Water quality prediction method based on river polymorphism calculation and Arima model

Abstract

The invention discloses a weight combination water quality prediction method based on a river course polymorphism calculation and Arima model. Firstly, historical water quality index monitoring data of river monitoring station is obtained, and predicted values of water quality data of next time period of river monitoring station are taken as initial predicted values based on Arima model; the predicted values of water quality data of next time period of river monitoring station are taken as initial predicted values. Then, based on the river space shape and the water quality index measured by the upstream monitoring stations, the flow ratio in the river is calculated, and the predicted value of the downstream water quality index is calculated. The spatial form includes tributaries and confluence; finally, the two predicted values are calculated by using the weight combination optimization method, and the final predicted values are obtained. Compared with the prior art, the invention notonly utilizes the characteristics of the data itself, but also considers the spatial relationship between the tributaries and the main stream, and the average error of the combined prediction value based on the Arima prediction value and the river polymorphism calculation is small, which has better accuracy and is suitable for the optimization of the prediction algorithm of the tributaries and theconfluence state.

Description

technical field [0001] The invention relates to a weighted combination water quality prediction method based on river multi-state calculation and Arima model, which belongs to the technical field of water environment monitoring. Background technique [0002] The prediction of river water quality is of great significance for real-time monitoring, early warning and early prevention of river water quality. Studies by scholars have shown that the diffusion of river water quality has nonlinear characteristics. The current concentration of water quality is not a simple gradual diffusion of the sum of concentrations, but a random disturbance with the river itself and the internal and external forces of the river. Most of the conventional methods are based on monitoring Based on the endogenous law of point historical data, the water quality prediction algorithm based on the Arima model is one of the conventional methods. The model first judges the steady state and unsteady state of ...

AI Technical Summary

Problems solved by technology

However, for the rivers in the south of the Yangtze River with many tributaries, the current research is relatively lacking, and there is no clear research conclusion on the spatial relationship between the tributaries and the main stream.

Images