基于BP-遗传算法的排土场边坡几何参数优化

Geometric Parameters Optimization of Dump Slope Based on BP Neural Network-Genetic Algorithm

在保证矿山安全生产的前提下,为发挥排土场最大经济效益,提出了基于BP-遗传算法的排土场边坡几何参数优化方法。以弓长岭大阳沟排土场为例,借助极限平衡法获取研究所需数据,利用BP神经网络建立边坡坡角、单段台阶高度及相应的安全系数间的非线性关系,并以此关系式为边界约束条件,建立了优化边坡几何参数的数学模型,利用遗传算法和传统优化算法进行寻优。结果表明,与传统优化算法相比,BP-遗传算法的优化结果更加精确、可靠,有效避免了传统优化算法在寻优时易陷入局部最优解的问题。提供了一种简单、精确、可靠的排土场边坡几何参数优化方法,具有较好的应用前景。

In order to obtain the maximum economic benefits of the dump while ensuring mine safety production,a method of geometric parameter optimization based on BP-genetic algorithm was proposed. With Dayanggou dump in Gongchangling Mine as an example,limit equilibrium method was firstly used to obtain a large amount of data. Then,BP neural network was used to establish a nonlinear relationship among slope angle, single bench height and corresponding safety factor,which was taken as the boundary restraint condition to establish a mathematic model to optimize geometric parameters of mine slope by using traditional optimization algorithm and genetic algorithm. Results show that compared with traditional optimization algorithm,BP-genetic algorithm can obtain more precise and reliable optimization result,overcoming the problem in traditional optimization algorithm. It is concluded that such a simple,precise and reliable geometric parameter optimization method shows a good prospect in application.