Prediction model generation method, system and device, storage medium and prediction method

Abstract

The invention provides a prediction model generation method, system and device, a storage medium and a prediction method, which are used for flow prediction of a water supply network comprising a plurality of nodes. The prediction model generation method comprises the following steps: acquiring topological structure images and a plurality of historical traffic values of all nodes; constructing an adjacent matrix and a feature matrix, and obtaining a training set and a test set; performing model training by adopting a simplified graph convolutional neural network and a long-short-term memory network to obtain an initial prediction model; inputting the test set into the trained initial prediction model, and performing precision evaluation; if the precision reaches the standard, determining that the initial prediction model is the prediction model. According to the prediction model generation method, system and device, the storage medium and the prediction method provided by the invention, traffic prediction of a plurality of nodes can be performed at the same time; by extracting the spatial features and the time features of the node flow data of the water supply network and performing precision evaluation on the initial prediction model, the predicted flow value obtained by the final prediction model is more accurate.

Description

technical field [0001] The present invention relates to the technical field of traffic forecasting, in particular to a forecasting model generation method, system, equipment, storage medium and forecasting method. Background technique [0002] As a key component of the urban water supply system, the water supply network is the link between users and water resources, and is responsible for the important tasks of water delivery and distribution. The water supply network is known as the "lifeline" of the city and ensures the normal operation of the water supply network. It plays a decisive role in ensuring the development of the national economy and ensuring the daily life of residents; with the continuous expansion of urban water supply scale, the efficiency of water supply network operation, water power and water quality safety and stability are also increasingly challenged, so it is necessary to The flow of the water supply network is predicted to prevent failures. [0003]...

AI Technical Summary

Problems solved by technology

[0003] In recent years, with the development of smart water affairs, the basic information and operation monitoring data of the water supply network have been continuously improved. In the context of smart city construction, artificial intelligence technology provides a new solution for the flow prediction of the traditional water supply network. However, common time series forecasting methods such as autoregressive model, autoregressive moving average model, gray forecasting, etc. cannot learn the nonlinear characteristics of time series data; machine learning algorithms such as support vector machines tend to converge to local optimum; further , in order to improve the accuracy of prediction results, deep learning is also widely used. Although the existing water supply network flow prediction methods can achieve good results, most of them are based on a single flow node. The attribute characteristics of the time dimension of the data, while ignoring the correlation between nodes, that is, the spatial characteristics of the flow data of the water supply network

[0004] In summary, the water supply network prediction methods in the prior art have problems such as forecasting only based on a single flow node, ignoring the spatial characteristics of the water supply network flow data, etc.

Images