A power transaction output value intelligent accounting method and system
By constructing a situational awareness and multivariate prediction model, combined with two-layer optimization and deep reinforcement learning algorithms, the problem of maximizing output and profit for thermal power enterprises in a complex electricity market was solved, achieving high-precision transaction decisions and resource optimization.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Filing Date
- 2026-04-24
- Publication Date
- 2026-07-17
AI Technical Summary
The existing power trading decision-making system has failed to effectively adapt to the characteristics of thermal power enterprises, and is unable to maximize output and profit in the complex and ever-changing power market environment. It lacks in-depth perception of market conditions and real-time cost linkage analysis, resulting in unreasonable bidding strategies, inefficient resource allocation, and difficulty in meeting the refined decision-making needs of 15-minute power block trading.
By collecting multi-source heterogeneous data, a situational awareness model and a multivariate prediction model are constructed. Combined with a two-layer optimization model and a deep reinforcement learning algorithm, the system achieves full-dimensional prediction and multi-objective optimization of market indicators, generates trading strategies, performs compliance corrections, and ultimately completes the intelligent accounting of power trading output value.
It accurately captures nonlinear market fluctuations, reduces forecasting errors, increases trading profits, lowers energy costs, responds quickly to market changes, enables coordinated optimization of cogeneration and multiple units, and provides multi-granular, high-time-series market data support.
Smart Images
Figure CN122415152A_ABST