Research On Optimization Method Of Virtual Manufacturing Cell Running Based On Multi-objective

To meet the unexpected change of consumer order and the dynamic adjustment of market environment, the traditional mass-production model must be changed. Multi-varieties and Small-batch production model has become the dominant mode of production manufacturing in information age and has also become one of important organization and management pattern in Industrie 4.0 Era. With the implementation of "Made in China 2025" strategy, Multi-varieties and Small-batch production mode has become the important content of intelligent manufacturing development action plan after the manufacturing base strengthened continuously and the depth integration of information and manufacturing technical characterized digitizing, networking and intelligent. Virtual manufacturing cell is one of new production model to achieve Multi-varieties and Small-batch production, and is also of technical support for digital, networked, intelligent manufacturing. This dissertation focuses on resource optimization, production scheduling and multi-attribute decision of virtual manufacturing cell system running and investigates systematically interrelated key issues and technologies. The content is as follows:1. The production mode based on personalized requirement has some significant characteristics. More specifically, there are often conflicted objectives in multi-objective optimization owing to the fact that personalized targets of customer are distinct and that there are different demands among companies or departments within an enterprise. On the other hand, the production process possesses dynamic characteristic because of both external and internal inevitable uncertain disturbances in ever-changing situation. Needlessly to say, the characteristics about multi-objective and dynamic increase the complexity of reality production. The dynamic reconfiguration method, which is aiming for achieving dynamic optimization combination of device resource, is introduced to generate virtual manufacturing cell. The manufacturing resource utilization has been improved and the production cost has been cut down through the conversion from multi-varieties and small-batch to mass-production mode.2. On the basis of analysis about cell formation, heuristic algorithm and Pareto dominance technique and so on, the modified free search algorithm based on Pareto-optimality is proposed to solve multi-objective resource optimization problem of virtual manufacturing cell. Furthermore, the convergence of the proposed algorithm is proved and the convergence speed and complexity are analyzed.3. The production scheduling is an important step in virtual manufacturing cell system running including both machine allocation and operation arrangement. The improved genetic algorithm based on Pareto-optimality is proposed to solve cell production scheduling problem. In the proposed algorithm, Pareto-optimality technology is applied, population maintenance strategy is improved and the operation operators are optimized. And the job shop benchmarks are employed to verify the effectiveness and feasibility of the proposed algorithm. To adjust to some uncertain disturbances, a rescheduling strategy is designed to study a class of uncertain processing time described by interval number. The rescheduling strategy is carried out by rolling horizon optimization technique and universal gray number conception and is examined via an example.4. To achieve overall optimization of virtual manufacturing cell system running, a unity decision-making model with double layers based on Pareto dominant solution set is proposed. Decision-making step by step and information feedback are combined.The decision maker’s preference formation and multiple attribute decision-making method are applied in the unity decision-making model.