School of Mining Engineering,Taiyuan University of Technology
矿区开采引起的残余下沉稳定时间长、潜在危害大,有必要准确地预测矿区地表的残余下沉。鉴于传统的残余下沉Verhulst模型建模误差大、适用性弱,在建模过程中以数据序列的首个数据保持不变导致预测效果差的缺陷,以直接离散Verhulst模型为基础,引入粒子群算法寻求模型迭代初始值的最优解,建立基于粒子群算法优化的矿区开采残余下沉直接离散Verhulst模型,并以山西阳泉和山东兖州矿区两个时间尺度的地表残余下沉监测数据集进行实例验证,最后利用Matlab App Designer工具实现模型算法的可视化。结果表明：基于粒子群算法优化的直接离散Verhulst模型的矿区开采残余下沉预测精度和稳定性增益明显,所开发的计算工具具有正确性和有效性。
The residual subsidence caused by mining has a long stabilization time and great potential harm. It is necessary to accurately predict the residual subsidence value of the ground of the mining area. In view of the large modeling error and weak applicability of the traditional residual subsidence Verhulst model, the first data of the data sequence is kept unchanged in the modeling process, which leads to the poor prediction effect. Based on the direct discrete Verhulst model, the PSO (particle swarm optimization) algorithm is introduced to find the optimal solution of the initial value of the model iteration, and the direct discrete Verhulst model of mining residual subsidence based on PSO is established. The surface residual subsidence monitoring data sets of two time scales in Yangquan, Shanxi and Yanzhou, Shandong are used for example verification. Finally, the visualization of the model algorithm is realized by using Matlab App Designer tool. The results show that the prediction accuracy and stability gain of residual subsidence in mining area based on direct discrete Verhulst model optimized by particle swarm optimization are obvious, and the developed calculation tool is correct and effective.
力帆,廉旭刚,韩雨. 基于粒子群算法最优化Verhulst模型的开采残余下沉预测[J]. 科学技术与工程, , ():复制