Playing PID control with a single-chip microcomputer - from ideal PID control to advanced PID control_14

3.6 Fuzzy Adaptive PID Control

The principle of fuzzy adaptive PID control is shown in Figure 1. The general design process of fuzzy adaptive PID controller is: 1. Transform the deviation and deviation change rate from physical quantity to domain, 2. Determine the membership function of deviation e, deviation change rate ec, kp, ki and kd, as shown in Figure 2, 3. Determine the fuzzy rule, 4. According to the fuzzy rule and the membership function of deviation e and deviation change rate ec, calculate the query table of Δkp, Δki and Δkd, and the output surface of Δkp, Δki and Δkd is shown in Figure 2, 5. Real-time query, and transform Δkp, Δki and Δkd from domain to physical quantity, 6. Add Δkp, Δki and Δkd to basic kp', ki' and kd' to obtain actual kp, ki and kd values. The effect of fuzzy adaptive PID control for a large pure lag system is shown in Figure 3. It should be pointed out that when the domain of e and ec is a continuous quantity, Δkp, Δki and Δkd should also change continuously, that is, C'=A'×B'×R, Ri=Ai×Bi×Ci, R=∪Ri, and should not be a table. However, most papers output Δkp, Δki and Δkd in tabular form. Since the data storage device of the 51 single-chip microcomputer is too small and the calculation speed is limited, it is impossible to calculate the continuously changing Δkp, Δki and Δkd. Such an example will be given using MATLAB later.

4 Outlook

This thing can be regarded as a special topic, and it ends here. However, due to the limitations of hardware, including program memory, data memory and ports, some things cannot be done, such as neural network PID control, and a real regulator requires many configuration parameters. In the future, if possible, I would like to play with STM32F429. The original development board costs about 200 yuan, or play with Raspberry Pi. I don’t know if I can stick to it. Playing with programs is very tiring, and my cervical spine is not in good condition.