基于随机森林的中俄东线管道河沟道水毁灾害评估研究

Study on river channel water damage disaster evaluation of China-Russia east line pipeline based on random forest

为科学评估河沟道水毁灾害,以中俄东线管道河沟道水毁灾害数据为依据,研究河沟道水毁灾害的影响因素,采用随机森林算法,构建河沟道水毁灾害评估模型,并结合准确度和ROC曲线对其性能进行评价。研究结果表明:基于随机森林算法,模型训练采用7折交叉验证法,并采用网格搜索算法进行参数优化,优化后的模型准确率和泛化能力均较好(准确率80.95%,AUC值为0.89)。研究结果可为河沟道水毁灾害预防和治理提供辅助决策依据。

In order to scientifically evaluate the river channel water damage disaster,the influencing factors of river channel water damage disaster were studied based on the river channel water damage disaster data of China-Russia east line pipeline.An evaluation model of river channel water damage disaster was established by using the random forest algorithm,and its performance was evaluated by combing the accuracy and ROC curve.The results show that based on the random forest algorithm,the model is trained with 7-fold cross-validation method,and the model parameters are optimized with grid search CV algorithm.The optimized model has good accuracy and generalization ability(the accuracy is 80.95%,and the AUC value is 0.89).The research results can provide auxiliary decision-making basis for the prevention and management of river channel water damage disaster.