基于遗传神经算法的压力钢管外压稳定性分析

Stability Analysis of Penstock under External Pressure Based on GA-NN Algorithms

压力钢管外压稳定性分析是重要而难以解决的问题。用遗传-神经网络模型对压力钢管外压稳定问题进行仿真计算,模型具有神经网络强大的函数逼近功能,同时利用遗传算法克服传统神经网络方法易陷入局部极小点的缺陷。实例仿真结果表明,用实数编码遗传神经网络模型分析加劲压力钢管的外压稳定性,具有效率高、鲁棒性好的优点,精度满足工程要求。

The stability analysis of the penstock under external pressure is an important and complex task.The model of GA-NN is proposed,the strong function approaching ability of neural network and inherit virtues are demostrated to break the limitation of local extreme in the traditional Neural Network.The experimental result indicates the efficiency,robustness and accuracy of the proposed algorithm.