| CPC G05B 23/0283 (2013.01) [G05B 23/0235 (2013.01); G05B 2223/06 (2018.08)] | 4 Claims |
|
1. A method for monitoring failure of assembly tooling for a mass-individualized production line, wherein a system for monitoring failure of assembly tooling is used, the system comprises: a manufacturing execution system (MES), a supervisory control and data acquisition (SCADA) system, an assembly tooling failure prediction system, a controller network, and an assembly line;
wherein the assembly line comprises a plurality of stand-alone devices for production and processing;
the SCADA system is connected with the assembly line through the controller network, and is configured to acquire, through the controller network, input/output (I/O) information of the stand-alone devices and further integrate the I/O information;
the assembly tooling failure prediction system is connected with the SCADA system, is provided with an assembly tooling failure prediction model, and is configured to monitor failure of the assembly tooling for the stand-alone devices based on the information acquired by the SCADA system; and
the MES is connected with the assembly tooling failure prediction system and the SCADA system respectively, and is configured to: dynamically formulate a production plan according to real-time status information of the assembly line acquired by the SCADA system and failure prediction information of the assembly tooling acquired by the assembly tooling failure prediction system; and send, by an industrial network communication protocol, a production task to the stand-alone devices through the Controller network; and
the method comprises:
acquiring and integrating, by the SCADA system, the I/O information of the stand-alone devices during an operation process of the assembly line through the controller network;
performing, by the assembly tooling failure prediction system, failure prediction for the assembly tooling for the stand-alone devices through the assembly tooling failure prediction model based on the information acquired by the SCADA system; and
dynamically formulating, by the MES, a production plan according to real-time status information of the assembly line acquired by the SCADA system and failure prediction information of the assembly tooling acquired by the assembly tooling failure prediction system; and sending, by an industrial network communication protocol, a production task to the stand-alone devices through the controller network;
building and optimizing the assembly tooling failure prediction model as follows:
A1: calling multi-source production data related to predicted assembly tooling as a status data input, wherein the multi-source production data comprises real-time working condition data and historical working condition data; analyzing the status data input, and extracting a feature; and optimizing selection of performance degradation data of the assembly tooling, so as to realize tooling status monitoring;
A2: predicting a performance degradation trend and remaining service life of the assembly tooling according to a nonlinear mapping relationship of the performance degradation data of the assembly tooling, so as to complete tooling status analysis;
A3: independently completing, by a tooling diagnosis and inference engine, an early warning diagnosis for assembly tooling that is about to fail based on prediction results about the performance degradation trend and remaining service life of the assembly tooling, combined with on-site diagnosis data and historical data, and generating a solution;
A4: evaluating performance of a current model, comprising a deviation between the performance degradation trend and remaining service life of the assembly tooling predicted by the model and actual data, as well as accuracy of status evaluation and failure warning; and
A5: determining whether the performance of the current model reaches a preset performance threshold; if not, iteratively optimizing the current model according to a model evaluation result, and returning to step A1; and otherwise, outputting a currently optimal assembly tooling failure prediction model to complete model building and optimization.
|