Feature Weighted Group Sparse Pattern Analysis Method And Its Application In Fault Diagnosis Of Hydroelectric Generating Sets

The motor act as the main equipment of hydroelectric generating, whose health is not only related with the security of hydro-power plant, but also with the fulfillment of providing safe and stable electricity. Many unfavorable factors are increasingly impacting the security and stability of power grid. For example: the complexities of the equipment, seasonal variations of the units and a variety of the abnormal vibrations etc. In order to make sure the motor work stably, it is of great significance to monitor the motor running and diagnose faults.The conventional ways of fault diagnosis, which still play the leading role in the hydropower units, are based on the experience and knowledge of professional and technical personnel. The disadvantage is evident. Some work has to be down to raise the degree of equipment automation and intelligence.In this paper, characteristics of feature weighted group sparse series analysis, which are used to fault diagnosis of hydroelectric generating sets. The main contributions of the work are as follow:(1) For the problems of the non-linear structure and lack of adequate train samples of noise samples, this paper proposed an efficient kernel discriminative common vectors dimensionality projection algorithm, which called efficient kernel discriminative common vectors algorithms. The main advantages of this algorithm included effectively solving the non-linear distribution of data, introducing the common vectors to solve the small sample and obtaining higher classification accuracy.(2) As the noise sample was easy to lose information, this paper proposed feature weighted group sparse pattern analysis algorithm, the sparse coefficients were produced by a new proposed feature weighted group sparse representation classification algorithm, FWGSC. which could robustly regress a given signal with regularized regression coefficients by introducing the feature weighted locality structure of the data. Experimental results on several datasets show that FWGSC not only enhanced the computational efficiency but also obtained higher classification accuracy.(3) In order to solve the problem of high complexity, this paper proposed feature weighted group sparse discriminative projection analysis algorithm, FWGSDP. The algorithm maximized the subtraction of inter-class reconstruction residual and intra-class reconstruction residual and thus enables data-in to achieve better separation. Finally, a sequentially iteratively re-restrained sparsecoding and eigen-decomposition strategy was developed to solve the FWGSDP model efficiently.(4) This paper proposed a design of fault diagnosis system for the hydroelectric generating sets, including designing the fault diagnosis algorithm of hydroelectric generating sets and a local control system and upper-supervision software. Besides, FWGSDP was applied to fault diagnosis for hydroelectric generating sets. Experiments demonstrated that FWGSDP was effective and feasible.