改进的PSOGA-SVM模型应用于露天矿区空气质量预测

Application of improved PSOGA-SVM model to air quality prediction in open-pit mines

为掌握露天矿区开采所造成的空气污染情况,必须对露天矿区空气质量进行准确预测。选取PM10质量浓度与平均温度、相对湿度、光照时间、风力作为影响空气质量的主要因素;基于收集到的露天矿区环境数据,采用支持向量机(support vector machine,SVM)建立模型,同时引入改进型惯性权重的粒子群优化(particle swarm optimization,PSO)算法作为变异算子来优化遗传算法,最终将该模型应用于实际场景。基于MATLAB建立了改进型惯性权重的粒子群遗传算法优化支持向量机(improve inertia weighted particle swarm optimization and genetic algorithm based optimize support vector machine,PSOGA-SVM)网络预测模型。结果分析表明,所提模型的预测精度优于交叉算法验证算法优化支持向量机(cross-validation support vector machine,SV-SVM)模型和粒子群算法优化支持向量机(particle swarm optimization for parameter optimization of support vector machine,PSO-SVM)模型,且预测精度可达到98.5%以上。

In order to grasp the air pollution caused by the open-pit mines, it is necessary to accurately predict the air quality of the open-pit mining area. PM10 concentration with the average temperature, relative humidity, light time, and wind were selected as the main factors affecting the air quality. Based on the collected data of the open-pit mining environment, a support vector machine(SVM) was used to establish a model. Particle swarm optimization(PSO) algorithm with improved inertia weight was introduced as mutation operator to optimize genetic algorithm, and finally the model was applied to the actual scenarios. An improved inertia weighted particle swarm optimization and genetic algorithm based optimize support vector machine(PSOGA-SVM) network prediction model was established based on MATLAB. The results show that the prediction accuracy of the proposed model is better than that of cross validation support vector machine(SV-SVM) and a particle swarm optimization for parameter optimization of support vector machine(PSO-SVM), and the prediction accuracy can reach more than 98.5%.