Simulation and Optimization of a Converging Conveyor System

Transcription

1 Simulation and Optimization of a Converging Conveyor System Karan Bimal Hindocha 1, Sharath Chandra H.S 2 1 Student & Department of Mechanical Engineering, NMAM Institute of Technology 2 Assistant Professor Department of Mechanical Engineering, NMAM Institute of Technology Nitte, India Abstract This paper describes a simulation based approach for reducing the production time on a converging type conveyor system. Basic design of a converging type conveyor system was modeled and was integrated into the Technomatic plant simulation software developed by Siemens and simulation run was carried out. Terms like blocking time and setup time are defined and the aim of this study is to reduce the blocking time and setup time by proper sequencing of the parts i.e., to optimize the production time using the equipments related to conveyor system and to heuristically determine the optimum buffer size required. The equipments are implemented to the basic design and the same process of simulation is carried out and the results are studied. The possible outcomes of this study are that by conducting multiple simulation runs, set up costs will be reduced, production time will be reduced which results in optimization of the production system. Keywords Discrete event simulation, dynamic job sequencing, converging conveyor system. I. INTRODUCTION Information technology and computerized automation have significantly improved flexibility in recent manufacturing. Today, most high-volume manufacturing systems may appear to be old shaped transfer lines but in fact they have become highly flexible, producing a large family of products such as electronics, automobiles, and other consumer goods. Purpose of striving for flexibility is to reduce the setup cost or time-to-respond to the growing diversity of customer demands. However, even the most flexible systems may still invite some setup cost in job changes. It is often advantageous to change the job sequence to further decrease the setup cost and time. The research on scheduling in the vehicle industry has mainly been focused on the car sequencing problem, which was first described by Parrello etal[1]. Enumerative approaches, such as the branch-and-bound algorithm of Carlierand Pinson [2], can only conquer small-scale problem instances. The most common heuristic methods devised in the early days include dispatching rules (Sculli, GreenandAppel, Kanet andhayya, Baker) [3], [4], [5], [6] and shifting bottleneck (Adams et al, Holtsclaw anduzsoy, Balas and Vazacopoulos, Liu andkozan) [7], [8], [9], [10]. Bianco, Dellolmo, Giordani, and Speranza [11] have minimized the makespan in a multimode multiprocessor shop scheduling problem. Each task can be undertaken by any methods in a set of predefined alternative modes, where each mode specifies a required set of assigned processors and a processing time. At any point of time, each processor can be used at most by only a single task. Akpan [12] has introduced a network technique that is a very useful tool for analysis of job shop sequencing problems. This technique offers a basis for finding an optimum solution. In this technique, the jobs priority on machines is determined by FCFS (First Come, First Serve) rule. Guinet [13] has proposed a procedure for solving jobshop scheduling problems. In this procedure, the objective is to minimize the maximum completion time of jobs. Lee, Piramuthu, and Tsai [14] have proposed to combine capabilities of genetic algorithms and machine learning techniques in order to develop a job shop scheduling system. Blazewicz, Domschke, and Pesch [15] have presented an overview of solution techniques for solving job shop problems. After a brief roundup of all the trends and techniques, they concentrated on branching strategies and approximation algorithms belonging to a class of opportunistic and local search scheduling. Sabuncuoglu and Bayiz [16] have studied reactive scheduling problems in a stochastic manufacturing environment. An active schedule has the property that no operation can be started earlier without delaying another job. Specifically they have tested several scheduling policies under machine breakdowns in a classical job shop system. In their researches, the performance of the system is measured for the mean tardiness and makespan criteria. Yu and Liang [17] have presented a hybrid approach combining neural networks and genetic algorithms to solve the expanded job shop ISSN: Page 109

