Real-time simulation test method of power system electromagnetic transient state based on cuda parallel computing
An electromagnetic transient simulation and electromagnetic transient technology, applied in concurrent instruction execution, machine execution devices, etc., can solve problems such as expensive prices, achieve the effects of model simplification, processing problem solving, and simulation speed improvement
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
Method used
Image
Examples
Embodiment 1
[0029] figure 1 As shown, CUDA decomposes calculations into thread kernels one by one: kernel_1, kernel_2...kernel_M, and each thread kernel is handed over to each grid Grid to complete, and the grid Grid divides the task into blocks one by one. Here, the blocks managed by each grid Grid are limited. The Block layer directly manages each thread Thread. When performing real-time simulation of the electromagnetic transient state of the power system, 2k calculations are allocated to different threads Thread. Go, because the number of CUDA calculation cores is very large, when 2k is less than this number, the calculation can be decomposed into more parallel calculations in each 2k calculation, by developing an algorithm suitable for the CUDA program execution model, and reasonably setting the Block And the number of Threads can make full use of the powerful parallel processing capability of GPU to solve the problem of processing electromagnetic transient massive data.
Embodiment 2
[0031] figure 2 As shown, the electromagnetic transient simulation of the power system must consider the electromagnetic transient process of the generator stator winding and the line, and all of them must be described by differential equations. These are linear ordinary differential equations, which can be written in the following form after simple transformation:
[0032] dX / dt=F(X)+G(U) (1)
[0033] X is the system state variable matrix
[0034] Using the two-point Euler method (the principle of other methods is the same), that is
[0035] x i n+1 =X i n-1 +(2h)(dX i n / dt) (2)
[0036] Substituting (1) into (2), we get
[0037] x i n+1 =X i n-1 +(2h)(F i (X n )+G i (U n )) (3)
[0038] In order to ensure the numerical stability of the calculation method, the prediction-correction method is used to solve equations (3) and obtain equation (4). Here, X ip n+1 predicted value, X icn+1 is the correction value
[0039] x ip n+1 =X i n-1 +2h(F i (X ...
PUM
Abstract
Description
Claims
Application Information
- R&D Engineer
- R&D Manager
- IP Professional
- Industry Leading Data Capabilities
- Powerful AI technology
- Patent DNA Extraction
Browse by: Latest US Patents, China's latest patents, Technical Efficacy Thesaurus, Application Domain, Technology Topic, Popular Technical Reports.
© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap|About US| Contact US: help@patsnap.com