The Research On Load Balance Controller of Single Roller Transmission System of Rolling Mill Based on Load Observer

In the aspect of steel rolling, in view of the slow speed response and poor synchronization performance of the upper and lower roller drag motors in the single-roller drive system of rolling mill, a control strategy combining load balance controller and load observer based on single neuron adaptive PID neural network is designed. In this control strategy, the actual load of the upper and lower rolls observed by the load observer is used as the input of the load balance controller, and the addition of feedforward compensation can suppress the speed fluctuation of the motor. The simulation results show that the speed response time and synchronization error of the motor are improved obviously by adding load observer, and the response time of the motor to steady state before rolling is decreased by nearly 26.5%. When rolling, the response time of motor returns to steady state is decreased by 16.5%, the maximum absolute value of synchronization error is reduced by 78%, and the synchronization performance is effectively improved.