Controller parameter self-adaption method, device and equipment in unsteady state and medium

The invention discloses a controller parameter self-adaption method and device in an unsteady state, equipment and a medium, and relates to the field of automatic production of intelligent control systems. The method comprises the following steps: acquiring real-time data of production; the real-time data comprises an environment variable and a material state variable; inputting the real-time data into a trained regression prediction model, and obtaining a steady-state phase manipulated variable set value according to the output of the regression prediction model; and based on the steady-state phase manipulated variable set value, the manipulated variable initial value, the total rising time length and the time interval between the current time and the manipulated variable change starting time, determining a real-time manipulated variable set value by utilizing a pre-constructed logic function, so that a controller operates according to the real-time manipulated variable set value. The manipulated variable is