Research On Real-Time Fault Diagnosis Of Dynamometer Based On Edge Intelligence

Dynamometer is a kind of loading dynamometer,which is mainly used to detect whether the engine can work normally.However,the dynamometer will have some faults after several times of use.If the fault cannot be found in time,it will not only fail to accurately detect the running state of the tested object,but also may cause damage to the tested object,resulting in unnecessary losses.Therefore,it is of great significance to propose a timely and effective fault diagnosis method for the running state of the dynamometer.First of all,due to the uncertainty of the direction of the signal when the dynamometer vibrates,and the complex network environment in the production site,this paper uses a threeaxis acceleration sensor to collect the signal when the dynamometer is running,and more comprehensively reflects the vibration of the dynamometer.For a large amount of collected data,MQTT,a lightweight transmission protocol with high transmission efficiency and low bandwidth requirements,is used to enable stable transmission of signals in an uncertain environment.Secondly,based on the vibration data collected,this paper proposes an intelligent fault diagnosis neural network model based on the depth learning method,which can accurately locate the fault location and achieve accurate identification of the fault type.The model extracts the time characteristics of the input signal through the long short memory network LSTM,further extracts the signal characteristics using the convolutional neural network CNN,and at the same time adds the CBAM attention mechanism to screen the important information of the convolution results.Normalization and linear correction units are used between the layers to increase the network sparsity,accelerate the network training speed,and effectively improve the overall performance of the model.Finally,the trained intelligent fault diagnosis model is deployed locally,the fault diagnosis model is used to identify the fault of the collected vibration data,and the identification results are transmitted to the cloud center,which monitors the dynamometer operation data and status in real time.Once the dynamometer has a corresponding fault,the fault alarm information will be displayed,so that relevant personnel can see the equipment fault information in time and make corresponding countermeasures,So as to avoid a series of losses caused by equipment failure.