Research And Development Of Public Bicycle Fault Analysis System By Data-driven

As a solution to the "last kilometer" problem of public transportation,public bicycle has become an important part of urban public transportation due to its green and convenient features.However,due to frequent use and harsh use environment,vehicles are prone to failure,and the lack of professional sensor equipment leads to the difficulty in fault identification and tracking of public bicycles,that is "fault bike detection difficulty",which leads to the maintenance difficulties of public bicycles,increases the maintenance cost of public bicycles,and reduces the efficiency of vehicle management.At the same time,the untimely handling of malfunctioning vehicles will seriously affect the normal use of users,thus reducing the quality of service of public bicycles.Therefore,it is of great significance to analyze the massive historical data of public bicycles,find out the law of vehicle failure from the data,and establish a fault diagnosis model based on the law to apply to the fault diagnosis of public bicycles,for solving the problem of "fault bike detection difficulty" of public bicycles and for maintaining the fault vehicle reasonably.Firstly,in order to find out the rule of the breakdown of public bicycles,the influencing factors of the breakdown of public bicycles are analyzed by using the method of structural decomposition.Taking the public bicycle system of Hangzhou as the research object,this paper makes statistics on the historical failure information of public bicycle in Hangzhou.Combined with the analysis of influencing factors of public bicycle failure,the effective features of public bicycle failure are selected.According to the feature attributes,combined with the original data processing method,a sample data set of public bicycle fault analysis based on data is established,which provides the data basis for the subsequent public bicycle fault identification and fault diagnosis.Then,aiming at the problem of "fault bike detection difficulty",a method of support vector machine fault detection for public bicycle is proposed.In combination with the public bicycle fault information archive,the fault information data of public bicycle is collected.The k-means clustering method is used to screen the sample data set,and thenthe sample data and support vector machine optimized by particle swarm optimization are used to obtain the fault discrimination model.The experimental results show that the support vector machine fault discrimination model optimized by particle swarm optimization algorithm can effectively distinguish the fault bike and non-fault bike of public bicycles,and the recall rate reaches 92.86%.The high accuracy of fault detection indicates that this method has certain engineering practicability.Then,in order to further analyze the fault types,a fault diagnosis model based on stacking automatic encoder is established for vehicle maintenance.Combining with the fault information file,the vehicle information data is collected to establish the sample data set.Through unsupervised training layer by layer,the stacked automatic encoder network automatically proposes effective features from the data set.BP algorithm is used to fine-tune the whole network and improve the accuracy of fault classification.Finally,the fault type is diagnosed and output according to the principle of softmax classifier.The experimental results show that the stacking automatic encoder is used to diagnose the fault categories of public bicycles,and its average accuracy is 84%,which indicates that this method can be applied to the diagnosis of the fault categories of public bicycles.Finally,on the basis of previous data analysis and fault diagnosis method research,the intelligent fault diagnosis system of public bicycle is designed and developed,and the fault diagnosis method proposed in this paper is verified.The results show that it has the expected effect in fault detection and fault diagnosis,and can be applied to engineering practice to solve the problem of "fault bike detection difficulty".In this paper,an intelligent fault diagnosis system is established to apply artificial intelligence,communication and computer technologies to the fault management of public bicycles,and a systematic and perfect fault diagnosis theory and application technology system of public bicycles is established to meet the requirements of smart city informatization.In the future,fault tracing can be further developed on the basis of fault diagnosis,so as to provide theoretical basis for fault prevention.