| With the comprehensive construction of smart grid, the information level of power grid is getting higher and higher, the production and operation of the power grid is increasingly dependent on the support of massive integrated power information, the wide-area massive information cause a great burden of data transfer and storage on power communication network, and even affect the safe and stable operation of power grid. Meanwhile, smart grid requires accurate collection and efficient transmission of panoramic information that descript the object, structure and status of smart grid, so as to achieve the integration and sharing of power flow, information flow and business flow.In order to meet the requirements of panoramic dynamic information for various advanced application systems of smart grid, the target scope, information type and information capacity of wide-area monitoring of smart grid has increased considerably compared with traditional power grid, the wide-area dynamic real-time monitoring information will face serious problems of transmission and storage in the power communication networks.Firstly, this paper proposed the light transmission mechanism of wide-area massive data based on the transmission principle of minimum feature information. The mode feature information is extracted from massive raw data using feature extraction and pattern recognition algorithms, then the vector mapping and parametric coding operations of mode feature information can be carried out based on improved CELP compression coding model. Heterogeneous information is integrated and light data is generated, in multi-protocol environment using information integration and information modeling technologies. The background information and feature information were distinguished and differentiated transmission control strategies were developed to achieve the light transmission and storage of wide-area massive data, which could meet the technical requirements of on-line monitoring of the panoramic information in smart grid.Then, this paper proposed a novel light treatment method of massive electric and physical data based on the ideas of mode feature extraction and CELP compression coding. The online feature extraction of electro-physical signal can be implemented using the online detection and pattern recognition method based on the event trigging mechanism and energy entropy criterion, taking different identification strategies of feature parameters for steady-state signal and transient disturbances. For steady-state signal, spectral analysis of pretreated initial electro-physical signal can be implemented using DFT algorithm with software synchronized sampling, to get the common feature parameters of all signal components. For transient disturbance signal, the mating pursuit of electro-physical signal in the related atomic library can be implemented using the fast atom decomposition algorithm with energy entropy criterion, to get the distinctive feature parameters of each signal component. The identification speed and accuracy of mode feature parameters of electro-physical signal can be enhanced by distinguishing steady-state signal and transient disturbance events. Furthermore, the related atomic library of electro-physical signals in smart grid was established, and the extraction process of feature information of electro-physical signal was introduced, the results of simulation examples verified the accuracy and effectiveness of the proposed method.According to the idea of object-oriented hierarchy modeling in IEC61850standard, considering the online monitoring requirement of panoramic dynamic information of smart grid, the paper established the data model, information model and communication service model of light data transmission system, from the perspective of application functions and communication services The generation mechanism of light protocol data model was analyzed, including the group coding of feature mode, the mapping of feature vector and the encapsulation of light protocol packets, etc. Meanwhile, the classified encoding storage, feature analysis and data recovery methods are induced. The results of dynamic tests verify the correctness of the information models of light data transmission system.Based on the analysis of the differentiated transmission characteristics of different types of information in smart substation, the network transmission control strategies of the integrated information in smart substation was established, from the degree of network transmission delay and network transmission quality. The network calculus theory based calculation method of delay upper bound of information transmission was introduced, furthermore, the information flow scheduling policy based on the strict priority queue and classification of service mechanism, and the control strategy of network traffic based on the leaky bucket shaping mechanism were proposed. The influence of electromagnetic interference and network transmission performance to application functions in the spacer layer and station layer of smart substation is analyzed, and then the corresponding suggestions and countermeasures were proposed.Lastly, based on OPNET network simulation platform, the intelligent substation communication network model is built, the simulation experiments of network transmission of panoramic dynamic power quality information in light data transmission system is carried out, and the network transmission performance is analyzed and evaluated. A physical prototype system of light data transmission system of smart substation is built, and the contrast dynamic transmission tests of massive electrical and physical data is carried out with traditional data transmission method. The results of network simulation experiments and dynamic transmission tests verify the feasibility, reliability and superiority of the proposed light transmission mechanism of wide-area massive data in smart grid. |
PDF Full Text Request