An effective discrete artificial bee colony algorithm for multi-AGVs dispatching problem in a matrix manufacturing workshop
Introduction
Automatic Guided Vehicles (AGVs) are computer-controlled driverless vehicles used to deliver materials. Since AGVs were introduced in 1955, they have been successfully applied in many different applications such as warehouse, container terminals, transportation and manufacturing systems (Draganjac et al., 2016, Yang et al., 2018, Cavone et al., 2018, Zhang and Li, 2018). Especially in recent years, with the rapid development of smart manufacturing industry, AGVs are increasingly favored by the numerous manufacturing enterprises due to their prominent features of simple operation, rapid response and high efficiency (Yang, Zhong, Dessouky, & Postolache, 2018). In a matrix manufacturing workshop, a fleet of m homogeneous AGVs of capacity Q, which is stationed at the depot, is waiting for the control system’s instructions to deliver the production materials to n tasks. Each AGV starts from the depot and returns to the depot after delivery of production materials. For each task, the AGV must deliver the production materials on time. For each AGV, its load cannot exceed a given boundary Q. It is important for the manufacturing industry to solve the above AGV scheduling problem (AGVSP) because an efficient AGVs scheduling can increase the productivity and reduce the delivery cost (Bechtsis, Tsolakis, Vlachos, & Iakovou, 2017). In these cases, it can be inferred that the AGVSP is a practical variant of vehicle routing problem (VRP) with time and capacity constraints (Ciancio, Lagana, & Vocaturo, 2018). However, the AGVSP is harder to solve than the VRP, as it considers both assignment and sequencing of tasks and routing of AGVs.
As mentioned in the literature (Fazlollahtabar and Hassanli, 2018, Malopolski, 2018), we divide the AGVSP into two continuous stages: dispatching stage and routing stage, which are tightly integrated together. In the dispatching stage, the tasks are assigned to AGVs and then the tasks assigned to each AGV are sequenced. Next, an optimal path is found in the routing stage based on the sequence produced in the dispatching stage. In this paper, we focus on the dispatching stage, namely the dispatching problems of tasks (AGVDP). The AGVDP is a NP-hard problem because it tasks into account assignment of tasks compared with the VRP. In general, it almost impossible for a NP-hard problem to obtain a solution by using exact solution methods in a limited amount of computing time (Braekers, Ramaekers, & Van Nieuwenhuyse, 2016). As a result, heuristics and meta-heuristics are highly appealed for solving such a complex problem (Szeto et al., 2011, Nalepa and Blocho, 2016, Yu et al., 2017, Brandao, 2018, Wang and Li, 2018). Most of the studies for VRP consider the travel distance (or travel time) as an objective, as this criterion is important for evaluating the performance of the VRP. However, for the AGVDP, the travel distance (or travel time) is only one of the key factors affecting its performance. Other factors, such as travel speed, delivery time, number of AGVs, and load of AGV so on, cannot be easily ignored. In the AGVDP, travel speed and load of AGV have been defined as a constant, so we take the following three important factors as the key indicators to evaluate the AGVDP. The first is to minimize the total travel distance of AGVs, which can reduce the energy consumption of AGVs. The second is to minimize the total difference between delivery time and arrival time, which can ensure the on-time delivery of AGVs. The third is to reduce the total number of AGVs, which can lower the investment in AGVs purchasing. In the current literature, there are no research on the AGVDP with the transportation cost minimization. Our study aims to fill this research gap by proposing new algorithms and benchmark results for this problem.
The current strategies adopted by almost all the plants are based on “First Come First Served (FCFS)”, that is, tasks that first send requests first obtain the service of AGVs. The FCFS is a simple heuristic responding to the production behavior of manufacturing workshop. Although it can obtain an acceptable solution by using the FCFS, this solution has still a lot of room for improvement. Therefore, in this paper, an effective discrete artificial bee colony algorithm is proposed for the AGVDP with the transportation cost minimization, as well as a mixed integer linear programming model. Numerous instances collected from real-world factory are used as benchmarks to evaluate the performance of the proposed algorithm. To compare with the proposed algorithm, we choose the FCFS method, the Gurobi solver and the four well-performing algorithms from the literature on the problems closely related to the considered problem as the competing methods. Moreover, a full factorial Design of Experiments (DOEs) is carried out to determine the best parameter values for the proposed and competing algorithms.
The remainder of this paper is organized as follows. In Section 2, we review the literature on the closely related problems. Section 3 formulates the problem under consideration and establishes a mixed integer linear programming model. In Section 4, we introduce the basic artificial bee colony algorithm briefly, which is followed by a description of the proposed discrete artificial bee colony algorithm in Section 5. Section 6 reports the computational results and comparisons. Finally, the concluding remarks are provided in Section 7.
Section snippets
Literature review
As mentioned in the literature (Fazlollahtabar and Hassanli, 2018, Malopolski, 2018), the AGV scheduling problem is divided into the following two sub-problems: one is AGV dispatching problem (assigning tasks to AGVs and sequencing tasks of each AGV), and the other is the AGV routing problem (finding an optimal path with a given sequence and determining arrival and departure times of AGV). Many researchers have treated separately or simultaneously these two sub-problems, and adopted many
Problem description
In a typical matrix manufacturing workshop, the workstations are arranged in a matrix distribution (see in Fig. 1). Each workstation consists of a material buffer and multiple computer numerical control (CNC) machines. The material buffer stores production materials for the CNC machines. The CNC machines consume production materials continuously. The AGVs are in charge of delivering production materials to the workstation, which is responsible for the production of products. When the stock
Basic artificial bee colony algorithm
The artificial bee colony (ABC) algorithm is a swarm-based intelligence algorithm, which is presented by Karaboga and Akay (2009) for continuous function optimization problems. In recent years, the ABC algorithm has been applied to many practical applications, such as the blocking lot-streaming flow shop scheduling problems (Gong et al., 2018), multi-objective flexible job shop scheduling problems (Li et al., 2011), the reverse logistics location and routing problem (Guo & Zhang, 2017),
The proposed discrete artificial bee colony algorithm
The basic ABC algorithm cannot be directly applied to the discrete/combinatorial cases due to its initial design for continuous function optimization. In order to settle the AGVDP with the transportation cost minimization, a discrete variant of the ABC algorithm, called DABC, is presented in this section. In the DABC, we first detail the solution representation, presented heuristic, and then initial population, neighborhood operators, employed bee phase, onlooker bee phase, scout bee phase and
Computational and statistical experimentation
In this section, we proceed with a comprehensive computational and statistical experimentation to check the obtained algorithm. We first describe the test benchmark, test methods and the experimental setting. Then, we calibrate the parameters of all the metaheuristics (proposed and competing). Finally, we carry out a detailed and comprehensive computational comparison of the proposed algorithm against the competing methods.
Conclusion
The automatic guided vehicle scheduling problem (AGVSP) is a very active research area in various practical applications, but there is still no study for the AGVSP on matrix manufacturing workshop in the existing research except (Li et al., 2018). The AGVSP is divided into two sub-problems, namely dispatching problem and routing problem, which is treated separately or simultaneously in the literature. In this paper, we have studied the AGV dispatching problem (AGVDP) with the aim of minimizing
CRediT authorship contribution statement
Wen-Qiang Zou: Methodology, