Internet Of Things-based Research And System About Electricity Management And Consumption Behavior For High-Speed Train

As the most important transportation foundation of the country,the railway industry has an exponential increase in the number of equipment and the entire energy consumption of the corresponding railway sector.Faced with the current situation of power consumption of a large number of equipment in the entire maintenance area,how to avoid the traditional manual recording of power consumption,how to effectively detect abnormal power consumption,and how to arrange reasonable power consumption strategies have become urgent problems to be solved.For the above problems,the main content of thesis is to realize automatic abnormal detection of electricity consumption,provide a best electricity consumption strategy and develop a power consumption management system,so as to improve the economic operation level of the entire railway sector and maximize energy saving and emission reduction.Aiming at the current energy consumption pain points of the railway sector,thesis conducted research on abnormal power consumption data detection and power consumption optimization strategies,and develops a power consumption management system for the railway sector.The main work is as follows:（1）For the electricity consumption of the railway sector,the problem of equipment failures is eliminated,and the artificial unreasonable electricity consumption arrangement often leads to the occurrence of excessive electricity consumption data.For a large amount of unlabeled data,this part explored the anomaly detection task from the perspective of unsupervised learning including one-class support vector machine,isolation forest and probability outlier detection.The abnormal data sets marked by relevant experts were used to evaluate the pros and cons of different algorithms.Finally,the isolated forest algorithm with the best effect on the existing data sets was used to detect abnormal data,and the accuracy of anomaly detection reached 0.97.（2）For the electrical equipment of the existing motor train maintenance workshop,this part will take the lighting equipment as an example to carry out research experiments on optimizing the power consumption strategy.Combined with the existing electrical equipment in the experimental workshop,this part is based on Genetic Algorithm and Particle Swarm Otimization to provide a reasonable lighting strategy for the lighting equipment in the experimental workshop.Different from the traditional purely quantitative power consumption recommendation,the optimization results proposed in this part provided a specific lighting scheme for each lighting device in space.The above method finally provided a practically available automated lighting strategy for the motor vehicle sector.Using the provided lighting strategy can save 56.03% of lighting power consumption and avoid redundant repeated power consumption.（3）On the basis of the above methods,a power consumption management system was developed for the experimental department,and anomaly detection algorithm and optimization algorithm were integrated into the whole system.Provide data collection,data time-sharing and divisional statistical query,anomaly detection,power consumption strategy recommendation and other functions for the experimental department.Through the development of this system,on the one hand,it solved the problem of manual meter reading statistics traditionally used by the railway department,on the other hand,it facilitated the railway department to detect abnormal data and conduct targeted maintenance,and finally provided intelligent information after the vehicle enters the queue.The lighting power consumption strategy scheme used the equipment control function provided by the system to achieve the purpose of energy saving and emission reduction to the greatest extent.