Regional annual water consumption forecast model

Annual water consumption forecasting is critical to the optimal operation and reasonable allocation of water resource. Based on the summary of annual water consumption studies at home and abroad, with the principle of maximal precision and minimal error, and using data of Chinese water consumption in the past 10 years, this paper analyzed the feature and property of time series prediction method in water consumption study, predicted China's water consumption in 2015, 2020 and 2030 and evaluated its accuracy and error through power function model, linear function model, logarithm function model and parabolic function model, and its prediction accuracy and error were analyzed. The results demonstrated that the parabolic curve fitting correlation coefficient was the biggest and mean absolute percent error was the smallest, and the fitting effect was the best when conducting the curve fitting of water use in 2015, 2020 and 2030. Research results were basically consistent with those of the actual annual water use, which will provide references and basis for the prediction of the regional annual water consumption.