Research On Injection Mold Workshop Scheduling Based On Improved Hopifeld Neural Network

The rapid growth of our national economy has also brought a lot of opportunities andchallenges to the domestic mould manufacturing enterprise. In a new competitive marketenvironment, in order to survive and develop, those enterprises have to rely on introducingadvanced production equipment, reducing mould manufacturing cost and shortening molddelivery cycle to win market place, but, even so, their mould manufacturing cycles are farfrom the large enterprises. The main problem is the management of the mould manufacturingshop production scheduling is still the most original and outdated styles.The present scheduling mode of the mould enterprise workshop production managementis firstly described in this paper, and the existing problems have been analyzed. The wholeprocess flows of mould manufacturing have been detailed mapped through digging intoproduction line of injection mold manufacturing workshop that is the highest domesticautomation degree at present. Acorrding to the processing method, the whole mold is dividedinto four parts, and the processing technology of each part has been discussed separately.After summarizing problems in the present mould manufacturing workshop, and combiningwith the function of production scheduling, the necessarily of take production scheduling inthe shop have been found. For tracking the workshop’s actual mould making project, theprocessing machine and processing time to each corresponding process are recorded.Secondly, Hopfield neural network optimization algorithm has been chosen aftercomparing with other intelligent optimization algorithms. It can quickly find out an optimalsolution in a short time, but it is easy to fall into the local minimum solution in the searchprocess. Simulated annealing algorithm can jump out the local minimum and finally searchthe global optimal value through using Metropolis criterion, but it need long time to calculate.Heuristie search method can omit many useless searching paths, and shorten the searchingtime. A kind of hybrid hopfield neural network optimization algorithm that based onsimulated annealing algorithm is proposed after combining hopfield optimization algorithm,simulated annealing algorithm and heuristie search method. The new improved algorithm avoids the hopfield network optimization catching in local minimum defects, and takes intoaccount the time performance of the algorithm.Finally, the model of the injection mould manufacturing workshop scheduling problemhave been established based on the improved hopfield neural network, as well as theconstraint condition and optimal objective function.The optimized results before and after the improvements are compared, then the finalfeasible optimal solution is further compared with the actual workshop scheduling time. TheGantt chart showed the cycle of mould manufacturing would be greatly reduced by using theimproved algorithm take workshop production scheduling.