提出了一种新颖的状态定义粒子群优化算法。该算法针对粒子群算法容易陷入局部最优和搜索精度不高的缺点,结合爬山算法和粒子群算法的特点,根据粒子状态的实时更新采用不同的搜索方法,在迭代过程中搜索到尽可能多的局部最优解,从而使算法可以更容易地跳出局部最优,更高效地搜索到全局最优解。对测试函数和非线性方程组求解问题进行实例仿真,仿真结果验证了算法的有效性,具有一定的实际应用价值。
It proposes a novel algorithm of mode definition particle swarm optimization. This algorithm solves the partial optimization and inaccurate search problem. It combines the characteristics of hill-climbing algorithm and PSO, uses different search method according to real-time particle mode to find as much local optimal solutions as possible during iteration. In this way, it can be easier and more efficient to search global optimal solution. Simulation examples of test functions and nonlinear equations solutions verify the efficiency of this algorithm.