A Neural Network-based Approach for Assessing the Energy Value of Regional Water Environment Pollution Losses and Its Application

Abstract

With social development and population growth, people's demands for quality of life and quality of life are increasing, and higher requirements for drinking water safety have been put forward. Therefore, it is important to study and establish WE pollution loss assessment models for WE management work and reasonable economic policy formulation. However, most of the traditional models use a single factor as the input variable to assess the trend of pollutant concentration and the intensity of pollutant discharge; while the WE pollution loss assessment method based on energy value analysis theory uses a single factor as the input variable for energy value analysis to assess the intensity and concentration of pollutant discharge, but this method cannot handle multi-factor and multi-variable data, and can only use the fuzzy mathematics in However, this method cannot handle multi-factor and multi-variable data, and can only use fuzzy mathematics to evaluate and calculate the model; and it cannot deal with the non-linear relationship between multiple pollutants and single pollutant concentration at the same time. Therefore, in order to better solve the above problems, this paper proposes a neural network (NN) model based on the regional WE pollution loss energy assessment method; the method uses NNs to assess the energy value of a single factor; at the same time, the energy value of the regional pollution economic value is calculated to obtain the average energy value of the whole region; finally, the weighted average method is used to calculate the corresponding weighted average energy value of each unit and pollutant discharge intensity and other information. Finally, the weighted average method is used to calculate the weighted average energy value and the pollutant emission intensity of each unit. Compared with traditional methods, this method is not only better able to deal with multi-source data problems and non-linear relationships, but also more efficient in solving complex problems.