Research On Online Fault Diagnosis Of Motor Based On Embedded Artificial Intelligence Algorithm

As an important equipment in industry and life,the health of motor's running status is related to the normal production and life.Therefore,it is of great significance to diagnose the fault and monitor the running status of the motor in real time.Taking the motor as the research object,this paper proposes an online fault diagnosis method based on embedded artificial intelligence algorithm,which mainly uses signal processing and neural network methods to deal with and identify various fault types of the motor,and deploys it to an embedded platform for online detection.The data preprocessing,feature extraction,optimization of the network structure,fault response time and other issues of the motor were studied respectively.This paper first introduces common motor faults.The main steps of traditional motor fault diagnosis methods are data preprocessing,feature extraction,and fault diagnosis.Although the traditional method has a good effect on a single fault detection task;However,with the increasing number of motor fault types,it is often difficult to choose the feature extraction method for motor fault diagnosis with traditional methods.Therefore,this paper proposes a method for extracting effective fault information and using BP neural network for data fusion,and deployed it on an embedded platform to realize online detection;The data acquisition system is used to collect vibration signals and phase current signals of three motors synchronously,after preprocessing multiple signal characteristics are extracted and fused into an indicator that can effectively distinguish various motor states.The experimental results show that the designed method has a high accuracy rate for identifying multiple types of motor faults,and can control the motor when an emergency fault is detected;In the entire embedded system,signal acquisition,preprocessing,feature extraction and fusion and fault identification can be completed within 0.25 s,which better reflects the real-time nature of online detection.In order to better realize the on-line detection and reduce the fault response time,this paper introduces the choice of embedded system processor.The low power microcontroller which does not support the operating system is selected for high real-time requirement.For the large amount of computation,choose microcontroller that support the LINUX,the systemsupports more complex computing framework;This paper provides a solution for realizing on-site diagnosis of motor fault on small,flexible and low-cost devices by using artificial intelligence technology.