A Nonlinear Dynamic Weight Particle Swarm Optimization Improvement Algorithm
In the traditional Linearly Decreasing Inertia Weight Particle Swarm Optimization(LDW-PSO)algorithm,the inertia weight is adjusted through a fixed linearly decreasing method.Although this approach is simple,it has limitations in adapting to specific prob-lems and reflecting the current state of the algorithm.To overcome these limitations,this paper introduces a power-law distribution function based on the linearly decreasing weight and proposes a new adaptive weight calculation method.This method allows the inertia weight to decrease gradually according to the nonlinear rule of the power-law function as the number of iterations increases,thereby enhancing the global search capability in the early stages of the algorithm and focusing more on local search in the later stages.Through this flexible weight adjustment,the improved method can effectively enhance the performance of the particle swarm optimization algo-rithm.To verify the effectiveness of the proposed method,this paper uses four benchmark test functions for performance analysis.Ex-perimental results show that,compared to the traditional linearly decreasing weight method and the standard particle swarm optimization algorithm,the improved algorithm demonstrates superior search capabilities and better convergence performance.
万方数据
维普
