基于改进WOA-SVM的导水裂隙带高度预测

Prediction of the height of the water-conducting fracture zone based on improved WOA-SVM

鉴于支持向量机(SVM)预测导水裂隙带高度时,存在参数选取困难、收敛速度慢、易陷入局部最优等问题,提出一种基于全局搜索策略的鲸鱼算法(GS-WOA)优化SVM的导水裂隙带高度预测模型。首先,通过加入自适应权重、使用变螺旋位置更新策略、引入最优邻域扰动策略提升WOA算法的收敛速度、全局搜索能力和跳出局部最优能力;其次,采用改进后的WOA算法对SVM的参数进行寻优;最后,利用优化后的SVM对导水裂隙带高度进行预测,并与未改进的鲸鱼算法优化的SVM、遗传算法优化的SVM模型进行对比。结果表明,在多种评价指标下,改进的WOA-SVM模型具有较高的预测精度。

In view of the difficulties in parameter selection, slow convergence speed, and easy to fall into local optimization when support vector machine(SVM) is used to predict the height of hydraulic fracture zone, a prediction model of hydraulic fracture zone height based on global search strategy whale algorithm(GS WOA) optimized SVM is proposed.Firstly, the convergence speed, global search ability and ability to jump out of local optimum of WOA algorithm are improved by adding adaptive weight, using variable spiral position update strategy and introducing optimal neighborhood perturbation strategy.Secondly, the improved WOA algorithm is used to optimize the parameters of SVM.Finally, the optimized SVM is used to predict the height of the hydraulic fracture zone, and compared with the SVM optimized by the whale algorithm and the SVM optimized by the genetic algorithm.The results show that the improved WOA-SVM model has higher prediction accuracy under various evaluation indicators.