2 scheduling problem. Satake, Morikawa, Takahashi, and Nakamura [18] have proposed a simulated annealing method based on the rescheduling activity of the human scheduler in order to void local search solution for solving job shop scheduling. Shiqing Yao,ZhibinJiang n, NaLi [20] have presented A branch and bound algorithm for minimizing total completion time on a single batch machine with incompatible job families and dynamic arrival. Flexible dispatching rules for minimizing tardiness in job shop scheduling were presented by Binchao Chen, Timothy I. Matis [21]. Rui Zhang, Pei-ChannChang, ChengWuc [22] presented A hybrid genetic algorithm for the job shop scheduling problem with practical considerations for manufacturing costs. Xinchang Haoa, Lin Linb, Mitsuo Genc, Katsuhisa Ohnoe [23] presented Effective Estimation of Distribution Algorithm for Stochastic Job Shop Scheduling Problem. Andrea Rossi [24] studied Flexible job shop scheduling with sequence-dependent setup and transportation times by ant colony with reinforced pheromone relationships. K. Li, J.Y.-T. Leunga, B. Y. Cheng [25] presented an agent-based intelligent algorithm for uniform machine scheduling to minimize total completion time. Wenchang Luo, Min Jib [26] presented Scheduling a variable maintenance and linear deteriorating jobs on a single machine with a goal to minimise the makespan and total completion time. II. PROBLEM DEFINITION There are certain terms or problems that a manufacturing unit will come across like blocking time, starving time, setup time etc. What are these terms? Blocking time is defined as the time that is calculated when a machine is undergoing processing of a part and on the conveyor, another part is waiting in line for it to be processed in the machine. Starving time is defined as the time that is calculated when a machine is undergoing processing of a part and another machine in line is waiting for that part to perform the required processing operation. Setup time is defined as the time the time a machine takes to set itself up every time a part enters it for the processing operation. Another question arises that why do we need certain equipments like a buffer or a pick and place robot. The production lines are subjected to disturbances arising from variations in processing times and failures of the workstations involved. This can cause the machines to be idle and can lower the throughput of the line. In order to mitigate the effect of this disturbance, in between machines, buffers are used. The throughput of a production line can be increased to a certain extent by allocating more buffer space between the machines. Robots are generally used to reduce the human effort and also reduce factors like human fatigue and more importantly human errors and boredom. III. METHODOLOGY: A. Designs In this study, firstly a basic converging type conveyor system is designed and is integrated into the plant simulation software using the tools available in the software. Simulation run is carried out and the results are studied. Followed by this, a buffer is added to the basic design. This is primarily done as mentioned earlier, to reduce the production time. Simulation run is then carried out and the results are studied again. In order to further reduce the production time, a pick and place robot is added to the previous design i.e. to the design with the buffer. In the similar way, simulation run is carried out and the results are studied. The parts are properly sequenced to reduce the setup time because if the parts enter in a random fashion in the machines, the machine has to change its setup every time a different part enters. If the parts are sequenced properly, then the machine has to change its setup according to the sequence defined and less number of times. B. Working In each design, five conveyors lines and five sources (for the parts) were used and in each line two machines were placed for the processing operation. Through each source two parts i.e. part 1, part 2 (similarly part 3, 4 for source 2 etc) enter the conveyor and goes through both the machines for the processing operation and all the ten parts are assembled in the assembly unit and the final product is sent out of the system through the drain. Sorter acts as a buffer storage space. C. Designs In first design, the source generates the parts that need to be processed. Lines 1, 2, 3 etc. are the conveyors that are responsible for mechanical mass handling from one station to the other. The sorters consequently sort the parts according to ascending order of part number. The results from the simulation run of the basic type of layout showed that a considerable amount of time was being wasted in setting up. This was happening because every time a type of part was processed and the next part happened to be of a different type, the setup had to be changed. This consumed time unnecessarily. Hence, installing a sorter would reduce the setup time by ensuring that similar parts are sent first followed by the next type of parts. In the third design, the working basically remains the same as the previous designs, with the added effect of the pick and place robots eliminating the effect of human fatigue on the system productivity and performance. The flowchart below shows the heuristic calculation of the optimum buffer size required. ISSN: Page 110

3 Fig 1- Design 1 of basic design Fig 2- Design 2: With sorter ISSN: Page 111

4 Fig 3- Design 3: With sorter and pick and place robot Fig 4- Flowchart to determine the optimum buffer size ISSN: Page 112

5 IV. RESULTS AND DISCUSSIONS Simulation run was carried out for all the three designs and the results shown below were studied. The bars in green depict working condition while in grey depict waiting and in yellow depict blocking and in brown depict setup time. Table1 Total simulation time of design Layout Design 1 Design 2 Design 3 Simulation time 1day:18hours:12minutes:05seconds 1day:13hours:50minutes:30seconds 1day:06hours:20minutes:11seconds Fig 7 Results obtained by simulation run of design 1 The heuristic determination of the optimal buffer size is shown below. Table 2- Size of buffer at each line Fig 5 Results obtained by simulation run of design 1 LINE SIZE (STARVING CONDITION) V. CONCLUSIONS The production time was observed to reduce with each modification applied to the basic converging conveyor design. The second design eliminated the blocking condition and reduced the production time by approximately 5 hours which is 11.9% reduction compared to the first design. The third design further reduced the production time to approximately 12 hours which is 28.57% reduction compared to the first design. Also, the optimal buffer size required was heuristically calculated. The setup time and production time was significantly reduced. Fig 6 Results obtained by simulation run of design 2 ISSN: Page 113

6 REFERENCES [1] Parrello, B.D., Kabat, W.C., Wos, L., Job-shop scheduling using automated reasoning: a case study of the car-sequencing problem. Journal of Automated Reasoning 2 (1), [2] Carlier, J., Pinson, E., An algorithm for solving the job shop problem. Management Science 35, [3] Sculli, D., Priority dispatching rules in job shops with assembly operations and random delays. Omega the International Journal of Management Science 8 (2), [4] Green, G.I., Appel, L.B., An empirical analysis of job shop dispatch rule selection. Journal of Operations Management 1 (4), [5] Kanet, J.J., Hayya, J.C., Priority dispatching with operation due dates in a job shop. Journal of Operations Management 2 (3), [6] Baker, K.R., Sequencing rules and due-date assignments in a job shop. Management Science 30 (9), [7] Adams, J., Balas, E., Zawack, D., The shifting bottleneck procedure for job shop scheduling. Management Science 34 (3), [8] Holtsclaw, H.H., Uzsoy, R., Machine criticality measures and sub problem solution procedures in shifting bottleneck methods: a computational study. Journal of the Operational Research Society 47 (5), [9] Balas, E., Vazacopoulos, A., Guided local search with shifting bottleneck for job shop scheduling. Management Science 44 (2), [10] Liu, S.Q., Kozan, E., A hybrid shifting bottleneck procedure algorithm for the parallel-machine job-shop scheduling problem. Journal of the Operational Research Society 63 (2), [11] Bianco, L., Dellolmo, P., Giordani, S., & Speranza, M. G. (1999). Minimizing makespan in a multimode multiprocessor shop scheduling problem. Naval Research Logistics, 46, [12] Akpan, E. O. P. (1996). Job shop sequencing problems via network scheduling technique. International Journal of Operations and Production Management, 16(3), [13] Guinet, A. (2000). Efficiency of reductions of job shop to flow shop problems. European Journal of Operational Research, 125(3), [14] Lee, C. Y., Piramuthu, S., & Tsai, Y. K. (1997). Job shop scheduling with a genetic algorithm and machine learning. International Journal of Production Research, 35(4), [15] Blazewicz, J., Domschke, W., & Pesch, E. (1996). The job shop scheduling problem: conventional and new solution techniques. European Journal of Operational Research, 93, [16] Sabuncuoglu, I., & Bayiz, M. (2000). Analysis of reactive scheduling problems in a job shop environment. European Journal of Operational Research, 126, [17] Yu, H., & Liang, W. (2001). Neural network and genetic algorithm-based hybrid approach to expanded job shop scheduling. Computers and Industrial Engineering, 39(3/4), [18] Satake, T., Morikawa, K., Takahashi, K., & Nakamura, N. (1999). Simulated annealing approach for minimizing the makespan of the general job-shop. International Journal of Production Economics, 60, [19] Jerry Banks., John S., Carson, Barrry L., David M. Nicol. (2007). Discreate Event System Simulation. Third edition pg-160. [20] Shiqing Yao,ZhibinJiang n, NaLi(2012). A branch and bound algorithm for minimizing total completion time on a single batch machine with incompatible job families and dynamic arrivals. Computers & Operations Research 39 pp [21] Binchao Chen, Timothy I. Matis(2013). A flexible dispatching rule for minimizing tardiness in job shop scheduling. Int. J. Production Economics 141 pp [22] Rui Zhang, Pei-ChannChang, ChengWuc(2013). A hybrid genetic algorithm for the job shop scheduling problem with practical considerations for manufacturing costs: Investigations motivated by vehicle production. Int. J. Production Economics 145 pp [23] Xinchang Haoa, Lin Linb, Mitsuo Genc, Katsuhisa Ohnoe(2013). Effective Estimation of Distribution Algorithm for Stochastic Job Shop Scheduling Problem. Procedia Computer Science 20 pp [24] Andrea Rossi (2014). Flexible job shop scheduling with sequence-dependent setup and transportation times by ant colony with reinforced pheromone relationships. Int. J.Production Economics 153 pp [25] K. Lia, J.Y.-T. Leunga B.Y. Chenga(2014). An agentbased intelligent algorithm for uniform machine scheduling to minimize total completion time. Applied Soft Computing 25 pp [26] WenchangLuoa, MinJib(2015). Scheduling a variable maintenance and linear deteriorating jobs on a single machine. Information Processing Letters 115 pp ISSN: Page 114