Research On Dynamic Bottleneck Analysis And Optimal Allocation Of Manufacturing Resources In Discrete Workshop

Currently,the domestic and foreign competitive environment in which our country’s manufacturing industry is becoming increasingly fierce and is facing an unprecedented predicament under the significant impact of COVID-19.Transformation and upgrading are essential for the manufacturing industry to "break the situation".Building an intelligent discrete workshop through industrial Internet of Things technology is the basis for transforming and upgrading the manufacturing industry.However,a discrete workshop comprises manufacturing resources with multiple types,different characteristics,and strong coupling.The relationship between each manufacturing resource node is complex,and the resource status will change dynamically,making production control more difficult.Furthermore,in discrete manufacturing processes,manufacturing resource bottlenecks can limit system throughput.However,these bottlenecks cannot be dynamically identified by the production system,which further leads to the inability to effectively allocate resources to the production line according to changes in production status,resulting in a waste of resources and failure to deliver orders on schedule.Moreover,the existing research is still insufficient in real-time state perception and integration,bottleneck dynamic tracking and quantification and resource dynamic allocation under disturbance.Therefore,this paper takes the resource bottleneck as an entry point to conduct further research on the production control methods of the manufacturing process.First,build an Io T sensing network at the bottom of the workshop to sense and store manufacturing resource status information in real-time.Then,considering the independent status characteristics of equipment and the coupling relationship between equipment,combined with the real-time status of resources in the manufacturing process,a dynamic identification method for resource bottlenecks is proposed.And build a bottleneck degree model to quantify its impact on system productivity.On this basis,combined with the resource bottleneck state to dynamically optimize the configuration of processing resources,a dynamic configuration driving mechanism for processing resources integrated with real-time status is proposed.The impact of status changes on the system is quantified through dynamic configuration thresholds,and appropriate configuration strategies are selected according to the degree of status changes.In this way,the delivery time and carbon emission under the disturbance can be effectively shortened;at the same time,according to the bottleneck interval where the manufacturing resources are located,the corresponding WIP quantity can be configured to reduce the production cost.The specific research contents are as follows:(1)A resource fine-grained analysis method is proposed to solve the problem of workshop resource status that cannot be accurately sensed and analyzed in real time.Under the generalized boundary of the discrete workshop manufacturing system,study the state space and space-time transition probability of the buffer zone and equipment,and form a real-time perception framework of resource status information with the main line of "object-perception-monitoring"(perception framework design,perception environment design and status key technology of information integration).Combined with the proposed theory and framework,a resource fine-grained state analysis method is proposed to realize migration perception,clustering and storage of fine-grained resource states.(2)To accurately identify the dynamic bottlenecks in the manufacturing process,a dynamic bottleneck identification method for discrete workshops driven by the finegrained state of resources is proposed.A dynamic bottleneck identification and decision-making mechanism driven by an event-data hybrid are proposed.The mechanism judges whether to re-identify the system bottleneck through the decision threshold.First,taking the effective buffer zone as the system boundary,the complex manufacturing system is decoupled into independent subsystems,and each subsystem is analyzed independently to improve the accuracy of identifying system bottlenecks.A comprehensive bottleneck degree calculation formula is proposed to quantify the impact of bottlenecks on the system.How restrictive the system is.Finally,the relationship between workshop entities and agents and the logical relationship between agents is studied,and the bottleneck identification effect is verified through simulation models.The results show that the proposed method can effectively identify dynamic multi-bottlenecks in manufacturing.(3)Aiming at the problem of optimal static allocation of manufacturing resources,a configuration framework driven by tasks,goals and resource constraints is proposed.Anion model is established,and an improvement to minimise completion time The improvement of the original imperial competition algorithm is reflected in the following two aspects: when the algorithm is initialized,a new cost normalization formula is proposed to solve the problem of the loss of some colonial countries during the initialization;a local search strategy based on mobile bottlenecks is introduced in the revolution link,retain the excellent genes in the empire through two bottleneck heuristic neighbourhood structures,and improve the optimization ability of the algorithm.Finally,through standard cases and workshop examples,it is verified that the proposed method is superior to the original algorithm and the comparison algorithm regarding search speed and solution quality.(4)To solve the problem of a resource allocation failure in a disturbed environment,a dynamic optimal allocation method of manufacturing resources based on the real-time bottleneck states is proposed.Establish a dynamic configuration threshold calculation model based on the delivery date and bottleneck state changes,and form a dynamic resource configuration driving mechanism based on the dynamic configuration threshold.An improved multi-objective hybrid evolutionary algorithm is proposed to solve the model,a multi-objective fitness function is constructed using the index weighting method,and an invasion-based population self-cleaning mechanism is proposed to avoid the multi-objective evolutionary algorithm from falling into "premature".At the same time,considering the impact of WIP resource allocation on the system,a precise allocation strategy of buffer WIP resources based on the bottleneck interval is proposed.Finally,through a workshop example,it is verified that the proposed method can effectively optimize resource allocation results,reduce production costs and increase system output.(5)Based on the above research,the resource entities with perceptual and interactive capabilities in the discrete workshop are used as the carrier to collect realtime data in the production process,build a platform for dynamic bottleneck analysis and optimization of manufacturing resources in the discrete manufacturing workshop,and verify the availability and effectiveness of each functional module in the discrete production plus workshop of engineering equipment manufacturing enterprise.