the network parameters and training the network off-line sufficiently the controller can be used to control the plant adaptively on-line, and the control accuracy and dynamic characteristics of the system is increased obviously. Moreover the controller has ability of self learning and self adjusting control parameters when the condition varies.
What problem does this paper attempt to address?