隧道爆破参数优化中遗传算法和微分进化算法的应用比较

Application Comparison Between Genetic Algorithm and Differential Evolution in Tunnel Blasting Parameter Optimization

隧道爆破参数优化技术发展至今,已经有多种优化算法被应用到该领域,其中运用最广泛的就是遗传算法。微分进化算法作为一种新型优化算法,很少被用来进行隧道爆破参数优化,而本文以遗传算法和微分进化算法为优化工具对十白高速狮子沟隧道进行爆破参数优化分析,比较两种算法各自的优缺点。实例分析表明,遗传算法和微分进化算均能有效地解决复杂的隧道爆破参数优化问题,得到全局最优解。相比之下,遗传算法收敛速度更快,而微分进化算法编程相对更简单、参数设置更简便,所得解的精度较高等。

There are various types of optimization algorithmusing in tunnel blasting parametersoptimizing,among the rest,genetic algorithm is the most popular using.As a newtype of optimization algorithm,differential evolutionis rarely used in tunnel blasting parametersoptimizing,this paper will analyze the blasting parametersoptimizingof Shizigou tunnel in Shibai super highway,using genetic algorithm and differential evolution,comparing the advantage and disadvantage of both optimization algorithms.By practical analysis,each of optimization algorithms can figure out the complex tunnel blasting parametersoptimizing problems,obtaining the globallyoptimalsolution.In contrast,rateofconvergence is quicker in genetic algorithm,however,usingdifferential evolution is easier in programming、more convenient in parameter setting,and higher precision of solutions,etc.