Machine Learning Based Power System Fault Diagnosis And Transient Stability Assessment

In recent years,with the large-scale access of flexible loads such as non-work air conditioners,large industrial users,electric vehicle charging stations(pills),microgrid,user-side energy storage,and interruptible load,the scale of UHV AC-DC interconnection continues.The structure and operation mode of the power grid is becoming more complicated,the source-network-charge side uncertainty increases.The safe and stable operation of power system is a huge challenge.Power system fault diagnosis and transient stability assessment are important guarantees for grid security risk situational awareness and safe operation.However,power systems are often disturbed or even interrupted by many factors.Traditional grid protection methods may cause protection misoperation or refusal due to insufficient accuracy of protection settings.And the safety and stability analysis method can not meet the demand for fast online transient assessment.Therefore,the research on new power system fault diagnosis and transient stability assessment methods has attracted wide attention from scholars at home and abroad.In recent years,due to the continuous development of computer technology and artificial intelligence,machine learning methods based on data mining technology provide new ideas for power system fault diagnosis and transient stability assessment.The research work and contributions of thesis paper are mainly summarized as follows:(1)A rough set-BP neural network fault diagnosis method based on genetic algorithm is proposed.Firstly,the genetic algorithm is combined with the rough set to preprocess the data;the genetic algorithm is used to optimize the initial weight and threshold of the BP neural network.The simulation results show that the model can accurately and effectively diagnose faults,which is feasible and effective.(2)The PSD-BPA simulation software,which is developed by China Electric Power Research Institute and widely used by utilities in China,is used to simulate the transient stability of the power system.The original data under different operating conditions are collected.Based on the transient stability mechanism of the power system,the key factors affecting the transient stability of the system are analyzed.Set and combine the research results of existing literature to establish the original feature set.The original sample of the transient stability assessment is obtained by performing systematic simulation analysis.(3)A power system transient stability assessment method based on principal component analysis-support vector machine is proposed.Firstly,the principal component analysis is used to reduce the dimensionality of the data,and then the optimization method of the kernel function parameters in the support vector machine model is improved.Simulation results verify the effectiveness of the model.(4)A method based on deep neural network for power system transient stability assessment is proposed.Using the powerful feature extraction and transformation capabilities of the deep neural network,the model is optimized by adjusting the appropriate parameters,the original sample set is trained,and finally the transient stability evaluation is performed.The simulation results show that the model has higher evaluation accuracy and practicability than other classification models.