Predictive control algorithm based on soft constraints for multivariable systems

According to the problem that solving optimization problem affected by interference is prone to be infeasible in multivariable system model predictive control, a new algorithm based on soft constraints is proposed in this paper. In predictive horizon the bounds of manipulated variable constraints are user-defined, then user-defined constraints are softened and added into objective function to solve the optimization problem and make predicted output in the future period tend to the set value, so as to realize the effective control of multivariable system. The simulation results show that, the algorithm proposed has better set point tracking effect and anti-interference ability. © 2013 Binary Information Press.