Research And Implementation Of Data Cleaning And Power Quality Assessment

Stable and high quality electricity is the main driving force for the development of social science and technology and the national economy.In order to ensure the power system to operate safely and reliably,it is necessary to carry out the management of power quality.The premise of power quality management can accurately assess the power quality,find the power quality problems,and then put forward the scientific basis and actual data support for the management of power quality.The assessment of power quality depends on the large amount of data generated by the power system,but there may be a variety of problems and impurities in the data.Only by using the data cleaning technology,the impurity and the error can be removed to improve the efficiency of the power quality assessment and power quality management level.This paper firstly the main contents of the data cleaning technology.Using the "local Chebyshev theorem" and "proximity data average method" for cleaning attribute value error,using the ideas of "sorting neighbor algorithm" cleaning duplicate data,finally,the relatively reasonable statistical data are obtained.Secondly,the meaning and nature of the power quality are studied,and the meaning of the index and the calculation formula of the evaluation system are mainly studied.Thirdly mainly explains the meaning of the index,research the weight of power quality assessment corresponding to entropy method,variation coefficient method and analytic hierarchy process.But also to remedy and amend the subjective objective,using the improved gray target theory,integrated weight method based on rough set and information entropy and the least squares method are used to obtain the comprehensive weight.Then this paper adopt the three assessment methods based on the average method,Copeland method and fuzzy Borda method are used to combine the above assessment results and get the best assessment result.Finally,a case analysis of data cleaning and power quality assessment of a class A monitoring point for 110 kV.The six assessment methods and the three combined assessment methods mentioned in this paper can obtain the most accurate assessment results.The results contribute to the improvement of power quality assessment efficiency and accuracy.Having a strong promotional value.