Optimal control

Neural control, fuzzy control, decoupling control, and so on are all related to matrices. Which one?

I also want to know if anyone is familiar with multivariable control?

It refers to multivariable control, which is used for controlling systems with nonlinear characteristics, multivariable coupling, large time lags, and time-varying properties.

Advanced control methods all involve matrix operations and modeling, as mentioned earlier. However, they are not widely used in industrial settings because the actual operating conditions change in real time, and the mathematical models established are themselves simplified, which makes it impossible to determine the optimal solution. These advanced control algorithms or optimal control algorithms can work fine for simulations on computers, and they are useful in the aerospace industry, but no one uses them in industry. What can barely be used at the moment is predictive control: it’s convenient to establish a model for it, and it employs a rolling optimization algorithm which gives it good robustness; theoretically, it is superior to PID control. However, whenever the operating conditions change, a new model must be created. This modeling process requires multiple samples, and it can also affect the normal operation of the device. Modeling should only be done when absolutely necessary; therefore, in the end, PID remains the better choice! ! Hehe

For these controls, collectively referred to as advanced control, it is first necessary to be familiar with the process and proficient in control principles; models are developed through observation and experimentation, followed by configuration and tuning.

The American company ASPON has a dedicated predictive control software package that enables the completion of predictive control tasks by establishing mathematical models and testing data from on-site instruments