News

Prefers to Proportion, I refers to Integral, and D refers to Differential.

In the motor speed control system, the input signal is positive. When the motor is required to rotate forward, the feedback signal is also positive (in the PID algorithm, error=input-feedback). At the same time, the higher the motor speed, the larger the feedback signal. In order to understand the principle of PID algorithm, we must first understand the meaning of each P, I, D and control law:

The proportional part is actually a large multiple of the difference between the preset value and the feedback value. For example, if the voltage across the motor is U0, the ratio P is 0.2, the input value is 800, and the feedback value is 1000, then the voltage and strain output to the two ends of the motor is U0+0.2*(800-1000). In order to achieve the purpose of adjusting the speed. Obviously the greater the ratio P, the faster the motor speed returns to the input value, and the higher the adjustment sensitivity. Thus, increasing the value of P can reduce the time from unsteady state to steady state. However, it may also cause the motor speed to oscillate around the preset value, so the integration I is introduced to solve this problem.

The integral part actually accumulates the time difference between the preset value and the feedback value. When the difference is not large, it does not cause oscillation. You can let the motor continue to run at its original speed. At the time, this difference was added to the integral term. When this sum is added to a certain value, it is processed once again. In order to avoid the oscillation phenomenon. It can be seen that there is a significant lag in the adjustment of the integral term. And the larger the I value, the more pronounced the hysteresis effect.

The derivative part is in fact the rate of change of the motor speed. That is, the difference between the two before and after the difference. In other words, the derivative term is based on the rate of change of the difference, giving a corresponding adjustment action in advance. It can be seen that the adjustment of the derivative term is advanced. And the greater the D value, the more pronounced the advance effect. The oscillation can be buffered to some extent. The role of the proportional term is only the magnitude of the amplification error, and what is currently needed to increase is the “differential term”, which can predict the trend of error variation. In this way, a controller with proportional + differential can control the suppression error in advance. Equal to zero, or even negative, thus avoiding serious overshoot of the controlled volume.