Design and implementation of an adaptive PID controller using single neuron learning algorithm

Qiang Sheng1, Zhuang Xianyi1, Wang Changhong1, X.Z. Gao2, Liu Zilong1
1Department of Control Science and Engineering, Harbin Institute of Technology, Harbin, China
2Institute of Intelligent Power Electronics, Helsinki University of Technology, Finland

Tóm tắt

In this paper, an adaptive single neuron-based PID controller for DC motor systems in vibratory stress relief equipment is proposed. Three inputs of the single neuron, which accord respectively with the proportion, integration and derivative of the feedback error, are associated with variable weights. The weight training algorithm is based on a combination of Delta and Hebbian learning rules. A motor speed control system with the WZ-86A DC motor was simulated using Matlab and Simulink, and further implemented on a microcomputer platform. Experiment results show that this system has satisfactory static and dynamical performances with strong robustness.

Từ khóa

#Algorithm design and analysis #Programmable control #Adaptive control #Three-term control #Neurons #DC motors #Stress control #Control systems #Neurofeedback #Hebbian theory

Tài liệu tham khảo

10.1109/60.41723 volkov, 1988, Vibratory stress relief of Titanium structures, Vibration Engineering, 59 10.1109/TSMC.1983.6313077 10.1109/21.260670 10.1109/ISIE.1992.279571 10.1109/72.207608 claxton, 1991, Vibratory stress relieving of welded fabrications, Welding Metal Fabrication, 541 hertz, 1991, Introduction to the Theory of Neural Computation klauba, 1983, Precise control of vibratory stress relief, Tooling & Production, 64