Research On Reliability Evaluation Of Power System Based On Improved Load Curtailment And Sequential Chain-Sampling

The reliability evaluation of power system can be described as a kind of measure of ability to satisfy power consumers or enterprises with sufficient electrical power and energy with an acceptable quality standard.Reliability evaluation of power system can provide a mass of significant reference information to power system planning and operation;as a result,it becomes an important auxiliary tool for the analysis of power system.It has become a prospective subject to evaluating reliability of power system as more accurately and more efficiently as possible.In reliability evaluation of power system,sequential Monte Carlo simulation(SMC)for estimating the reliability of power system can effectively simulate the timing transfer process between operational system state and fault system state.SMC can not only achieve an unbiased estimation of reliability index of power system,but also effectivity obtain the Probability density distribution about relevant reliability index.It's no hard to see that SMC has a wider application prospect for estimating reliability of power system.However,the weakness of traditional SMC is that its calculation speed is so slow and convergence of index is so weak,which severely limiting the engineering application of this method.Therefore,this paper carries out some research about SMC,and puts forward a modified state sampling arithmetic and state analysis method to accelerate the reliability evaluation of power system.From the prospective of state analysis of power system: the paper divides the global optimization process of traditional method for system state analysis into a serious of dynamic optimization sub-process within dynamic correction region,combining the static equivalent technique with optimal linear programming theory,the paper puts forward a modified load curtailment model to analysis system state.These models ensure the consistence of correctional optimization before and after the calculation of global power flow and improve the efficiency of corrective analysis by reducing the calculation scale.At last,the Results of the RBTS,the IEEE-RTS79 and the IEEE-RTS96 systems demonstrate that the proposed model can achieve good accuracy as well as significant speed-up.From the prospective of state sampling of power system: this paper transforms traditional process of system's random simulation about the annual sequence samples into sampling process of chain simulation to characterize evolutional behavior of random fault of system states.Combining the technology of conditional kernel density estimation(CKDE),this paper provides a new sequential chain sampling algorithm.This method can avoid the repeated calculation of system state samples,and improve evaluation efficiency of system state sampling to some extent.At last,the RBTS,IEEE-RTS79 reliability test systems are taken as examples to verify the effectiveness and fastness of this algorithm.