摘要
为提高最小二乘支持向量机(LSSVM)预测模型的精度,准确预测煤炭开采成本.利用改进的自适应粒子群算法(IAPSO)的全局搜索能力,寻找LSSVM最优的惩罚因子r和高斯核函数的半径σ,提出一种IAPSO-LSSVM预测算法.在分析影响煤炭开采成本的空间因素、时间因素和定性因素的基础上,构建基于IAPSO-LSSVM的煤炭开采成本预测模型,并以TF煤业集团数据进行仿真实验.结果表明:与LSSVM、PSO-LSSVM算法相比,该模型预测效果更好.
In order to improve the prediction accuracy of least squares support vector machine （LSSVM） model, this paper used the global search ability of improved adaptive particle swarm optimization （IAPSO）, searched the most optimal r and σ, and put forward a IAPSO-LSSVM prediction algorithm. According to the factors affecting the coal mining cost, spatial, temporal factors and qualitative factors, this study established the coal mining cost forecasting model based on IAPSO-LSSVM and carried out the simulation experiment with the data of TF coal mining group. The results show that the proposed model is better than the LSSVM and PSO-LSSVM method.
出处
《辽宁工程技术大学学报（自然科学版）》
CAS
北大核心
2017年第5期554-560,共7页
Journal of Liaoning Technical University (Natural Science)
基金
国家科技支撑计划(2013BAH12F01)
