基于粒子群优化的在线支持向量回归预测控制方法

• 作 者: 陈进东 潘丰
• 机 构: 江南大学轻工过程先进控制教育部重点实验室,
• 出 处: 信息与控制
• 发表日期: 2013年06期
• 关键词: nonlinear model predictive control online support vector regression particle swa

针对非线性系统模型预测控制中,预测模型容易失配和目标函数难以求解的问题,提出了一种基于粒子群优化算法的非线性系统在线支持向量回归模型预测控制方法.该方法利用在线支持向量回归建立被控对象的非线性预测模型,并通过在线学习实现模型的在线自校正;同时采用粒子群优化算法求解目标函数,完成滚动优化.对非线性系统的仿真结果表明,该方法是有效的且具有良好的自适应性.

Aiming at the problem that the prediction model is easy to mismatch and the objective function is difficult to solve in nonlinear system model predictive control, a particle swarm optimization algorithm based on online support vector regression model predictive control is proposed. The method uses online support vector Regression to establish the nonlinear prediction model of the controlled object and online self-tuning of the model through online learning.At the same time, the particle swarm optimization algorithm is used to solve the objective function to complete the rolling optimization.The simulation results of the nonlinear system show that this method is effective And has good self-adaptability.