Method, system, device and medium for phased prediction of leakage of sewage treatment system

Abstract

The application discloses a sewage treatment system leakage loss stage prediction method, system, device and medium, the method comprises the following steps: S1. Based on historical rainfall and sewage treatment quantity data lag test, get adaptive lag window; S2. According to the adaptive lag window, the rainfall and sewage treatment quantity data are analyzed, and the sample data in the dry season are obtained; S3. After the sample data in the dry season are divided into two stages, the two-stage prediction result is obtained by inputting the deep learning model, the two-stage prediction result is combined and evaluated, and the leakage loss prediction performance evaluation is obtained. The present application can realize high-precision prediction of the leakage loss of the sewage treatment system, thereby providing effective guidance for pipe network maintenance and repair, facilitating the positioning of potential leakage high-incidence areas and time periods, taking maintenance measures in advance, reducing the occurrence of sudden leakage events, and optimizing maintenance strategies and resource allocation, which can effectively reduce operating costs.

Images