Research on Intelligent Control Strategy of Plasma Cutting Process
By taking into account these problems such as the nonlinear volt-ampere property of plasma arc, the multi-parameter close coupling of cutting technology process and the difficulty to determine the optimal parameter, we present a control strategy based on the combination of RBF neural network and expert system. The RBF neural network and expert system can resolve a certain tape problems respectively. Mutual combination can take full advantages of the high logical reasoning ability of expert system and the good robustness and a strong learning ability of neural network, overcoming the disadvantages of the expert system as poor fault tolerance and learning ability. This control algorithm avoids the redundant modeling process of nonlinear system, having the ability of multi-parameter decoupling. Expert system has high adaptive ability and self-learning ability so it can achieve parameters trained by neural network during the reasoning process and then it obtain optimal output value, having quite strong guidance to productive practice.
multi-parameter close coupling RBF neural network expert system optimal parameter
Deli Jia Jinsong He
Harbin University of Science and Technology, Harbin 150080 China
国际会议
The 22nd China Control and Decision Conference(2010年中国控制与决策会议)
徐州
英文
3409-3413
2010-05-26（万方平台首次上网日期，不代表论文的发表时间）