论文部分内容阅读
提出了一种训练模糊神经网络的加速进化规划算法。该算法是进化规划算法的一种改进 ,适合于多维高精度的数值优化。经仿真实例验证 ,其收敛速度有了明显的改善 ,且不易陷入局部极小。
An accelerated evolutionary programming algorithm for training fuzzy neural networks is proposed. The algorithm is an improvement of evolutionary programming algorithm, which is suitable for multi-dimensional high-precision numerical optimization. The simulation example shows that the convergence rate has been significantly improved, and not easy to fall into the local minimum.