Research On State Estimate And Error Detection Of Line Loss In Urban Distribution Network

The function of distribution network is to connect the transmission net work and end-users. With the rapid development of electric automation technology, the power supply company's demand of electric automation system becomes much higher. End-users'demand also becomes much higher. As our country has stated the project of energy-saving and ejection-decreasing and the plan of smart grid, how to decrease the line loss in urban distribution network becomes a critical subject for power supplier. The line loss in Shanghai consists of two parts. One is primary line loss, which refers to the line loss in transmission network. The other is district line loss, which refers to the line loss in distribution network. The line loss can be classified into two classes, the statistic line loss and theory line loss. The statistic line loss is calculated from the measurement data of power meter. The theory line loss is calculated from the parameters and operation mode of power grid.The power supplier in Shanghai has implemented the line loss delicacy project for the purpose of decreasing the line loss in urban distribution network. This project is to find the causes of line loss through delicacy comparison of different lines'line loss. The main job of line loss management has transformed from theory line loss to static line loss comparison. But the data quality of energy measurement couldn't be guaranteed during the operation of the project. The problems are as followers,1. the data loss caused by the communication breakdown2. the abnormal data caused by the mistake in system parameter maintance 3. the abnormal measurement data casued by the fault of measurement device.In this paper, a state estimate algorithm based on weighted least square algorithm will be discussed. According to the statistic characters of line loss, an error detection based on hypothesis testing will also be discussed. The main researches of this paper are as followers.1. The analysis on the mathematic mde of the usual electric components will be made. Some optimization work will be done according to the distribution network's specialization.2. The analysis on different data from three different systems, which have different sampling period, will be done. The research of how to convert different type of data to same calculation will be made.3. Three state estimate algorithms, weighted least square algorithm, rapid decomposition algorithm and measurement transform algorithm will be discussed. The weighted least square algorithm wiil be choosen as the state estimate algorithm of urban distribution network, due to the measurement speciality of the distribution network automation system. A state estimate calculation will be discussed.4. Research of the statistic characters of line loss will be made. An error detection based on hypothesis testing will be discussed. And An error detection based on real date will be discussed.