摘要
针对传统信息融合技术在煤矿井下环境等级评价中的局限性,文章提出了一种智能算法:通过粒子群优化算法对最小二乘支持向量机进行参数寻优,建立多传感器信息融合模型PSO-LSSVM,克服参数选择的主观性、盲目性,从而提高算法的分类精度和收敛速度。实验结果表明,相比未经参数优化的最小二乘支持向量机模型、网格算法优化最小二乘支持向量机模型,PSO-LSSVM模型能很好地解决煤矿井下环境等级评价中小样本的高维、非线性、不确定性等方面的问题。
For the limitation of traditional information fusion technology in the coal mine underground environment level evaluation, an intelligent algorithm was proposed in the paper: Particle Swarm Optimization algorithm was adopted to optimize the parameters of the least squares support vector machine, and a multi-sensor information fusion model PSO-LSSVM that overcomes the subjectivity and blindness on parameter selection was established for improving its classification accuracy and convergence speed. Experimental results showed that compared with the least squares support vector machine model not been optimized and the least squares support vector machine model optimized by the grid searching algorithm, PSO-LSSVM model could be a good solution on the issue of the high-dimensional, nonlinear and uncertainty of the small samples in coal mine underground environment level evaluation.
出处
《测绘科学》
CSCD
北大核心
2014年第7期150-154,共5页
Science of Surveying and Mapping
基金
国家自然基金项目-神华煤炭联合(51174257)
国家重点基础研究(973)计划项目(2010CB732002)
安徽高校省级自然科学研究重点项目(KJ2010A083)
安徽理工大学青年科学基金(2012QNY34)
关键词
信息融合
粒子群算法
最小二乘支持向量机
参数优化
交叉验证
information fusion
Particle Swarm Optimization
least squares support vector machine
parameter optimization
cross validation