Auto-tuning PID controller design using a sliding-mode approach for DC servomotors

Purpose: The proportional-integral-derivative (PID) controller has been a practical application in industry due to its simple architecture, being easily designed and its parameter tuning without complicated computation. However, the traditional PID controller usually needs some manual retuning before being used for practical application in industry. The purpose of this paper is to propose an auto-tuning PID controller (ATPIDC) which can automatically tune the controller parameters based on the gradient descent method and the Lyapunov stability theorem. Finally, a field-programmable gate array (FPGA) chip is adopted to implement the proposed ATPIDC scheme for possible low-cost and high-performance industrial applications, and it is applied to a DC servomotor to show its effectiveness. Design/methodology/approach: To ensure the stability of the intelligent control system, a compensator usually should be designed. The most frequently used compensator is designed as a sliding-mode control, which results in substantial chattering in the control effort. To tackle this problem, the proposed ATPIDCsystem is composed of a PID controller and a fuzzy compensator. The PID controller can automatically tune the gain factors of the controller gains based on the gradient descent method, and the fuzzy compensator is utilized to eliminate approximation error based on the Lyapunov stability theorem. The proposed fuzzy compensator not only can remove the chattering phenomena of conventional sliding-mode control completely, but also can guarantee the stability of the closed-loop system. Findings: The proposed ATPIDC system is applied to a DC servomotor on a FPGA chip. The hardware implementation of the ATPIDC scheme is developed in a real-time mode. Using the FPGA to implement, the ATPIDC system can achieve the characteristics of small size, fast execution speed and less memory. A comparison among the fuzzy sliding-mode control, adaptive robust PID control and the proposed ATPIDC is made. Experimental results verify a better position tracking response can be achieved by the proposed ATPIDC method after control parameters training. Originality/value: The proposed ATPIDC approach is interesting for the design of an intelligent control scheme. An on-line parameter training methodology, using the gradient descent method and the Lyapunov stability theorem, is proposed to increase the learning capability. The experimental results verify the system stabilization, favorable tracking performance and no chattering phenomena can be achieved by using the proposed ATPIDC system. Also, the proposed ATPIDC methodology can be easily extended to other motors.