基于改进PSO算法的函数极值寻优研究

Recognition of Functions Extreme Optimization Based on Improved PSO Algorithm

针对许多工业生产过程的数学模型难以得到最优极值的问题,研究利用PSO算法实现对于复杂非线性函数的极值寻优。为了提高PSO算法的搜索能力和搜索精度,提出了一种惯性权重的自适应调整策略,并利用Sigmoid形式变异算子实现种群的重新初始化,避免早熟收敛。仿真结果证明了所提改进措施的正确性和有效性,改进后的PSO算法,在处理复杂非线性函数极值寻优问题时的能力大为提升。

Mathematical model for many industrial processes is difficult to get the optimal extremes, so the PSO algorithm was used to solve the problem of extreme optimization for complex non-linear function. In order to improve the search accuracy and capabilities of the PSO algorithm, a new adaptive adjustment strategy was proposed for the in- ertia weight, and Sigmoid form of mutation operator was used to achieve re-initialize the population, which avoids premature convergence. The simulation results prove the correctness and effectiveness of the proposed improvement measures, and the ability of the PSO algorithm improved is enhanced greatly when handling the problem of complex nonlinear function extreme optimization.