基于粒子群-支持向量机(PSO-SVM)的苯储罐泄漏事故风险预测

将支持向量机(SVM)模型运用于事故前苯储罐泄漏事故风险预测,为使模型性能最优, 用粒子群算法PSO优化SVM模型参数,建立了PSO-SVM风险预测模型。为验证模型风险预测性能,分别采用遗传算法(GA)和网格搜索法(GS)优化SVM参数,并比较测试集与PSO-SVM、GA-SVM、GS-SVM三种模型预测结果的均方误差及相关系数。然后进一步探讨模型中权重调整方式、种群规模对PSO-SVM模型预测性能的影响。研究发现,权重线性递减所建PSO-SVM预测值与测试集相关系数更高、均方误差更小、预测效果更好,种群规模没有影响PSOSVM模型预测值但会影响计算时间,这为危化品泄漏事故的风险预测提供了一种新的方法。

Risk prediction of benzene storage tank leakage accident based on particle swarm optimization-support vector machine (PSO-SVM)

It is proposed to apply the support vector machine (SVM) model to the risk prediction of the benzene tank leakage accident before the accident. To optimize the performance of the model, we use the particle swarm algorithm PSO to optimize the SVM model parameters, and establish the PSO-SVM risk prediction model. In order to verify its model risk prediction performance, we used genetic algorithm (GA) and grid search method (GS) to optimize SVM parameters, and compared the test set with the prediction results of mean square error and correlation coefficient for PSO-SVM, GA-SVM and GS-SVM.. Then further explore the influence of the weight adjustment method and population size on the prediction performance of the PSO-SVM model. The study found that the PSO-SVM predicted value built by linearly decreasing weight has a higher correlation coefficient with the test set, a smaller mean square error, and a better prediction effect. The population size does not affect the predicted value of the PSO-SVM model but affects the calculation time. This is The risk prediction of hazardous chemical leakage accidents provides a new method.