Research On Tool Wear Warning System Based On Edge Computing And Cloud Optimization

With the improvement of equipment intelligence and the rapid development of workshop awareness,the perception of a large amount of data and information will inevitably lead to higher requirements for data transmission,analysis and processing in the workshop.The data business processing mode combining the cloud and the edge is a solution with superior performance and strong applicability.This article takes workshop tool wear as the research object.Aiming at the current situation of frequent tool change in the laminated material drilling process,lack of tool wear monitoring methods,low production efficiency and high tool costs,this paper proposes a workshop tool wear early warning strategy combining edge computing and cloud optimization.The main work of the thesis includes:1.Using the workshop as the edge node,establish a tool wear early warning system under the cloud and edge framework.In this system,the data acquisition and transmission layer provides data during tool processing,the workshop edge layer realizes real-time early warning,and the cloud service layer optimizes the early warning model online.2.Set up for plant equipment and the edge node layer transport layer data collection,real-time data acquisition and processing of the tool,and the data is filtered,splicing,and finally transmitted to the computing node Drag edge.3.Drag edge node acquired tool monitoring data includes flow characteristics,so using a Storm in large data architecture compute engine,wavelet packet transform and time and frequency acoustic emission analysis of the signal,extracts characterizing the state of wear feature amount,BP neural network based on the predicted value of the tool wear.4.The establishment of mechanisms for distributed data storage and processing tool based HBase distributed database.Spark computing platform,after training tool wear prediction model,to achieve self-optimization model and optimization parameters dynamically.This article builds a tool wear early warning prototype system,realizing real-time wear early warning at the edge nodes of the car,with small system response delay,high concurrency performance,and small early warning error;distributed data storage and self-optimization in the cloud.