With the rise of smart grid in recent years,intelligent s ubstation appeared simultaneously.As the focus of the ‘power flow,information flow,business flow' three-flow collection,intelligent substation plays a very important role in building a strong smart grid.Meanwhile,some urgent problems are exposed,such as the fault detection problem of primary equipment,the interference warning problem of secondary equipment and the spectrum sensing problem caused by the sho rtage of spectrum resources.Research shows that by monitoring the electromagnetic spectrum of the substation,the above problems can be solved effectively.However,the existing electromagnetic spectrum monitoring system of substation has so many disadvantages,such as high complexity,large volume,narrowed frequency domain and high cost.It is difficult to meet the requirement of multi-point spectrum real-time monitoring in substation.To this end,this paper presents a substation electromagnetic spectrum monitoring system based on the theory of sparse signal processing.After systematic analysis of the theory of sparse signal processing,this paper makes a field measurement of three substations of different voltage levels in a city along the southeastern coast and obtains the actual data of the electromagnetic spectrum of the substations.After the statistical analysis,it came to a conclusion that the electromagnetic spectrum of the substations is sparse,block-sparse and location-correlated.This paper constructs a joint sparse model of the electromagnetic spectrum of the substation,that is,the spectrums of different locations in a substation not only contain the common part with same sparse support set,but also contain separate sparse parts.Furtherly,based on this model and the requirement of multi-point real-time monitoring in the substation,a monitoring model is proposed for the wireless sensor nodes in the substation to acquire the el ectromagnetic spectrum.Finally,a joint binary-tree decomposition block orthogonal matching pursuit algorithm is proposed to reconstruct the electromagnetic spectrum of the substation,and the superiority of the proposed algorithm is verified by Matlab simulation. |