A method for generating a reservoir optimal dispatching decision model based on depth learning

Abstract

The invention discloses a method for generating a reservoir optimal dispatching decision model based on depth learning, and relates to the the field of reservoir dispatching technology. Based on the Depth learn algorithm, the historical operation rules of reservoirs are extracted from the historical operation data of reservoirs, while optimizing the computational accuracy of the reservoir dispatching model, the computational time-consuming of reservoir operation model is optimized, The method can quickly extract reservoir dispatching rules from massive historical dispatching data, and is suitable for real-time dispatching of reservoirs and short-term, medium-term and long-term dispatching of reservoirs, and improves the calculation precision, calculation efficiency and practicability of the reservoir dispatching model, and solves the defects of low calculation precision, low calculation efficiency and low practicability existing in the prior art.

Description

technical field [0001] The invention relates to the technical field of reservoir dispatching, in particular to a method for generating a reservoir optimal dispatching decision model based on deep learning. Background technique [0002] Reservoirs, as an important water conservancy engineering measure for human utilization and management of water resources, have effectively resolved the contradiction between water resource allocation and human economic and social development needs. As a powerful tool to guide the operation of the reservoir, the reservoir operation plan is one of the key technologies to realize the comprehensive benefits of the reservoir. Therefore, it is very important to formulate a scientific reservoir scheduling plan for fully utilizing the comprehensive benefits of the reservoir. [0003] At present, the reservoir dispatching work is mainly guided by dispatching diagrams or dispatching models (such as linear functions, machine learning algorithms), both ...

AI Technical Summary

Problems solved by technology

[0005] (1) The ability to learn the historical operation and management experience of the reservoir is limited, and the decision-making results given deviate greatly from the actual operation;

[0006] (2) The actual scheduling scenarios that the reservoir needs to deal with are complex, and it is difficult to give reasonable and accurate scheduling decisions when dealing with extreme inflow conditions;

[0007] (3) Most of them only focus on the simulation accuracy, and pay insufficient attention to the time-consuming calculation, which limits the practical application of the model;

[0008] (4) Since there are many factors affecting reservoir scheduling, and there is a complex nonlinear relationship between the influencing factors and reservoir scheduling decisions, the difficulty of solving the scheduling model and the time-consuming calculation are greatly increased. However, it is rarely used in the real-time scheduling of reservoirs because it is difficult to meet the rapid decision-making requirements of real-time scheduling of reservoirs.

Images