On Research Of Buffer Capacity Allocation Problems Of Line Manufacture System

With the development of global economy, the collaboration between enterprises becomes more and more frequent and the relationship grows more and more closely, which constitute a complex network. In the complex production relationship, the allocation of production line have a great influence for the performance of production. A hybrid algorithm has been proposed to slove the problems of buffer capacity distribution who has network structure with random event for line manufacture system. The proposed algorithm is capable of simulating conceivable stochastic states of syste m structure sunch as equipments by inherits the works of unreliable manufacturing sys tem and represents the random faults utilizing nalysis model constructed by hybrid pro duction decomposition technique on the basis of improved initial solution determinatio n technique. The works of this thesis are as follows.Firstly, in order to evaluate the performance of linear manufacturing system, the algorithm of decomposition and synthesis method have been mastered. A comparison between similar production line and non similar production line has been analyzed to summarize the features of the two methods by simulation expriments.Secondly,An improved genetic algorithm has been proposed. A operator which ca n maintain the diversity of population is exsited in the proposed algorothm for its effe ctiveness to restrain the survival rate of early-maturing model in evolutionary process.The improvement of the survival rate of dominant mode and convergence speed when the diversity is high has been testified by thoery. When the diversity is low, the survi val rate of dominant mode is decreased and the probability of overstep the local is incr eased.Thirdly, A genetic simulated annealing algorithm is designed based on the the adaptive genetic algorithm and simulated annealing algorithm for line manufacture system. The local optimal solution is selected utilizing simulated annealing algorithm by the mutation operator to achieve good local search and global search ability. The convergence of the algorithm is proved.The result shows that production efficiency is improved using the genetic-simulated annealing algorithm in linear hybrid manufacturing system.Finally,Using GUI of MATLAB and GUI programming, the Buffer capacity allocation technique for linear production are designed by software, achieving related functions by operating the front desk, a modular design meet the needs of all segments of users and with better visual resistance and scalability.