DWT-PCA-LSTM-based water supply quantity prediction device for water supply company

Abstract

The invention discloses a water supply company water supply prediction device based on DWT-PCA-LSTM, and the device comprises a data collection module which is used for generating a time sequence of daily water demand data through collecting the daily water supply data of a water supply company in a city, meteorological data corresponding to the daily water supply, holiday and workday conditions;a data preprocessing module used for removing abnormal values of the daily water supply data time sequence and reducing noise of the daily water supply data time sequence; a prediction variable determination module used for carrying out correlation analysis on the time sequence of the daily water demand data obtained by the data preprocessing module, calculating a residual error sequence accordingto the time sequence of which the abnormal value is removed and a noise reduction sequence, and preliminarily determining a prediction variable input into the prediction model; and a data predictionmodule used for completing prediction of the future water supply amount by establishing an LSTM neural network model. The daily water supply amount of the water supply company is predicted through a deep learning method, and a water supply adjustment basis can be provided for the water supply company.

Description

technical field [0001] The invention relates to urban water supply monitoring technology, in particular to a DWT-PCA-LSTM-based water supply forecasting device for water supply companies. Background technique [0002] Accurate urban water demand forecasting plays a key role in building an urban smart water supply system, and an efficient and reliable smart water supply system can effectively promote the construction of a smart city. For water supply companies, decision-making activities related to water supply schemes, green energy conservation, and optimal use of water resources all depend on the forecast of water demand. At the same time, accurate water demand prediction results also help to improve the quality of water supply, which can minimize the residence time of water in the pipeline and improve the quality of domestic water for residents. [0003] Urban water demand data has strong nonlinearity and randomness. Traditional linear prediction methods based on mathemat...

AI Technical Summary

Problems solved by technology

However, due to the tendency of gradient explosion or gradient disappearance in the training process, the RNN model does not have the ability of long-term dependence, and it is easy to lose key information in the sequence

[0006] The above methods have solved the problem of time series prediction to a certain extent, but the historical data of urban water demand has strong random fluctuations and is affected by seasonal factors, and has a certain degree of periodicity, so the prediction effect of the above models is still difficult. satisfactory

Images