Cognitive Tuning Method Based on Case-Based Reasoning for PID Controller's Parameters

The proportional-integral-derivative (PID) tuning process has inherent problems, such as difficulty in identifying the model, a long adjustment time, and difficulty in determining the critical stable point. In this study, a cognitive tuning method is proposed based on case-based reasoning (CTCBR) for the parameters of the PID controller. Firstly, a new case-based reasoning (CBR) model with dynamic learning functions is designed, which adopts multiple attributes decision-making to improve the case retrieval strategy. Multiobjective evaluation criteria are then used to evaluate the expected effect after parameter tuning. Finally, a cognitive tuning method for PID parameters with a self-learning ability is obtained. Compared to other typical methods, an improved performance control is achieved, which improves the adaptability and tuning success rate of a PID controller system.