Adjustment Parameter Choices Of GPC Algorithms

摘要

Generalized predictive control (GPC) is the most popular of the prediction methods, especially for industrial processes. However, the tuning parameters (essentially a prediction horizon, a control horizon and a weighting parameter) have a determining effect on the behavior of closed loop system. But, it is not always easy to find parameters optimal values. In this paper we propose to use a genetic algorithm to optimize them.