Risk Assessment of Sudden Water Pollution in Northwest Inland River Basin Based on IPSO-RBF Neural Network
The sudden water pollution accident damages the ecological environment and endangers human life and health,conducting risk assessment of sudden water contamination in inland river basins is particularly important for maintaining frag-ile ecological security in the western region.In response to the sudden water pollution problem in the northwest inland river basin,the PSR model was used to select 18 factors to establish the risk evaluation index system of sudden water pollution,and the risk evaluation model of sudden water pollution was constructed based on the radial basis neural network model(RBF).To further ensure the model accuracy,the neural network model parameters are optimized using the particle swarm algorithm(IPSO)with improved inertia weighting factors and learning factors.Finally,an IPSO-RBF neural network risk assessment model for sudden water pollution in the northwest inland river was established,using this model to evaluate the risk class of sudden water contamination in the Wuwei Section of Shiyang River Basin from 2017 to 2022.The results show that the risk level of sudden water contamination was level Ⅱ from 2017 to 2019,and level Ⅲ from 2020 to 2022.The results are consis-tent with the entropy weight TOPSIS method and the watershed governance situation.The research results are beneficial for improving the prevention and control level and emergency response capacity of sudden water pollution in Shiyang River Basin,and are of great significance for water resource management in the northwest inland river basin and ecological pro-tection in Qilian Mountains.
sudden water pollutionrisk assessmentRBF neural networkIPSO algorithmcontinental river basin
万方数据
维普
