Power Quality Analysis And Improvement Of Hybrid Power System In Distribution Network

With the grid connection of distributed generation,more power quality problems will be brought to the grid.The PQ disturbance signals generated by connecting different types of distributed power sources and distribution networks under different operating conditions are different.In this paper,firstly,the IEEE power distribution system based on photovoltaic and wind turbine is built to extract the PQ disturbance signal.Secondly,the the PQ disturbance in distribution network including distributed power is analyzed.It mainly includes the PQ disturbance detection,PQ disturbance classification and PQ disturbance control.In terms of the PQ disturbance detection: In this paper,LMD(local mean decomposition)algorithm is used to detect power quality disturbances in microgrid.By analyzing the single and compound disturbance signals in time domain and in frequency domain respectively,the algorithm is proved that it has higher accuracy.However,the requirements for data processing and calculation are increased and the operation time is longer in the proposed method.In order to overcome the defect,the novel detection algorithm based on VMD initialization S-transform is proposed in this paper.VMD can be used to set the number of iterations in order to avoid the disadvantage in the algorithm LMD.Through detecting 9 types of disturbance signals in distribution network,it is found that the proposed algorithm has more advantages in the PQ detection on the hybrid power system in distribution network,and the initial signal is provided for the subsequent PQ classification.In the aspect of the PQ disturbance classification,two algorithms based on PQ disturbance classification are studied.The advantages and disadvantages of the different classification algorithms are discussed as well as the scope of application.The initialization of FCM(fuzzy c-mean)clustering data is relatively lower,only part of the main features can be extracted to complete the more accurate classification.The amount of calculation is reduced.The PQ disturbance classification is based on the analysis of de-trend fluctuation.The algorithm has higher accuracy in the recognition and is easier to operate.Without using the traditional classifier,the PQ events can be easily interpreted by adopting two-dimensional or three-dimensional scatter diagram.In the aspect of the PQ disturbance management: a hybrid power control system based on shunt active power filter and dynamic voltage regulator is built.The photovoltaic and wind turbine are included in this system to simulate the PQ disturbance generated by new energy.An intelligent control algorithm combined with shunt active power converter is proposed,which is managed by artificial intelligence technology.Meanwhile MPPT(maximun power point tracing)is realized in photovoltaic and wind power systems.The dynamic performance of SAPF is optimized by using fuzzy logic,neural network and adaptive neural fuzzy inference system control algorithm.Therefore the PQ disturbance isgoverned.The harmonic distortion rate in the output side of the linear and nonlinear load is reduced to0.20% and 2.05%,which meets the PQ requirements in the distribution system.