Application of PID neural network in headbox multivariable decoupling control

摘要

Aiming at the control problems of headbox in papermaking process, such as the nonlinearity and serious coupling relationship between pressure and liquid level, a kind of PID neural network decoupling controller is designed. The experience data of PID control is used to determine the initial value of the weight and make the controlled variables vary with the given values, which realizes the decoupling control of the multivariable system. The simulation results show that the PID neural network decoupling controller meets the requirements of the control performance index, which has high practical value.