A Study On The Environmental Quality Assessment Model Based On BP Artificial Neural Networks

Environmental quality assessment was one of the most important environmental managing institutions in China. It can make a scientific assessment for the environmental quality of a certain region, thus provide scientific foundations for environmental management, environmental engineering, environmental standards, environmental planning, comprehensive prevention and treatment, and ecological environment construction etc. Scholars both domestic and foreign have done many researches on methods for environmental quality assessment in the past years, but most of the methods lacks of maneuverability. However, the traditional assessing methods used in practical works had emerged a certain extent localization and irrationality. Thus, this paper tried to introduce artificial neural networks (ANN for short) theory into environmental quality assessment, set up an environmental quality assessment model based on Error Back Propagation arithmetic and to use it in practical works, thereby provides a new method for environmental quality assessment.In this paper, the BP-ANN environmental quality assessment model was improved in two aspects: firstly, the expert samples (training samples) of BP-ANN model were extended to improve model's robustness and veracity of distinguishing. That is, created 20 training samples randomly between samples based on environmental standards instituted by government. Secondly, the amount of nodes in network's covert layer was optimized using an arithmetic based on golden section theory to make training error of the model got to minimal value after less times cyclic iteration, and to improve the performance of BP-ANN model. Based on the improved arithmetic, BP-ANN model for air quality assessment and BP-ANN model for water quality assessment was built.