Deep learning model generation method for dynamic reactive power reserve demand calculation of power grid
A deep learning and model generation technology, applied in computing, reactive power compensation, circuit devices, etc., can solve problems such as long computing time and poor generalization ability, and achieve the effect of reducing solution scale and time complexity
Active Publication Date: 2021-05-28
HOHAI UNIV +2
View PDF4 Cites 1 Cited by
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
[0004] Purpose of the invention: Aiming at the problems existing in the prior art, the purpose of the present invention is to provide a method for analyzing the reactive power reserve demand of the power grid based on deep learning, and to calculate the reactive power reserve demand by simply using the transient time domain simulation software. The method was converted to the artificial intelligence method for automatic calculation, and the reactive power reserve demand analysis model based on deep learning was established, which solved the dynamic reactive power reserve demand model for maintaining the transient voltage stability of the power grid in the case of large disturbances in the transient time domain simulation calculation Technical issues of poor generalization ability and long computation time
Method used
the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View moreImage
Smart Image Click on the blue labels to locate them in the text.
Smart ImageViewing Examples
Examples
Experimental program
Comparison scheme
Effect test
Embodiment Construction
[0107] The implementation of the present invention will be described in detail below, and the technical solution of the invention will be described in detail in conjunction with the accompanying drawings.
[0108] A deep learning model generation method for power grid dynamic reactive power reserve demand calculation of the present invention, its process and sample generation method are as follows figure 1 shown, including the following steps:
[0109] Step 1: Construct the expected fault set, and construct the scenario set for different operation modes of the power grid.
[0110] Contains the following procedures:
[0111] Step 1-1: According to the needs of training the deep learning neural network, set the required number of scenes N, and use s n (n=1,2,...,N) represents N different scenes. The power output of new energy plants in the power grid will be P re and each node load demand P d are regarded as independent random variables, and there are a total of N R a rand...
the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More PUM
Login to View More Abstract
The invention relates to a deep learning model generation method for dynamic reactive power reserve demand calculation of a power grid. The implementation method comprises the following steps: constructing a scene set for different operation modes of a power grid; constructing a sample set which takes the scene set as input and takes critical dynamic reactive power reserve of each dynamic reactive power device of the power grid as output; and constructing a deep learning neural network, and performing parameter adjustment and training on the network. According to the invention, a power grid reactive power reserve demand evaluation model based on deep learning is established, changes of a power grid operation mode and a system reactive power reserve demand calculation method considering transient voltage stability constraints under various offline scene sets are considered, the minimum reactive reserve value required in the transient state of the power grid is quickly and effectively calculated by using an artificial intelligence algorithm, and a basis is provided for reliable and safe operation of a power system.
Description
technical field [0001] The invention relates to a power system security and control method, in particular to a method for generating a deep learning model for calculating dynamic reactive power reserve requirements of a power grid. Background technique [0002] In recent years, the scale of the interconnected grid has been increasing and the structure has become increasingly complex, and the operating state of the power system is gradually approaching its stability limit. After the disturbance, the ability to ensure the transient stability of the power grid further declines. Most serious grid stability accidents are uncontrollable cascading accidents caused by a sharp drop in the voltage of some nodes in the network. Taking the DC receiving-end grid as an example, when the receiving-end AC grid fails and the voltage of the converter bus drops, it may cause Simultaneous commutation failure of multiple DC circuits. During the restoration process after the fault is cleared, th...
Claims
the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More Application Information
Patent Timeline
Login to View More Patent Type & Authority Applications(China)
IPC IPC(8): H02J3/16H02J3/46G06Q10/04G06Q50/06
CPCH02J3/16H02J3/46G06Q10/04G06Q50/06H02J2203/20Y04S10/50
Inventor 赵晋泉赵泽麟陈睿张振安崔惟单瑞卿徐鹏
Owner HOHAI UNIV