Ứng dụng mạng nơron hồi quy tổng quát và thuật toán nội suy đánh giá chất lượng nước mặt sông và các chi lưu sông Đồng Nai, tỉnh Bình Dương năm 2012 - 2018

Abstract

The Dong Nai River is one of the main water supply sources for living and production activities in Binh Duong province. However, the Dong Nai River is currently under pressure by emissions from industries, urban, agriculture and service activities. In this paper, the generalized regression neural network model (GRNN) and the interpolation algorithm was used to evaluate water quality of the Dong Nai river and tributaries. Monitoring data used was for 7 years from 2012 to 2018 at 12 monitoring stations. The results of the study indicated that the GRNN model was a good performance to assess the water quality of the Dong Nai river and tributaries. Comparing the results of the GRNN model with the water quality index showed that the GRNN model was reliable and in close agreement with the actual results with R2 = 0.938, RMSE = 0.055, E = 0.935. Water quality interpolation map also illustrated that the water quality in the neighbor area was still good. Besides, the GRNN model has a lower cost and less time consumption compared to the WQI method of the Vietnam Environment Administration.