Research On Information Processing Method And Dynamic State Estimation Technology Of Hybrid Measurement In Distribution Network

At present,the information source of distribution network is mainly based on Supervisory Control And Data Acquisition(SCADA).However,with the improvement of the level of power distribution automation,higher requirements have been put forward on the time delay,accuracy,and frequency of measurements.Therefore,GPS-based synchronous phasor measurement unit(PMU)is gradually applied to the distribution network.Its high-frequency,high-precision,low-latency,and complete measurement characteristics provide a wealth of information for the distribution network,and also lay a solid data foundation for the dynamic state estimation of the distribution network.But in the short term,PMU cannot replace SCADA to achieve the observable requirements of the entire network.Therefore,the hybrid measurement composed of PMU and SCADA is an effective method to improve the state estimation and other advanced applications with a single SCADA measurement.However,the two types of measurement have large differences in sampling frequency,delay,measurement structure,time scale,etc.Therefore,preprocessing of the mixed measurements is a necessary prerequisite for improving the performance and accuracy of the state estimation of the distribution network.At the same time,due to the high-frequency characteristics of the PMU,it is also necessary to compress the massive PMU data.Therefore,this paper studies the problems in the unified use process of hybrid measurement and the dynamic state estimation of distribution network based on hybrid measurement.The main work is as follows:1)Based on the existing research,the characteristics of PMU / SCADA hybrid measurement are analyzed,and the research status of asynchronous data synchronization processing,mass data lossless compression processing,and distribution network state estimation are briefly introduced.2)Aiming at the problem that the two types of measurement are asynchronous due to the sampling frequency and time delay between PMU / SCADA mixed measurements,the characteristics of mixed measurements are analyzed,and an asynchronous data synchronization method is proposed based on the Spearman correlation coefficient method;Uncertainty propagation method models the uncertainty of the line parameters to spread the synchronization function to neighboring nodes.Aiming at the storage pressure brought by PMU high-frequency measurement,a pre-processing based LZW compression algorithm(PBLZW)based on measurement preprocessing is proposed.After pre-processing the measurement,the data redundancy is reduced initially,and the data is rotated through the Burrows-Wheeler invertible transformation(BWT).Finally,the lossless compression of the measurement data is achieved by the LZW lossless compression algorithm.3)The dynamic state estimation method of distribution network based on the quadratic constraint quadratic estimation model is studied.A PMU / SCADA hybrid quadratic estimation model is established by adding quadratic constraints.In order to solve the problem of weak resistance of the UKF algorithm,a robust unscented Kalman filter(QCQE-UKF)based on the quadratic constraint quadratic estimation model is proposed.And an adaptive correction of the gain matrix is achieved by proposing a corresponding measurement identification method.