将一种改进的基于差分进化算法的非线性预测控制应用到酸碱中和反应pH值控制系统.算法充分利用滴定曲线模型,指导优化过程搜索的初始化空间.同时,在变异和选择等操作中改进了差分进化算法,解决了一类有边界约束的非线性优化问题.在发酵罐实验装置中进行了测试实验,取得了较好的效果.
Differential evolution (DE)-based nonlinear model predictive control (NMPC) is applied to a pH neutralization system. The algorithm fully makes use of the titration curve model to obtain the appropriate initial values of nonlinear optimization problem. Meanwhile, a modified DE algorithm is proposed to solve a boundary constrained optimization problem in the NMPC. The proposed method is validated in a real fermentation reactor, and the testing results show its effectiveness.