Research On Power Quality Disturbance Identification Method And Reactive Power Compensation Strategy For Subway Power Supply System

The subway provides solutions for urban transportation p roblems,including high-speed,safe,comfortable,convenient,and low-cost.Becoming an important component of urban transportation and infrastructure for urban development in China.Operating conditions such as locomotive start stop,full load or no load can cause power quality issues such as harmonic generation and voltage fluctuations;As the departure interval of trains decreases,the increase in the number of trains operating simultaneously has an increasing impact on the system,causing harm to the equipment safety,power supply stability,and system reliability of the power system.Therefore,it is urgent to carry out research on power quality disturbance identification and compensation strategies to ensure the stable and reliable operation of the subway power supply system.However,the current power quality monitoring methods require feature extraction before recognition,which requires some human intervention.The processing process does not have a certain method guidance,and only relying on experience to construct feature samples will affect the accuracy of the subsequent classification process;There may be many problems such as negative impact on the classification results and low accuracy due to operator errors.According to the power quality classification and timing correlation,this paper designs a classification method of power quality disturbance signals based on long-term and short-term memory networks in the subway power supply environment.Simplified the recognition process of power disturbance signals,compared to traditional power disturbance signal classification methods,it has the advantages of simple structure,reducing the negative impact of operator errors on classification results,and being able to directly classify power signals with good classification results.It can better meet the requirements of high-quality power and real-time recognition in subway power supply systems.In response to the characteristics of rapid load changes in the subway power supply environment and the tendency of AC DC hybrid traction power supply systems to generate harmonic voltage and voltage fluctuations and other power quality disturbances,in order to improve power quality,achieve tracking compensation,and achieve higher economic benefits,this article identifies different types of disturbances and conducts reactive power compensation for them,analyzing their compensation effects.Firstly,the working principle and control strategy of static var generator(SVG)are introduced,and particle swarm optimization neural network algorithm is used to optimize SVG;Then,simulate and model the subway power supply system,mainly studying the disturbance impact of train startup on the grid side;Finally,the types of electrical disturbances were identified in the subway power supply environment,and SVG was used to compensate for the grid side.Compensation strategies for different types of disturbances were studied.