GPU-Accelerated Batch Solution for Short-Circuit Current Calculation of Large-scale Power Systems

摘要

With the growing scale of the power systems, short-circuit current calculation becomes one of the most important, but intensive computational tasks for the power system planning and operation. Calculating the large-scale short-circuit current accurately and efficiently has great significance in practice. Graphics processing unit (GPU) has great potential in power system applications due to its superior performance in float-pointing calculation and memory bandwidth. To actively adapt to the unique characteristics of GPU’s software and hardware, a batch solution for short-circuit current calculation is designed and implemented by parallelizing kernel algorithms without compromising computational accuracy. Large-scale test systems are simulated to verify the accuracy and efficiency. Especially, compared with the counterpart CPU implementation, the proposed method can achieve about 36 times speedup ratio for the 10000-bus power system.