| Ventilation is an important technical means to achieve tunnel safety,health and comfort design standards.The service status of ventilation facilities will directly affect the quality of tunnel environment.Studying and analyzing the service status and evolution trend of tunnel ventilation facilities can provide data support for the refined management and maintenance of tunnel ventilation facilities,ensure the reliability of ventilation facilities,and also help to improve the safety level of tunnel operation.In this thesis,the tunnel fan is taken as the research object,and the evaluation method of tunnel fan service state based on analytic hierarchy process and fuzzy mathematics theory is proposed.This method constructs the evaluation system by selecting the evaluation index of tunnel fan service state,and then combines the state of each evaluation index to obtain the service state of tunnel fan.In this thesis,a fault diagnosis model of key parts of tunnel fan based on convolutional neural network is proposed to solve the problem that it is difficult to obtain the state of some evaluation indexes.The diagnosis model can use the vibration data of tunnel fan to obtain the fault information of some key parts,so as to provide the basis for the overall state evaluation of fan.The main research contents of this thesis are as follows :(1)Based on the comprehensive consideration of the influencing factors of the tunnel fan,the evaluation index of the service state of the tunnel fan is selected based on the relevant specifications and standards,and the evaluation system of the service state of the tunnel fan is constructed.After determining the standard range of each evaluation index by specification,the weight,deterioration degree and membership degree calculation method of each evaluation index are reasonably selected.Finally,the fuzzy comprehensive evaluation model of the service state of the tunnel fan is established,and the comprehensive score of the service state of the tunnel fan is obtained by the membership degree,weight and the weighted processing results of the evaluation set.(2)When evaluating the service status of tunnel fan,it is difficult to obtain the health status of some key evaluation indexes by conventional methods.Therefore,this thesis constructs a fault diagnosis model of tunnel fan based on convolutional neural network.After the vibration signal of the tunnel fan is filtered and processed by continuous wavelet transform,it is input into the fault diagnosis model for diagnosis and identification.The diagnosis results are used to judge and locate the fault of the key parts of the tunnel fan.The diagnosis results can provide data support for the overall state evaluation of the tunnel fan.(3)In order to ensure the feasibility of the fault diagnosis model in the research of tunnel fan state evaluation,the experimental verification is carried out with the fault data of the tunnel fan motor bearing.After the experimental platform is built based on the Python algorithm module,the bearing vibration signal is dimensionally converted and input into the fault diagnosis model for training.After determining the optimal model parameters of the diagnosis model,the fault diagnosis results of the tunnel fan bearing are obtained.The experimental results show that the diagnostic accuracy of the tunnel fan fault diagnosis model reaches 99.3 %,and its accuracy and diagnostic speed are significantly better than support vector machine,BP neural network and other fault diagnosis models.(4)Based on the maintenance data of tunnel axial fan with two different motor power in a tunnel,combined with the evaluation method of tunnel fan service state proposed in this thesis,the evaluation of the service state of a tunnel fan is carried out.The subjective weight of the evaluation index is determined by AHP analytic hierarchy process combined with expert scoring,and the combined weight value of each evaluation index is determined by the objective weight value determined by the entropy weight Topsis method.Combined with the fault diagnosis model,the deterioration degree and membership degree of each evaluation index are obtained.Finally,the comprehensive score of each fan in the tunnel is obtained through the evaluation model.The evaluation results can accurately and effectively describe the current health status of the tunnel fan,indicating that the evaluation method of the service status of the tunnel fan proposed in this thesis can provide a reference for the operation and maintenance of the tunnel fan. |
PDF Full Text Request