摘要
以输电线路作为研究对象,针对输电线路的覆冰监测推导出覆冰厚度的理论计算解析式。由于传感器易受温度或其他外力影响,使得监测值产生误差。采用神经网络,利用覆冰历史数据,对覆冰厚度进行误差修正。实验结果表明:修正后的数值与实际覆冰值接近,覆冰厚度绝对误差的范围在0.000~0.087mm,相对误差值的变化范围在0.00%~6.56%。
Taking the transmission line as the object of study,the theoretical calculation and analytical formula of the ice thickness on the transmission line was derived.Considering the fact that the sensor is susceptible to the temperature or other external forces which resulting in an error in the monitored value,the neural network together with ice thickness’historical data were employed to correct the thickness of the ice coating.The experimental results show that,the corrected value becomes close to the actual icing value and the absolute thickness of the icing thickness ranges from 0.000mm to 0.087mm along with the relative error ranging from 0.00%to 6.56%.
作者
马泽楠
张长胜
田庆生
MA Ze-nan;ZHANG Chang-sheng;TIAN Qing-sheng(Faculty of Information Engineering and Automation,Kunming University of Science and Technology;Yunnan Electric Test&Research Institute Co.,Ltd.;Yunnan Electric Power Research Institute,Yunnan Power Grid Co.,Ltd.)
出处
《化工自动化及仪表》
CAS
2019年第12期1017-1022,共6页
Control and Instruments in Chemical Industry
基金
国家自然科学基金项目(KKGD201503106)
