Research Of Enable Technologies And Applications In Agile Manufacturing

Economy globalization and global informatization are making the environment of competition, pattern of development, space of activity and efficiency of run in manufacturing change fully and deeply. More and more enterprisers and engineers are facing new situation and challenge because of the competition, division and reengineering trend in manufacturing. How to develop and manufacture products quickly to respond to the complex and uncertain market demands? How to utilize the exiting manufacturing resource most economically and effectively in dynamic manufacturing environment? How to make the manufacturing system run high-effectively, safely and stably during the manufacturing process with high speed, accuracy and integration? On the background of these, this paper studies on the theory and application of many enable technologies in agile manufacturing.Agile manufacturing in the environment of information economy is a systematic engineering. The author proposes the framework of agile manufacturing methodology and the enable technologies in agile manufacturing, by analyzing the concept and features of modern manufacturing and American strategic plan of agile manufacturing, especially by analyzing in-deeply the competition character of modern manufacturing in our country and the core competitive edge supporting the enterprise's agile manufacturing. The agile manufacturing methodology is comprised of three levels, which are enabling technology level, profession skill level and application level, where enable technology level is the basic and core. Three enable technologies that have outstanding effect in realizing agile manufacturing are presented for the economic objectives time, quality, cost, service and environment. They are flexible technology of workpiece installation for developing new product quickly, dynamic scheduling technology and technology of process monitoring and technique optimization.On the side of developing product quickly, in order to shorten the technique preparation period of process, quicken the introduction time of new products, create new process technique and tool, give product designer more extensive design space, utilize and recombine exiting resource maximally, reduce the investment and wasting of special equipment and tool set up, the author proposes and realizes flexibleworkpiece fixturing method with initiative location for agile manufacturing. The initiative location method based on the actual position of the regular locating box and its components proposed and realized can avoid the effect of fixture's manufacturing error on workpiece's precision fundamentally. Generally, the working accuracy can be improved up to 2 levels and more. In the locating box system, the modeling of geometric converted information and its information processing method proposed and realized can solve effectively the problem of error compensation during switching of workpiece's processes. It provides reliable and accurate processing information basic for the technology of initiative location and fixturing. The stuffing formulary and technique with good combination property proposed and realized can improve the operating condition and processing quality of small rigid jobs such as thin wall etc. The flexible installation system developed with independent intellectual property technique as the core has been put into actual application, which has been used to solve many precise manufacturing problems, such as the thin wall including wing panel of guided missile, artificial jawbone, sliding tobbom of automatic wefting etc. and the complex parts with many processes and no location reference, which can not be solved by traditional fixtures.On the side of production planning and scheduling, in order to optimize the allocation and utilization of production process, optimize production progress, the author presents and realizes an intelligent scheduling method and system responding to multi-objectives, multi-constraints and multi-disturbances, which provides effective solution method and supporting system for solving the obvious conflict among multi changes, short period, high quality and low cost in the environment of agile manufacturing. A scheduling model is proposed and built, which can reflect the practical demands in modern manufacturing system. The model can solve partially the scheduling problems considering batch, the arriving time of job blanks, insertion of urgent job, machine breakdown and optimization objective, while these problems can't be solved by many other scheduling models. A hierarchical optimization decision and scheduling model considering production planning and tracking is presented and built. By reducing the complexity of agile manufacturing system in handling problems, it can adapt better to the multivariate global market and production operations target. The job scheduling process model based on multi agents is proposed and realized. It becomes theory base and technique for supporting and realizing the concurrent, rapid and effective operation of hierarchical optimizationdecision and scheduling. The intelligent scheduling optimization strategy and the intelligent scheduling algorithm based on knowledge and biological intelligent computation given have perfect computing speed and optimization accuracy. The intelligent scheduling method and system have been verified by well-known scheduling case base and actual scheduling problems with multi-objectives, multi-constraints and multi-disturbances.On the side of process monitoring and process optimization, in order to identify and evaluate device status on-time, exactly and fully during the complex and multivariant manufacturing environment and process, realize dynamic, associated and distributed process monitoring and fault diagnosis, eliminate the effect of variant operation and time change character on monitoring and decision, identify, classify and decide when the prior sample of state and fault can't be got completely, the author presents and realizes the process monitoring and process optimization, which can be used in dynamic production environment and modern manufacturing equipments. Traditional model of quality monitoring must depend on enough monitoring sample, and can only analyze and control limited parameters, and can't coordinate multi-objective optimization. But the concept and model of quality efficiency monitoring given can avoid all these problems. Process monitoring of machining system and machining optimization model are built based on reinforcement learning. The model can be evolved and perfected by reinforcement learning based on data base and processing character. It not only can adapt to everchanging dynamic manufacturing process, but also can optimize machining process and find machine potency based on accumulated processing data and experience. Multi-sensors are used to monitor multi variants during dynamic machining process. Process statistical theory and computational intelligence technique are integrated to solve the problem of feature extraction of non-structural data and information fusion. The modified reversal selection algorithm given makes the preferred state monitoring population closer to state sample. So the state detector that is more sensitive to non-self can be determined to improve the validity of state monitoring to dynamic manufacturing environment.