基于信赖域-遗传算法的雷电冲击试验波形参数估算  

Parameters Estimation of Lightning Impulse Test Waveform Based on Trust Region Method Combined with Genetic Algorithm

在线阅读下载全文

作  者:李忠诚 刘馨然 张穆昕 LI Zhongcheng;LIU Xinran;ZHANG Muxin(State Grid Liaoning Electric Power Company Measurement Center,Shenyang 110000,China)

机构地区:[1]国网辽宁省电力有限公司计量中心,沈阳110000

出  处:《电瓷避雷器》2022年第5期31-35,共5页Insulators and Surge Arresters

摘  要:合理的雷电冲击波形参数提取对于雷电冲击耐压试验的有效进行至关重要,针对带过冲和振荡雷电冲击波形处理过程存在的争议,本研究提出一种基于信赖域-遗传算法的雷电冲击波形平均曲线拟合方法,构造双指数函数形式波形平均曲线,采用信赖域算法优化拟合曲线误差基础上,引入遗传算法解决信赖域算法局部信赖域半径减小的缺点,分别对雷电冲击波形存在过冲、波前振荡、波峰振荡、波尾振荡情况下拟合效果进行验证。分析结果表明:信赖域-遗传算法能够较好拟合雷电冲击波形平均曲线,几种情况下拟合得到的波形基本参数均与原始波形参数较为接近。拟合效果来看,对于伴有波前振荡的曲线拟合效果最佳。引入遗传算法后,迭代计算收敛速度得到明显加快。Proper estimation of lightning impulse parameters is very important for the effective implementation of the lightning impulse testing. For the controversy on the treatment measures of lightning impulses with overshoot or oscillation, a curve fitting technique which combined the trust region methodwith genetic algorithm was proposed to determine the lightning impulse voltage waveform parameters. The mean curve was constructed using double exponential form and the fitting error was optimized by means of trust region method. The genetic algorithm was also introduced to solve the constraintofthe trust regionradius. The effectiveness of the proposed technique was examined on different cases of lightning impulses with overshoot or oscillation. Research results show that: the proposed technique is effective in estimating the peak value, front and tail times of the impulse waveform in all three cases. This method gives best estimation result especially in the lightning pulse with oscillations on the front. The results also show that the proposed technique converges faster than the traditional trust region method.

关 键 词:雷电冲击试验 电压波形 信赖域算法 遗传算法 双指数函数 

分 类 号:TM86[电气工程—高电压与绝缘技术]

 

参考文献:

正在载入数据...

 

二级参考文献:

正在载入数据...

 

耦合文献:

正在载入数据...

 

引证文献:

正在载入数据...

 

二级引证文献:

正在载入数据...

 

同被引文献:

正在载入数据...

 

相关期刊文献:

正在载入数据...

相关的主题
相关的作者对象
相关的机构对象