Research On Fast Calculation Method For Static Safety Analysis Of Power System Considering Uncertainty

With the expansion of power system scale and the increase of power generation,it has become a very important problem to establish a reliable power system operation monitoring,analysis and control system to ensure the safe and economic operation of the power grid.The development of computer technology provides a very favorable condition for the operation and management of power systems,and many power systems have adopted computer systems with excellent on-line performance and advanced man-machine interfaces.It can not only be used to improve the security of system operation,but also to minimize the cost of power production.The large-scale access of renewable energy increases the uncertainty and risk of power system operation.The on-line security evaluation and analysis of the system,including static instability,frequency stability,transient stability,voltage stability,small interference stability,put forward higher requirements.If the traditional analysis method is used,the calculation is large and complex,which is strictly limited in practical application and is not suitable for practical on-line scanning analysis,so the machine learning method is proposed and applied to static security analysis of power system.1.in this article,a static security analysis and evaluation method based on decision tree and source load uncertainty is proposed to improve the calculation speed of static security analysis in power system.The problem of slow performance of static security analysis based on traditional power flow technology is solved.2.on the basis of the traditional static security analysis method,the decision tree is sampled and the method of adding Gao Si noise is used to simulate all kinds of uncertain factors caused by complex environment in reality.The training sample of decision tree is formed and the sample is generated by simulation analysis.3.the principal component analysis(PCA)technology is proposed to reduce the dimension of the sampling data based on the traditional analysis method,find the principal component in the data,and use the principal component data to characterize the original data to reduce the dimension of the data set.Simplify samples and improve algorithm efficiency.4.the traditional static security analysis method is combined with the decision tree to compress the sample analysis results into a set of security rules,and the decision tree is constructed according to the rules for online analysis.The results of tens of thousands of simulations are compressed into a set of security rules that are used to quickly and online divide uncertain network states into secure or unsafe states to cope with the effects of uncertain factors and emergencies.Then the traditional detailed static security analysis is carried out to reduce the influence of uncertain factors on the calculation speed of static security analysis.