嵌入共轭梯度算子的遗传算法

分析病态线性方程组的机理,将原线性方程组的求解问题转化为一个等价变分问题的极少值点寻优问题。在遗传算法产生的子代群体的个体以固定的概率采用共轭梯度法产生新子群,即采用共轭梯度法在局部进行搜索。将共轭梯度法局部搜索能力与遗传算法全局搜索能力有机结合,从而实现了混合算法的优化。算例结果表明,该算法对于病态方程组的求解效果明显优于一般的遗传算法和共轭梯度法。

Genetic Algorithm with Conjugate Gradient Operator

By analyzing the machanic（mechanism） of ill - conditioned hnear equations, this paper change the problem of solving the linear equations into the problem of the equivalent differencial by searching the most optimation solution. In the genetic algorithm, the individuals of the generation group generate new generations with conjugate gradient method by fixed possibility, namely searching locally by conjugate gradient method. The local searching ability of the conjugate gradient method combined with the global searching ability organically, then the opfimation of mixed algorithm is realized. The numerical experiments shows（show） that the algorithm is much more efficient than the general genetic algorithm and conjugate gradient method.