Multi-Bottleneck Identification And Resources Optimization Reconfiguration Of Job Shop Production Network

As an advanced manufacturing system,modern job shop production process contains a large number of resource nodes and complex interaction between nodes,being a cyclic motion based on dependence relation in the form of disordered chaos on the whole.In order to achieve rapid response to production schedule and optimized dispatching of production task in job shop operation process,it is necessary to properly understand a series of states including operation state of equipment,circulation state of material,execution progress of task,dynamic change of bottleneck in shop manufacturing process,to make a study of inherent correlation,transmission characteristic,extraction and analysis method of the information generated from manufacture process in terms of system,so as to offer scientific support for optimum allocation of resource.The paper takes operation optimization of complex production process in production shop in manufacturing industry as specific subject and breakthrough for research to carry out network analysis and modeling for complex manufacturing process in job shop based on the novel idea on integration of system complexity with complex network,so as to explore influence mechanism of change in physical system structure and configuration on operation efficiency of manufacturing system;make scientific researches on dynamical characteristic of information transmission in complex manufacture process and network to reveal diffusion mechanism of dynamic information in manufacturing system;establish dynamical equation for information to obtain the basis to determine information circulation in manufacturing process;take node attribute,network propagation and network topology into comprehensive consideration to present real-time identification and prediction method for bottleneck;achieve dynamic optimum dispatching of overall manufacturing process based on bottleneck control point to take full advantage of production processes at upstream and downstream of bottleneck control point by advantage of heuristic algorithm and genetic algorithm;and apply research result to guidance of engineering practice and offer new idea on shop operation optimization in discrete manufacturing,the details are as follows:(1)A dynamic model for Job-Shop based on network evolution analysis method is presented according to randomness,complexity and dynamic characteristic of Job-Shop manufacturing system in terms of complex network.A dynamic network model for Job-Shop is set up based on multi-hierarchy production data such as order attribute,technological line,logistic path and resource allocation in terms of complex network;Mathematical description of network model for job shop is achieved by data structure based on parameters such as order attribute,node service ability,resource load and mapping relation etc.;dynamical equation for network structure based on degree distribution of network and network centricity as well as key attribute of manufacturing system based on manufacturing load,throughput and delay period are set up to evaluate manufacturing system network for Job-Shop.(2)An identification method of bottleneck in manufacturing based on network characteristic is presented in terms of novel view on integration of complexity of manufacturing system with complex network.Bottleneck identification algorithm based on network bottleneck efficiency matrix is presented by expansion of bottleneck connotation based on network model.The method has taken relation among node attribute,network propagation mechanism and network topology structure into comprehensive consideration,with degree of bottleneck represented by manufacture load of node and network bottleneck efficiency matrix between nodes used to overcome the deficiency of other algorithms,namely bottleneck node identification is only dependent on adjacent node.(3)A dynamic identification method of multi-bottleneck is built based on network characteristic to tackle the problem that it is difficult to identify multi-bottleneck in Job-Shop in disturbance environment and that bottleneck identification after bottleneck shifting is not comprehensive and efficient to a certain extent.A dynamical equation for Job-Shop network is set up to obtain basis to determine disturbance factor on the move.A bottleneck identification algorithm based on CML is set up by expansion of bottleneck connotation and taking propagation mechanism of disturbance in production network,influence mechanism of topological coupling between nodes and dynamical characteristic of node into comprehensive consideration,so as to achieve quantitative description and continuous prediction of bottleneck in Job-Shop.(4)Theory of constraint dominated multi-stage optimization method integrating advantage of genetic algorithm and heuristic algorithm is presented to solve problem in resource optimization and allocation in job shop.Firstly,diversity of Pareto set is ensured by genetic algorithm characterized by high strength solved and high robustness.Optimum resource allocation in first stage is carried out by preparation of coding scheme and decoding scheme for algorithm and crossover operator;secondly,secondary dispatching optimization for optimum resource allocation is achieved by making the best of bottleneck resource at upstream and downstream according to heuristic rule;finally,experimental verification and contrastive analysis are carried out to verify that the algorithm is advanced and effective.(5)A prototype system for job shop network modeling and optimum resource allocation is designed and developed based on the method and technology mentioned above,the system contains two major functions namely network modeling and production dispatching optimization.The research content and system mentioned in the paper have been applied to optimum resource allocation in certain job shop,having obtained favorable application effect,indicating that the optimum allocation method for network and resource in manufacturing process is more identical with the existing situation of job shop and that the research findings have a bright future in application.