Multi-mode heat supply unit load distribution optimization method based on artificial neural network

Abstract

The invention discloses a multi-mode heat supply unit load distribution optimization method based on an artificial neural network. The method comprises the following steps of S1, building a database, and storing unit operation data and accessed meteorological data, and S2, building a BP artificial neural network model, calling data in the database as input data of the BP artificial neural network model, and predicting the heat supply load and the power supply load of the unit in the next 24 hours. The invention provides a multi-mode heat supply unit load distribution optimization method based on an artificial neural network. The method predicts a future load based on meteorological information and a historical load of a power plant; pre-adjustment and pre-planning scheme making are conducted on the power plant according to the predicted load; reasonable load distribution is conducted on the power plant unit according to the MEGC instruction and the predicted load; and according to the predicted load, the power plant is guided to make a decision, whether to participate in peak regulation bidding of the second day and determine the bidding price is judged.

Description

technical field [0001] The invention relates to an artificial neural network-based multi-mode heating unit load distribution optimization method, which belongs to the field of thermoelectric load prediction of thermal power heating units. Background technique [0002] At present, with the steady increase in urban heating demand year by year, the problem of insufficient heating capacity has become increasingly prominent, affecting people's livelihood security. To this end, on the one hand, power generation companies increase their heating capacity through equipment and technological transformation to meet more heating load demands; on the other hand, they participate in auxiliary services in the power market through flexible transformation, and actively respond to the power system peak-shaving compensation policies successively introduced by the state. According to the needs of different power supply loads and heating loads in different periods, how to quantify and guide the ...

AI Technical Summary

Problems solved by technology

However, due to the existence of multiple heating modes, the strong coupling characteristics of heat and electricity, and the differences in unit performance and the need for deep peak regulation, it is difficult for power plants to reasonably allocate and optimize the heat and electricity loads of different units based on operating experience.

[0004] 1. A calculation method and system for dynamically predicting the peak-shaving interval of thermal power units (application publication number: CN112465662A, application number: 202011091746.X). , providing a calculation method and system for dynamically predicting the peak-shaving range of thermal power units, which only considers the load prediction and peak-shaving space of the power grid itself, and does not discuss the peak-shaving mechanism under the background of thermoelectric coupling in heating units;

[0005] 2. A thermal and electrical load monitoring system for heating units based on GIS weather information (application publication number: CN209746598U, application number: 201920165213.8). The field of this patent is the thermal and electrical load monitoring of thermal power heating units. In this method, based on the collection The obtained GIS meteorological information is used to predict the heat load and monitor the electric load in the heating area. Based on the heat load prediction, it can provide certain guidance for the operator's peak-shaving ability of the power grid. This method does not predict the electric load at the same time but only Therefore, it has a limited role in predicting and guiding the actual power grid peak shaving capacity

Images