Power market price prediction method, device, equipment, storage medium and program product

By combining data preprocessing and ARIMA modeling with over-limit correction and error monitoring retraining, an adaptive and updated electricity price prediction model is constructed, which solves the problems of inaccurate electricity price prediction and insufficient real-time performance in existing technologies, and realizes efficient electricity spot market price prediction.

CN122434591APending Publication Date: 2026-07-21SHENZHEN POWER SUPPLY BUREAU
0 Cites 0 Cited by

Patent Information

Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
SHENZHEN POWER SUPPLY BUREAU
Filing Date
2026-04-30
Publication Date
2026-07-21

AI Technical Summary

Technical Problem

Existing electricity spot market price forecasting models fail to fully incorporate various external factors, resulting in significant deviations between forecasts and actual market operations. Furthermore, they struggle to rapidly complete data processing and model inference within high-frequency clearing cycles, thus failing to meet real-time requirements and adapt to dynamic market changes.

Method used

By preprocessing data, performing ARIMA modeling, over-limit correction, and error monitoring and retraining, and by combining missing value imputation and outlier correction of multidimensional historical datasets, an adaptive and updated electricity price prediction model is constructed to achieve real-time electricity price prediction.

Benefits of technology

It significantly improves the accuracy and practicality of electricity price forecasting, meets the high-frequency clearing cycle requirements of the electricity spot market, provides quantifiable uncertainty information, and facilitates risk-controlled decision-making by market participants.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN122434591A_ABST
    Figure CN122434591A_ABST
Patent Text Reader

Abstract

The application relates to a power market price prediction method, device, equipment, storage medium and program product. The method comprises the following steps: acquiring a day-ahead market time-of-use price sequence and a real-time market time-of-use price sequence of a target power market, collecting historical data in a preset time period to form a multi-dimensional historical data set; performing missing value filling and abnormal value correction on the multi-dimensional historical data set; constructing an ARIMA data model according to the preprocessed data set; performing over-limit correction processing on the price prediction value output by the model, determining a prediction interval according to the corrected price prediction value, and taking the prediction interval as a price prediction result; determining an error value according to the price prediction result and an actual price; if the error value exceeds a preset error threshold, returning to the previous step and continuing to execute. The scheme realizes real-time price prediction conforming to market rules, having uncertainty expression and being self-adaptive and updated, and significantly improves the prediction accuracy and practicability while ensuring real-time performance.
Need to check novelty before this filing date? Find Prior Art