A self-tuning immune feedback controller for controlling mechanical systems

Summary form only given. The application of biological-information processing mechanisms to control systems promises greater flexibility and may make it possible to construct a control system whose performance is better than that of conventional control systems. Biological immune systems have learning, memory, and pattern-recognition abilities. The application of some of these abilities to control/sensing systems has been studied; we have focused on the immune feedback mechanism. An immune feedback mechanism simultaneously responds to foreign materials and stabilizes itself. We examine an engineering application of a biological immune system and propose an immune feedback controller. We propose an immune feedback law based on the functioning of biological T-cells; it includes both at active term, which controls response speed and an inhibitive term, which controls stabilization effect. We also describe a self-tuning immune feedback controller based on the immune feedback law whose parameters are automatically tuned by using neural networks. Experimental results for velocity tracking control of a DC servo motor confirmed the validity of our immune feedback law and also demonstrated the effectiveness of the self-tuning immune feedback controller for controlling practical systems.