并联机构求正解 用遗传算法来对神经网络算法的权值选取过程进行优化，既能缩短训练时间，提高网络收敛速度，又可以有效避免训练中陷入局部最优解。(According to solving results, the solving precision of neural network algorithm which is optimized by genetic algorithm is able to reach nanoscale and has a higher solving precision and a faster solving speed than Newton iteration method and neural network algorithm.)