Application Research of Multi-Model Fusion Technology in Real-Time Prediction of Dispatch Order Operation Duration

Abstract

Abstract Power grid dispatching is a challenging task that requires dispatchers to be forward-looking in multiple aspects, with the operation of dispatch orders being one of its core tasks. Currently, the prediction of dispatch order duration relies mainly on the manual maintenance of experienced dispatchers, which is not only labor-intensive but also demands a high level of expertise. As a result, the effectiveness of existing systems in this regard is difficult to guarantee. However, due to the complexity of business scenarios, there has been no deployment of systems based on artificial intelligence algorithms. To address this issue, this paper employs multi-model fusion technology to achieve real-time prediction of dispatch order operation duration and has deployed it in Zhejiang Electric Power Company. The main contribution of this paper lies in proposing a strategy to solve the problem of regression prediction after differentiating feature fusion. Additionally, the paper provides a detailed description of how the relevant algorithms were deployed in Zhejiang Electric Power Company and presents the results of practical applications, demonstrating the effectiveness and practicality of the proposed solution.