基于AWDO-SVR算法的石羊河流域生态系统健康评价

Evaluation of Ecosystem Health Based on AWDO-SVR Algorithm in Shiyang River Basin

针对流域生态系统健康评价问题,采用自适应风驱动优化(AWDO)算法和回归支持向量机(SVR)耦合算法对石羊河流域进行流域生态健康评价。针对支持向量机参数不易确定的问题,利用自适应风驱动优化算法寻找支持向量机的最优参数,提出基于AWDO-SVR算法的评价模型。利用该模型运算速度快、简捷,不需要赋权的优势对石羊河流域水资源禀赋、水资源开发利用、社会经济功能3个方面共9个指标进行评价。结果显示石羊河上游生态健康程度为亚健康,中游为病态,下游为病态。其评价结果与可变集模型评价结果一致,说明AWDO-SVR算法能够较好地应用于流域生态系统健康评价。

This paper evaluates the ecological health of Shiyang River Basin by the adaptive wind-driven optimization(AWDO)algorithm and support vector regression(SVR)coupled algorithm for problems in health assessment of watershed ecosystem,finds the optimal parameters of support vector machine(SVM)by AWDO algorithm for uncertainty of parameters from SVM,proposes an evaluation model based on AWDO-SVR algorithm,and evaluates nine indexes such as water resource endowment,water resource development and utilization,and social and economic function of Shiyang River Basin by the model with advantages of fast and simple operation and no need of weight.The results show that the ecological health is sub-health for the upper reaches of Shiyang River,and morbid for the middle and lower reaches respectively.The evaluation result is the same as that of the variable set model,indicating that AWDO-SVR algorithm can be effectively applied to the ecosystem health evaluation of the river basin.