Dynamic Data-driven Modeling And Optimization Methods For Efficient Operation In Stamping Workshop

Stamping production has been widely used in the manufacturing process with a series of advantages,e.g.,high productivity,low processing cost,stable product dimensional accuracy,convenient operation,and easy to achieve mechanization and automation.Stamping workshops are the basic unit to execute stamping activities,however,the long production lines,complex processes,and unreasonable selection and use of equipment have made it energy-consuming.Meanwhile,due to its unique production characteristics,scheduling in stamping workshops is difficult,production schemes are often unreasonable;there is more room for improving the production efficiency.In the current context of resource and energy shortage,the implementation of energy saving and production efficiency improving in stamping workshops is of great significance to achieve sustainable development.Stamping production belongs to the category of discrete manufacturing,and its status is closely related to multiple factors,e.g.,materials,machines,people,environment,and orders.During the operation of the stamping workshop,dynamic events such as production fluctuations,failures,and order insertions often occur.It is difficult to ensure the efficient operation of the workshop if these dynamic events cannot be sensed and processed in time,resulting in key indicators in the workshop(e.g.,energy consumption and operation efficiency)being greatly affected.Therefore,this paper establishes a dynamic data-driven modelling and optimization method for efficient operation of in stamping workshop,which have considered the random and dynamic characteristics during the operation.A theoretical model of key indicators(energy consumption and operating efficiency)for efficient workshop operation and a discrete event model driven by dynamic data was established.Based on the models,a method was proposed for energy consumption and operation efficiency prediction in the workshop under fluctuations,and an optimization method for efficient operation in the workshop to improve operation efficiency and reduce energy consumption was studied.The key technology of establishing data service flow in the workshop was studied and the management software was developed,to support the modeling and optimization.The verification of the above methods and technologies was realized in a stamping workshop that produces body covering.The research results of this paper provide system-level analysis and energysaving,and efficiency-increasing solutions for efficient operation in stamping workshop,which is of important practical significance.The main research work of the paper is as follows:(1)A dynamic data-driven modeling method was proposed for efficient operation in stamping workshop.The components of the workshop were analyzed,and the dynamic factors existing in the operation of the workshop and their influence on the operation were clarified.Key indicators for efficient operation in the workshop were identified,and theoretical analysis models were built,i.e.,the energy consumption model and the makespan model,based on the manufacturing characteristics of stamping processes.On this basis,a prediction model of key indicators for efficient operation in the workshop was built based on discrete event simulation,aiming at the current deficiencies in the prediction and analysis of key indicators for efficient operation in stamping workshop.Furthermore,comprehensively considering the variations of data in the workshop,the dynamic data-driven modeling method was formed for efficient operation in stamping workshop,and an overall strategy for dynamic data-driven modeling was established.The proposed method and strategy can ensure the timeliness and accuracy of the prediction,analysis,and optimization processes in the workshop for the efficient operation.(2)A method was proposed for dynamically predicting energy consumption and operation efficiency in the workshop under fluctuations.A decision rule was built based on an “offline and online” mode,to integrate the fluctuating data in the discrete event model of the workshop in a timely and effective manner.Relational networks were constructed based on historical data and multi-factor sensitivity analysis to determine the threshold,which provides a theoretical basis for identifying “effective variations” of fluctuating data and reducing computational costs.Meanwhile,the latest probability distribution model of data that obeys the gamma distribution was built by combining Bayesian Inference and Markov Chain Monte Carlo method.The feasibility of the proposed method and the established mechanism was verified by using a simplified workshop as an example.(3)An optimization method was proposed for efficient operation in the workshop.Considering the production characteristics of stamping,a strategy based on event-driven rescheduling and complete rescheduling approaches was built.An efficient production scheduling model in stamping workshops was established.A two-stage approach based on ranks of embodied energy of parts during manufacturing(EEPM)that can quickly select more efficient production solutions was proposed,which enables the actual production to tend to be high efficiency and low energy consumption.(4)The data service flow of the workshop was established and the management software was developed.The type of data at each level in the workshop was determined.A data acquisition network was built to acquire real-time data in the workshop and sense the variations.A multi-parameter-based data classification rule was proposed to accurately classify each stage of the stamping process based on the perceived real-time data and to efficiently obtain the steady-state data of each stage.A service flow engine was determined and a multi-level data service flow in the workshop was built.The service flow can provide dynamic and real-time data for the prediction and analysis of key indicators,integration of fluctuating data,and efficient operation scheduling in the workshop.Meanwhile,a management software platform was built for supporting the production monitoring,and efficient operation management and control in stamping workshop.(5)Applications of the aforementioned methodologies and technologies were achieved in an enterprise.The modeling and optimization methods for efficient operation in the workshop were applied to a stamping workshop that produces body covering parts of forklifts.The production monitoring,energy consumption and operation efficiency prediction and analysis,and production scheduling in the workshop were achieved,respectively.Results show that the proposed methodologies and technologies have laid a solid foundation for efficient operation management and control in the workshop.