A Data Driven Model for Predicting Tool Health Condition in High Speed Milling of Titanium Plates Using Real-Time SCADA

Carbide tools are easy to be damaged and quick-wear in the process of high speed machining (HSM) of titanium plates, as it is difficult to predict its working condition accurately. The machining quality requirement of thin wall parts is so high, and the health condition of the cutter is closely related to the quality of processing. Predicting the change of tool health condition is very important to a controllable processing quality. In this paper, a data model of machine and tool, communication framework and access strategy based on the OPC UA was developed to predict tool healthy condition. A prognostics and health management (PHM) technology was applied to deal with related machine data. A BP neural network model was built to reflect the relationship between machine condition and the parameters of the tools. The tool healthy condition parameters, dimensions errors, surface roughness and surface texture of the workpiece were inspected to revise the prediction model, which developed the accuracy of the model and make it to be more valuable. The model realized the monitoring and predicting of important information of tool healthy condition and made the intelligent judgment to machining quality. The technical level of intelligent monitoring of tool management and processing cycle was improved while the efficiency and quality of machining titanium were increased.