Research On Non-parameter Model Condition Monitoring Of Wind Power Unit

As a clean and renewable utilize way, the wind power has been developed worldwide in recent years, and the condition monitoring of the large equipment of the wind power units has become the important part. On the bases of the research achievement of condition monitoring on the wind power units in recent years, this article aims at the requirements on the control, operation and the simulation of the wind power unit, and research on the main issues of condition monitoring, the main content of work and the innovate achievement are as follows:1. It analysis the current situation of the condition monitoring on the wind power unit in details, and come up with two issues in the field of condition monitoring, one is regarding the selection of input variable and another one is regarding the dispose of residual.2. The key point of establishing accuracy model is determined by the measure parameter of input variable sets when the forecase models of the main measure parameters of each component is established. The article analysis the judge method and the relative relationship among the variable sets, establish the IPSO-BP condition monitoring model regarding the temperature of bearings of wind power units. It will selected the input variable sets of the IPSO-BP models according to the analysis results of the relative relationships, and it will refine the selection ways of non-parameter model input samples in contrast to the results of residual.3. In the statistical methods of residual regarding non-parameter forecase models of the wind power units, the suitable of the slipping windows can eliminate the impacts caused by measure errors and occasional reasons, for example the shut down of the units. The article analysis the mothed and principle of the duplex slipping windows in details, it indicates the suppression fuction on the abnormal residual by using the duplex slipping windows, then, confirms it by using the IPSO-BP forcase models.