Research On Digital Prototype For Fault Diagnosis Of Injection Device Of Injection Molding Machine

As an important part of the injection molding machine,the injection device of the injection molding machine is also the high-frequency area where the failure of the injection molding machine occurs.After the failure,it needs to be diagnosed and repaired in time to avoid major losses and hazards.The traditional fault diagnosis methods of injection molding machines are mainly manual diagnosis and expert diagnosis.Due to their lack of intelligence,they have the shortcomings of "time-consuming,labor-intensive,labor-intensive" and low diagnostic accuracy,which can no longer meet the increasing degree of intelligence.High demand for fault diagnosis of injection molding machines.As an efficient simulation method in the manufacturing industry,digital prototyping provides a new idea for the fault diagnosis of the injection device of the injection molding machine.This technology is used to construct a virtual model of the physical entity to simulate the running state of the entity,and then the data obtained by the digital prototype is substituted into the model.Fault diagnosis of injection device of injection molding machine in neural network.This paper takes the injection device of BOY15 S injection molding machine as the research object,discusses the working process of the hydraulic system of the injection molding machine;analyzes and simplifies the basic structure of the digital prototype;then uses the co-simulation method of AMESim and 3Dmotion to build the digital prototype of the injection device of the injection molding machine;The mathematical model of the PNN neural network and the programming method of MATLAB are used to design and construct the fault diagnosis model of PNN and BP neural network suitable for the injection device of the injection molding machine;Determine the fault feature information used for fault diagnosis by the neural network,and design a table to organize and store the fault feature information in the form of a mat file;choose unable to inject glue,incomplete injection molding,nozzle plastic solidification,no feeding Four kinds of faults and normal states are used as fault diagnosis output results,and the fault feature information obtained from the operation of the digital prototype is substituted into the designed neural network model for training,and the test data is selected for testing.The experimental results show that using the data obtained from the digital prototype for fault diagnosis has a better fault diagnosis effect.When using small data samples for fault diagnosis,the probabilistic neural network has higher diagnostic accuracy,reaching more than 70%.When the samples are trained and diagnosed,the BP neural network has a high fault diagnosis rate,and its fault diagnosis rate is above 90%.This paper takes BOY15 S as the research object,builds its digital prototype,and uses the obtained data to diagnose the failure of the injection device of the injection molding machine.The research results will have a certain guiding effect on the fault diagnosis research of injection molding machines,and have a certain reference effect on the research of digital prototype novelty.