Research On Technologies Of Fault Warning For Automobile Assemble Conveying Equipment Based On Big Data

At present,intelligent equipment is more and more widely used in the manufacturing field,such as automobile assembly workshops.Intelligent equipment has many parts and complex internal structures.It is no longer feasible to rely on traditional manual experience to determine the operating status of equipment and provide early warning of failures.Aiming at the problem of complex structure,difficult operation and maintenance of automobile final assembly transportation equipment and the need for real-time early warning of faults,this paper studies the technology of early warning of automobile final assembly transportation equipment faults based on big data.Theoretical research is carried out from three aspects of fault early warning technology,big data technology and fault early warning system,and innovative theoretical methods are proposed.The use of big data technology has completed the development of automobile assembly and transportation equipment failure early warning system.First,a fault early warning method for automobile assembly and transportation equipment based on the growth neural gas clustering algorithm and improved LS-SVM regression model is proposed.After the production line data collected by the sensor is subjected to feature extraction and dimensionality reduction processing,it is passed into the growing neural gas clustering algorithm model to obtain the similarity trend.At the same time,it is passed into the improved LS-SVM regression model to obtain the residual value.The difference and similarity trends,the risk coefficient is derived,the equipment status is evaluated,and the equipment failure is pre-warned.Finally,an example of a hoist is used to verify the effectiveness and feasibility of this method.Then,from the aspects of optimizing the calculation process of the fault early warning algorithm and improving the fault tolerance performance of the system,the application of Flink in the fault early warning system is introduced.Then selected throughput and latency as performance indicators,designed two test scenarios of input and output and word count,conducted performance comparison tests,studied the performance difference between Flink and Storm,and gave the test results.Finally,the development of four early-warning system modules for the final assembly and transportation equipment of real-time data collection and transmission,real-time data calculation,real-time data storage,and visualization was completed.The feasibility and effectiveness of the method proposed in this paper were verified through practical application of enterprise engineering.