【摘要】 作为一种先进的计算机控制算法,预测控制具有良好的自适应性和鲁棒性,一直是自动控制领域的研究热点.针对非线性系统,文中提出了一种基于菌群优化算法和模型学习的单步预测控制算法,算法在不需要系统数学模型前提下,根据非线性系统输入输出数据,通过最小二乘支持向量机建立系统数学模型,获得系统输出估计值,利用反馈校正减少预测误差,最后利用菌群优化算法进行滚动优化得到控制量,得到有效的预测控制器.由单变量和多变量非线性系统的仿真结果证明了算法鲁棒性和自适应性较好.更多还原

【Abstract】 As a kind of advanced computer control algorithm,predictive control has been a research hotspot in the field of automatic control because of its good adaptiveness and robustness. For the nonlinear system,a predictive control algorithm based on BSFO( bacterial swarm foraging for optimization) algorithm and model learning is proposed in this paper. According to the system input and output data,the algorithm establishes mathematical model by the LS-SVM( least squares support vector machine),forecasts the output value,and reduces the error through the feedback and error correction with unknown mathematical models. The control values are obtained by the rolling optimization of BSFO algorithm,and effective controllers are obtained. The simulation results for univariate and multivariate nonlinear systems show that the proposed algorithm has an excellent adaptive ability and robustness.更多还原