Study On Preventive Maintenance Of Intelligent Production Line Oriented To Personalized Customization

The intelligent production line realizes the deep fusion of information and physics through ubiquitous Io T and network collaboration.It is characterized by the highly interconnected manufacturing equipment,the deeply integrated manufacturing data and the dynamic reconfiguration of production line,which meets the requirements of mix-flow manufacturing paradigm featuring multi-variety,small batch and customized production.Under the personalized customized production mode,the intelligent production line puts forward higher requirements for equipment reliability,operation stability and production adaptability.The conventional passive operation and maintenance mode can no longer meet the complex operation and maintenance requirements of the intelligent production line.This paper focuses on the key technologies of the preventive maintenance of the intelligent production line.Based on the premise of ensuring the efficiency and the equipment utilization of the personalized customization production line,the early degradation evaluation of equipment is carried out.Through the reconfigurable preventive maintenance,the production interruption caused by the unexpected downtime of production line is avoided.The purpose of this paper is to realize the autonomous perception,state evaluation,self-organized running,and load balance.The contributions of this paper are as follows.(1)The system framework of the preventive maintenance based on deeply integrated equipment information for intelligent production line is explored.The information transmission based on OPC UA,Machine to Machine(M2M)communication,and software-defined industrial heterogeneous network are applied on the information transmission.In the deep fusion of multi-source heterogeneous sensor data,the edge-enabled data fusion method and the data fusion mechanism of edge-cloud cooperation are proposed.Furthermore,the framework contains the theory of equipment operation monitoring based on equipment electrocardiogram mechanism and health evaluation method based on deep learning,which helps feedback the urgent maintenance information of the equipment in real time.In order to ensure the stable operation for the personalized customized production line,the reconfigurable maintenance mechanism is proposed to realize the systematic operation and maintenance for the adaptive control of the production process.(2)Similar to the use of electrocardiogram(ECG)for monitoring heartbeat,the equipment ECG(EECG)mechanism reveals the degradation process based on fine-grained collection of data during the entire operating duration.This paper expounds the EECG mechanism,including the granular division of the duration of the production process,the matching strategy for process sequences,and several important working characteristics(e.g.,baseline,tolerance and hotspot).Based on the EECG mechanism,the optimization method for the cycle time of the production process and the online monitoring method of equipment performance degradation are proposed.The performance of the EECG was validated using a laboratory production line.The experimental results have shown that the mechanism of the intelligent ECG can well support the implementation of EECG.The automatic production line EECG(APL-EECG)system of the intelligent production line can monitor the operation status of the equipment in real time and provide scientific guidance for the maintenance of the equipment.(3)Based on the time series sensing data of equipment,the performance prediction based on deep learning is proposed.The popular machine learning framework,namely Tensor Flow,is introduced to build the deep learning model framework.The work status of a cylinder,an important part of a small trolley in the automobile assembly line,is evaluated by the DNN model using Keras,and the key implementation technologies are described as follows.Furthermore,the status evaluation strategy of cylinder is developed and the accuracy rate of prediction result reached industrial application standard.It is found that DNN can be used to analyze sensing data of multi-source heterogeneous equipment with weak correlation in the incomplete observation environment of the rich data sets.(4)Based on the formal semantic model of domain ontology,the reconfigurable maintenance method for preventive maintenance is constructed.First,the manufacturing resources and production process of the intelligent production line is systematically analyzed.The formal semantic model is constructed by using domain ontology,and the manufacturing resources of production line are abstracted and described by a semantic network structure.Next,the dynamic fusion of the information of physical resources of the production line is promoted by using data-driven semantic model,which provides the basic model for the status acquisition and self-organization reconfigurable of the production line.Furthermore,the self-organization and self-adaptive operation mechanism of intelligent production line based on multi-agent system is established.Finally,for the predictable state degradation and performance imbalance,the reconfigurable method of path dynamic planning and task switching is formulated to realize the dynamic reconfiguration of the mixed flow production line.(5)Being focus on the load imbalance in the process of self-organizing and self-adaptive operation of the equipment cluster of intelligent production line,this paper explores the collaborative optimization method of intelligent production line based on edge computing.and the energy-aware load balance and scheduling(ELBS)strategy is formulated.Specifically,an energy consumption model related to the workload is established on the fog node,and an optimization function aiming at the load balancing of manufacturing cluster is formulated.the improved particle swarm optimization(PSO)algorithm is used to obtain an optimal solution,and the priority for achieving tasks is built towards the manufacturing cluster.The multi-agent system is introduced to achieve the distributed scheduling of manufacturing cluster.The proposed ELBS method is verified by experiments with candy packing line.Considering the energy and workload,experimental results showed that the proposed method provides optimal operation and load balancing for the mixing work robots.Summary,based on the underlying information interaction,a bottom-up reconfigurable maintenance method from single equipment to cluster equipment is proposed,which realizes the key maintenance technologies including the autonomous perception,state monitoring,preventive maintenance and load balancing.This paper contributes a theoretical and technical foundation for the preventive maintenance of the personalized customized production line.