基于人工鱼群牛顿迭代混合算法的架空光缆应力计算研究

Research on Stress Calculation of Overhead Optical Cable Based on Artificial Fish Swarm and Newton Iterative Hybrid Algorithm

在工程计算中,常常会碰到非线性方程组的求解问题,通常来说牛顿迭代法是比较好的选择,但是其对初值非常敏感,如果迭代初值选取不当,常常导致结果发散、不收敛。文章针对架空光缆应力方程在不同档距、不同温度下的求解问题,提出一种基于人工鱼群算法与牛顿迭代法的混合优化算法。首先将光缆状态方程转化为模函数,依靠人工鱼群算法(artificial fish school algorithm,AFSA)良好的全局寻优能力求得不同环境条件下的模函数的最优解,再将最优解作为迭代初值代入牛顿迭代法进行计算,最终求得光缆在不同环境条件下的应力结果。仿真结果表明文章所提混合算法有效地降低了结果发散概率以及减少了牛顿迭代法的迭代次数,可在实际工程中提供参考。

In engineering calculation,we often encounter the problem of solving nonlinear equations.Generally speaking,Newton iterative method is a better choice,but it is very sensitive to the initial value.If the initial value of iteration is not selected properly,it often leads to divergence and non-convergence.In this paper,a hybrid optimization algorithm based on artificial fish school algorithm (AFSA) and Newton iterative method is proposed to solve the stress equation of overhead optical cable under different spans and temperatures.Firstly,the state equation of the optical cable is transformed into a mode function,and the optimal solution of the mode function under different environmental conditions is obtained by relying on the good global optimization ability of AFSA.Then,the optimal solution is substituted into the Newton iterative method as the initial iterative value for calculation,and finally the stress results of the optical cable under different environmental conditions are obtained.The simulation results show that the proposed hybrid algorithm can effectively reduce the divergence probability and the number of iterations of Newton iterative method,which can provide a reference in practical engineering.