Rolling prediction of single water quality parameter based on neural network

Abstract

The research of the single water quality parameter prediction has important directive significance to the evaluation, planning, early warning and management of water environment. This paper proposed a neural network rolling predictive method for the single water quality parameter. It integrates the rolling predictive control idea into the nonlinear neural network parameter modeling. The trend and the rule of the water quality in the future day can be well known on some historical measured data. The tests in the specific water area showed that, this method can be used for the single water quality parameter prediction in a rolling way.