Design And Development Of Prediction Software On Voltage Quality

With the rapid development of our national economy and people’s living standards rising, not only put forward higher demands for the security, reliability, economy of power grid operation, but also make the requirements of the network voltage quality higher and higher. Therefore, in order to prevent the generation of voltage failure, it is necessary to predict the trend of the voltage quality and take preventative measures to the monitoring points whose forecast results cannot meet requirements. The prediction ensures users use electricity in a safe manner and the power system operate safely, and has a high application value to power system’s stability and security.At present, the research of analysis and management system on voltage quality is still in the preliminary statistics and simple summary report for voltage qualification rate, which calculate the upper limit rate,ultra-lower limit rate and qualification rate of voltage according to the day, month and year statistics of each monitoring point or the whole network. The analysis system does not reflect the distribution of voltage within the qualified range, nor do they reflect the degree of voltage fluctuation, but it will produce very serious consequences to electrical equipment or power system and even damage over the crash in excess of a certain limits. Besides, it also cannot meet the electric power authority and management’s demands of assessment and monitoring in data application levels. In this thesis, a research about the voltage quality prediction is done under the background of Ningbo Electric Power Bureau. The main content of this article is as follows:Firstly, the voltage indexes that reflect the voltage quality are determined. In order to assess voltage quality from multiple angles, voltage qualification rate that widely used at present must be the first choice. In addition, we choose voltage standard deviation and voltage probability density as other forecasting objects of voltage quality. Voltage standard deviation reflects voltage fluctuations during a period of time, whose value directly reflected voltage volatility relative to the ideal. The probability density curve of Voltage not only can get statistics of the fluctuation of voltage in a particular time, but also reflects the distribution of voltage in the acceptable range.Secondly, the method of Combined Forecast on voltage quality is researched. The method can forecast voltage quality depending on quarterlt、monthly、daily, which consists of the short-term、mid-term、long-term forecast. In addition, the amount of data that the voltage capture system collected may be more or less. Therefore, a single prediction model cannot meet the actual requirements. Accordingly, it is necessary to propose combination forecast model that is the combination of weighted array model and BP neural network model. Weighted array model composed by ARIMA model and GM(1,1)gray model predicts the value of the voltage indicator; the BP neural network model corrects the deviation of the weighted composite prediction model.Thirdly, the software of the voltage quality prediction is designed and developed. Based on the clear design objectives and design principles, the software design include the software architecture design, functional design, web-interface design, database design and interface design, its major functional modules are designed and developed, including the prediction module of voltage quality, the voltage quality management module, the database management module and the user login module. The front interface provides a multi-style display, such as tables, curves, bar graphs.Fourthly, the software of voltage quality prediction is verified. By the means of functional verification and performance verification of this software, the software meets the design requirements and performance requirements. Based on application verification, the voltage qualification rate, voltage standard deviation, voltage probability density are verified and analyzed, which assess voltage quality from the voltage level、voltage fluctuation and voltage distribution.