多策略PSO-BP算法在尾矿坝监测系统中的应用

Multi-strategy PSO-BP algorithm application on the monitoring system of tailings dam

BP网络能较好地解决尾矿坝监测系统的预报问题,但容易陷入局部极小值;PSO算法具有全局优化的特点,对PSO算法改进后更加突出了这一特性,将改进后的PSO算法与BP网络结合弥补BP网络的缺陷,从而得到全局最优解。把多策略PSO-BP算法用在尾矿坝的监测系统中,通过以往收集的数据,对尾矿坝体上各个测点的水位规律进行模拟,了解坝体浸润线规律,并预测下一级坝体的浸润线情况,为尾矿坝的稳定性提供依据。以齐大山尾矿库为例,证明了该算法应用在尾矿坝的监测系统上能取得较好的效果。

BP network can solve the problem of forecasting tailings dam monitoring system, but it is easy to fall into local minima; PSO algorithm has the characteristic of global optimization, the feature become more prominent after improving the PSO algorithm, combining the improvement of PSO algorithm with BP network make up deficiencies of BP network and get the optimal solution. The multi-strategy PSO-BP algorithm used in the railings dam monitoring system, by the data collected in the past, simulate water level laws of measurement points on each railings dam body, learn about the rule of dam seepage line, predict the next level of the dam seepage line situations, provide the basis for the stability of railings dam. Qidashan tailings dam, for example, to verify the algorithm used in the monitoring system of tailings dam can achieve good results.