Statistical Evaluation and Prediction of Audible Noise of Hvdc Transmission Lines

Yong Yi,Zhengying Chen,Liming Wang
DOI: https://doi.org/10.1109/ceidp.2016.7785468
2016-01-01
Abstract:Audible noise of high voltage direct current (HVDC) transmission lines is a significant parameter for electromagnetic environment and has caused wide public concern. In order to offer a comprehensive statistical evaluation of audible noise for HVDC projects, a continuous test platform of audible noise considering various weather conditions has been built to record automatically. The long-term statistical measurements have been carried out based on the full scale HVDC operating transmission lines. The whole audible noise test platform has a low uncertainty, less than 1%. The data base is constructed using effective audible noise data. The prediction model of audible noise by using data fitting based on the linear mixed effect model (LME) which mainly involves data grouping, model fitting, parameters calibration and model validation. In addition, the least absolute shrinkage and selection operator algorithm is introduced to select the variables which are the influence factors of audible noise. Lastly, the calibration of prediction model is proposed to enhance the prediction accuracy.
What problem does this paper attempt to address?