| The data information of the power system is the basis of power system identification.The use of fault recording device is one of the main ways of recording system state.In order to meet the requirements of safety and high efficiency of the smart grid,the device in the power grid should timely determine the failure and communicate with the higher level system to remove the fault and make power supply security when the fault occurs.In this paper,a new type of recording device is designed,and the time series algorithm is used to analyze the acquired current data.Based on the analysis of domestic and foreign existed recording device,this paper designs an on-lined recording device with simple structure,high efficiency and low energy consumption,referencing to the technical requirements of recording equipment in China and international standard.The device uses MSP430 as the core of data processing,using electromagnetic mutual inductance coil to collect data signal of transmission line current.The device uses a thought of modular designing,which is made up with processing module,data acquisition module,power module,reset module,communication module and other parts.The software design of device follows the idea of modularization,and the main program and the subroutines of device are designed.The light weight and autonomous characteristic of device make it can be installed on the line of distribution network.With actual debugging and using,the device can record the circuit current and realize judging the fault.According to the characteristics of data acquired by device and the need for fault analysis,the paper decides using time series algorithm,and describes the identification,estimation and verification of three basic time series models chiefly.The paper uses MATLAB simulation software to analyze the current data of distribution network.The paper establishes a time series model based on the seasonal model,and the model is used for prediction and error analysis.By calculating the fault data and normal data of distribution network with time series model,and compared with the traditional data heuristic mutation criterion,the method is proved to be efficacious. |