利用遗传-粒子群优化混合算法求取剩余静校正量

剩余静校正是一个具有多参数、多极值的全局优化问题,当大量未知参数存在时,常规的遗传算法（GA）几乎无法避免早熟现象,难以保证收敛于全局最优。结合粒子群优化算法（PSO）和遗传算法的优势,提出了一种新颖的粒子群优化-遗传混合算法。混合算法利用了粒子群优化算法的速度和位置的更新规则,并引入遗传算法里的交叉变异思想。用混合算法和遗传算法分别对两个理论模型进行试处理,处理结果表明,混合算法比遗传算法具有更好的性能,是一种求取复杂地形条件下剩余静校正量的实用方法。

Application of PSO-GA Hybrid Algorithm to Residual Statics Correction

The residual statics correction is a global optimization problem with multi-parameter and multi-extreme values.The ordinary genetic algorithm(GA) is almost impossible to avoid the premature phenomenon because of the large number of unknown parameters,so it is difficult to convergence to a global optimum.This paper presents a PSO(particle swarm optimization)-GA hybrid algorithm(PGHA) based on the advantage of PSO and GA.PGHA combines the rules of updating the velocity and situation of PSO with the idea of cr...