基于PSO-NP机械加工尺寸的PI型广义预测控制

摘要

For machining process of workpiece size error prone to defects, in this paper, through the testing data of machining dimension sequence for on-line modeling, this paper puts forward a kind of particle swarm optimization ( pso) algorithm and non-linear programming type PI constrained generalized predictive control ( PIGPC) , the algorithm with nonlinear programming processing machinery processing control system in the process of input and output constraints, resulting in type PI constrained generalized predictive control, the control law by particle swarm optimization ( pso) algorithm for the secondary search again to the constrained generalized predictive PI type of scaling factor and integral factor to optimiz-e the setting, through matlab simulation results show that the algorithm can accurate projections for processing the change of the size, mostly m-ake the processed parts size control within the permissible error.%针对机械加工过程中工件尺寸容易产生误差的缺陷，文章通过对机械加工尺寸的检测数据序列进行在线建模，提出了一种基于粒子群算法和非线性规划的带约束的PI型广义预测控制( PIG-PC)，该算法通过非线性规划处理机械加工过程中控制系统输入输出的约束条件，从而得到带约束的PI型广义预测控制的控制律，再通过粒子群算法进行二次搜索，对约束PI型广义预测的比例因子和积分因子进行优化整定，通过matlab仿真结果表明该算法能够对加工尺寸的变化进行准确的预测，使被加工零件尺寸基本上都控制在允许误差内。