Research On Key Techniques In Cyber-physical Power System

With the improvement of intelligent level in power system,the information and communication technology plays an increasingly important role in the power system,which urges the traditional power grid towards to cyber-physical power system.Cyber-physical power system realizes the information acquisition and perception of power system through the massive heterogeneous intelligent terminals,and transmits the massive multi-source heterogeneous monitoring data to the upper layer through the communication layer to make the decision,thereby implementing the upper level's instruction to realize the optimal operation of the power physical system.The cyber-physical power system promotes the deep integration of the primary system and the information system,and finds a new way to achieve the goal of smart grid.However,the research and development of cyber-physical power system is still at primary stage,and there are still many key issues need to be solved.As a system of deep integration of computing resources and physical resources,the basis of its safe and reliable operation is the panoramic data acquisition,transmission,storage and analysis.Aiming at the problem of low perceptive precision of physical equipment,great pressure in middle layer transmission and upper layer calculation and storage process brought by massive data,this paper focuses on the research and exploration in three parts: bottom layer perception,middle layer transmission and upper layer storage.The main research contents and contributions of this dissertation are as follows:(1)The condition detection of cyber-physical power system equipment has been a challenging problem for researchers over decades.We also point that the equipment signature can be uncertain and diminishing due to noise and other interference.To address this issue,this chapter proposed a novel method for condition detection of cyber-physical power system equipment based on building a feature knowledge database.Firstly,this chapter studies the electrical equipment and investigates the unique signatures of the healthy and faulty equipment.Secondly,this chapter proposed a novel scheme for condition detection of electrical equipment,and both fast fourier transform(FFT)and independent component analysis(ICA)analysis are leveraged to obtain the features.Finally,taking induction motor for example,the resulting FFT-ICA features contain rich information on the signatures of the healthy and faulty motors,which are analyzed to build a feature knowledge database for online fault detection.Furthermore,the proposed scheme can improve the efficiency of online fault detection at each inverter frequency and under each load level after finishing the offline training.(2)To solve the shortages of large data storage capacities and high complexity of compression in sampling of the data transmission process under the Nyquist sampling framework,this research chapter presents a novel data transmission method.Firstly,the sparsity of the multi-sourced heterogeneous data in cyberphysical power system is numerically proved first.This is followed by providing a proof of the matching satisfaction of the necessary conditions for compressed sensing.Secondly,this chapter proposed a data efficient transmission scheme based on compressed sensing theory.Finally,taking power quality signal for example,The results of the experiment shows that the proposed scheme effectively enhances the precision of harmonic and inter-harmonic detection with low computing complexity,and reduce the large amount of data storage and transmission.(3)Hadoop distributed file system(HDFS)has been widely used in datacenters for storing large amounts of data.However,these storage systems usually adopt the same replica and storage strategy to guarantee data availability,i.e.creating the same number of replicas for all data sets and randomly storing them across data nodes,which will lead to waste of cluster storage resources.To address this issue,this chapter presents a resource efficient utilization storage strategy.Firstly,this chapter develops a hypergraph-based storage model for cloud datacenters,which can precisely represent the many-to-many relationship among files,data blocks,data racks,and datanodes.Secondly,this chapter proposed a resource efficient utilization storage strategy for cloud datacenters based on a novel hypergraph coverage model.According to users' requirements of data availability in different applications,our proposed algorithm can selectively determine the corresponding minimum hyperedge coverage.We have also implemented improved implicit enumeration algorithm to solve it.Finally,experimental results show that the variable hypergraph coverage based strategy can not only realizes the efficient utilization of cluster resources,but also improve the storage performance in the datacenter.Based on the above three aspects,this paper realizes the combination of the bottom equipment precision perception,the middle layer data efficient transmission and the upper layer data optimization storage,and realizes the optimal utilization of multi-source heterogeneous data in the cyber-physical power system.Besides,the proposed method improves the optimal utilization of resources in power system and the intelligent level of power grid.