Discussion on failures under various types of machine tools
There are n parts Ji ( i = 1, 2, n) that are to be machined on m machine tools M j ( j = l , 2, m). According to the process requirements, the processing of each part consists of several processes and passes through these devices in their respective order, which is known as the nm ordering problem. If the p-process is required at most in the part processing, each machine has Kj ( j = 1, 2, l ) different processing functions related to this processing. The above problem can simply describe the problem, where k= max{ Kj}( j= 1, 2, l ). The job plan optimization goal is to minimize the passage time of all parts, that is, the time from the start of the first part to the end of the last part is the shortest. At the same time, consider various constraints in the processing process, such as the process can not be exchanged, the process can not be interrupted. For the workpiece to be processed, all that is concerned is to complete all the machining processes in the specified process sequence, and there is no requirement as to which machine to perform these tasks. The final result of the production operation plan is to obtain the processing order of each process of each workpiece on each machine. The representation of the processing order is not unique. Because the genetic algorithm is used to optimize the search plan, according to the characteristics of the multi-function machine, the natural number coding is used to represent the processing order of the workpiece, and the solution of the problem, that is, the processing of the parts. The order corresponds to the arrangement state of the individual bit strings in the algorithm. The specific method is that in a nk production operation planning problem, a bit string of length n is used to indicate the processing order of the parts, and the bit string is encoded by the natural number table {1, 2, 3, 4, 5,}. The length of the bit string of the individual is nm, and there are m, 1, m, 2, n n in the bit string; the digital a of the jth occurrence in the bit string represents the jth processing of the ath part. Since the processing sequence of each process in each part and the processing time of each process are known, according to the scheduling rules of the process machine, each bit string represents a sorting scheme and the passage of all the parts corresponding to the sorting scheme. time. The initial state of each bit string is a random arrangement. Such a representation can naturally include many constraints in the encoding of the bit string. 2 Description of processing resources In the processing system of multi-function machine tools, the traditional method of distinguishing machining resources by machine type is not enough to effectively describe the common and unique machining functions between machine tools. It is necessary to study a new method to describe the various machining machine groups. Processing functions. At the same time, it must also be considered that in the process of parallel random search of genetic algorithm, the occurrence of the bit length of the individual or the number of individuals of the whole population is prevented from increasing, resulting in a large increase in the amount of calculation required for each iteration. Various factors that may lead to an unacceptable slow convergence rate, because the length of the bit string or the size of the population will directly affect the calculation speed and search efficiency. In this paper, the concept of the process machine is used to describe the necessary resources of the processing system. Each different process of all the workpieces to be processed is described as a single process machine, so that each different process corresponds to a function-specific process machine. . Each process machine may contain several actual machine tools, and the actual machine tools between the process machines are allowed to cross each other, that is to say, each process machine corresponds to a collection of actual machine tools that allow each other to cross each other. If all the workpieces need to complete the five processes of P1, P2, P3, P4 and P5 during the machining process, one of the machine tools M k has the functions of completing P1 and P5 at the same time, this paper thinks that the machining system of the workpiece is composed of five The machine tool consists of Pm1, Pm2, Pm3, Pm4, and Pm5. The actual machine tool Mk belongs to two process machines, Pm1 and Pm5. In the actual scheduling process, when the workpiece comes to the front of the process machine, the process machine automatically matches the workpiece with the original machine that is the earliest to be completed according to the completion time of the actual machine tool. If there are multiple machines with the same function, they will still be treated as different actual machine tools. When considering the process machine, they will be classified into the corresponding similar machine according to their functions. The scheduling process of the process machine is actually the decoding process of the individual bit strings in the iterative process of the genetic algorithm. In order to take a code Ji from the bit string (ie, the workpiece number, the number of occurrences k indicates the kth process), according to the set rules, select one of the earliest completed machine tools and the actual machine tool to which the process machine belongs. The kth process of Ji(i= 1, 2, n) is matched. When all the codes are matched, it means that all the workpieces have been processed, and then the passage time of all the workpieces can be obtained. In this scheduling process, the basis of optimization is based on the order in which the codes appear in an individual bit string given by the optimization process, that is, the processing order of the workpiece, which is more reasonable to realize the process machine, the actual machine tool, and the workpiece to be processed. match. 3 algorithm implementation If the number of individuals in the population is N, the number of iterations is K, the probability of crossover is Pc, and the probability of mutation is Pm. The initial data required for the calculation process is set, and N individuals are randomly selected as the initial population. Two individuals are randomly taken from the population as parental bit strings to cross and cross to obtain two child generations. From the two parental individuals and the two descendant individuals, two of the larger values ​​are selected to enter the new generation of populations, and this operation performs PcN/2 times. Then, PmN/2 are randomly taken out from the remaining individuals in the original population for mutation. The mutation operation is carried out by means of two gene exchanges within the same string. This method can ensure the number of various genes in the string. The unchanged, individual after mutation is also selected into a new generation of population. All individuals in the remaining original population are directly selected for the new generation. After completing these operations, the individual bit strings are decoded according to the scheduling rules of the process machine, and the workpiece passing time and the appropriate values ​​of the various strings in the new population are calculated, and then the maximum suitable value and the retained maximum maximum value in the new population are calculated. Comparison: If the retained value is less than the maximum value in the current population, the maximum suitable value in the current population and other corresponding data are replaced by the original retained data; otherwise, the original data is retained. If the number of genetic operations has reached a predetermined value, the iteration is stopped, and the optimized value corresponding to the maximum suitable value and other data of the bit string are output; otherwise, the next iteration is shown. 4 numerical experiments The Job Shop job planning model for multi-function machine tools based on genetic algorithm reflects the development trend of modern manufacturing systems and new features of job planning problems, and improves the dynamic response capability of the processing system. In order to verify the validity of the proposed method, this paper calculates several Job Shop job scheduling problems of different scales. The optimization result of the Job Shop job scheduling problem, in order to simplify the calculation process, assume that each machine has two functions, the number of machines is the same as the maximum number of steps in all parts, except for the two functions of the machine, the other raw data is J10 M The 10 standard questions are the same. Selecting different initial populations, the optimization results are generally within 741; 100,000 sorting schemes are randomly selected, and the average value is 1114. 3. Similar results are obtained for the calculation of other scale problems. The calculation results show that the genetic algorithm combined with the dynamic scheduling optimization method of the process machine is more effective for solving the Job Shop operation plan of the multi-function machine tool. The coded bit string method is used to represent the processing order of the parts, and the relationship between the processes is considered. It can make its optimization scheme more reasonable. (Finish) Cap Embroidery Machine,Embroidery Machine,Flat Embroidery Machine HEFENG COMPUTERIZED EMBROIDERY MACHINES LIMITED , http://www.hk-embroiderymachine.com