摘要
为进一步提升0.4kV配网不停电可靠性评估,通过改进粒子群优化算法用于0.4kV配网不停电可靠性评估分析中。实验结果表明,PSO算法电网故障率预测的最大相对误差(12.742%)主要集中在迭代第27次,而利用GPSO算法进行可靠性评估,可以发现在相同迭代次数下,GPSO算法的相对误差仅为0.49%,较PSO算法降低94.15%。配网不停电可靠性检测指标中,GPSO算法的运行时间仅为0.124s。与PSO算法相比,GPSO算法的精度更高,平均运行时间和空间开销内存较PSO算法极大减少,且检测精度高达98.56%。GPSO算法能更准确地预测配网线路老化过程,修正后的累积老化误差仅为5.10%,较LSTM算法与ARIMA算法分别降低72.84%、83.26%。
In order to further improve the reliability evaluation of 0.4 kV distribution network without power outage,the improved particle swarm optimization algorithm was used in the uninterruptible reliability evaluation and analy⁃sis of the 0.4 kV distribution network.The experimental results showed that the maximum relative error(12.742%)of the PSO algorithm in predicting the power grid fault rate was mainly concentrated in the 27th iteration,while the reliability evaluation of the GPSO algorithm shows that the relative error of the GPSO algorithm was only 0.49%un⁃der the same number of iterations,which was 94.15%lower than that of the PSO algorithm.In the reliability detection indicators for uninterrupted power supply in the distribution network,the running time of the GPSO algorithm was only 0.124 seconds.Compared with the PSO algorithm,the GPSO algorithm had higher accuracy,significantly reduced average runtime and space overhead memory,and a detection accuracy of 98.56%.The GPSO algorithm could more accurately predict the aging process of distribution network lines,with a corrected cumulative aging error of only 5.10%,which was 72.84%and 83.26%lower than the LSTM algorithm and ARIMA algorithm,respectively.
作者
舒畅
赖飞伟
SHU Chang;LAI Feiwei(Guangdong Power Grid Customer Service Center,Guangzhou 510000,China;Guangdong University of Technology,Guangzhou 510643,China)
出处
《粘接》
CAS
2024年第10期165-168,共4页
Adhesion
关键词
粒子群优化算法
配网不停电
可靠性评估
particle swarm optimization algorithm
no power outage in the distribution network
reliability assessment
