Research And Implementation Of Jobs Intelligent Dispatching System For A Kind Of Hyrbid Assembly Lines

As one of labor intensive manufacturing industies, apparel manufacturing depends heavily on the production supervisors's knowledge and experience, even intuition for conducting jobs dispatching and production control. This leads to the non-optimality and unreliability of the production plan. Meanwhile, as the trend of the apparel production mode with the small order size and fast changeover of garment styles, the jobs dispatching based solely on manual efforts cannot satisfy the demands of fast apparel manufacturing. Aiming to the requirements of jobs dispatching in apparel manufacturing, this paper firstly formulates the problem of the jobs dispatching problem and then proposes the jobs intellginet dispatching model based on a modified discrete differential evolution algorithm. Finally, a kind of production jobs dispatching expert system is developed by integrating the discrete event simulation model with the jobs intelligent dispatching model to provide more scientific decision for the jobs dispatching of the agile apparel manufacturing. The outstanding contributions and innovations of the paper are as follows.1. Formulate the mathematical model of the jobs dispatching problem in the hybrid assembly lines and present the objective functions and the constraints.For the overall evalution of the multiple objectives, the utility function method is proposed to synthesize the utility values of all objective functions. The definition of the utility function is introducted and explained in detail2. Propose the discrete event model which serves as the dynamics of the hybrid assembly lines to solve the difficulties of acquiring the dynamic performance objectives of the hybrid assembly lines without mathematical formulations. In addition, the discrete event simulation model provides a friend way for production managers to adjust the dispatching strategy, which is a shotcut for checking the performance of manual dispatching jobs.3. Propose a modified discrete differential evoluation algorithm based Pareto optimality which solves the prematurity problem of the traditional differential evolatuion algorithm. The modified discrete differential evolation is applied into the jobs dispatching of apparel manufacturing. The results of the experiments demonstrate the correctness and effectiveness of our proposed approach. 4. Develope a kind of jobs intelligent dispatching expert system for the sewing assembly lines in apparel manufacturing. The program structure and the implementation of main modules are presented. The application results are analyzed and discussed.The proposed Pareto selection trategy discrete differential evolatuion algorithm and the jobs intelligent dispatching expert system will benefit domestic apparel manufacturing enterprises with improving production agile and efficiency by the scientific jobs dispatching strategy and enhance their competition edges facing the global environments.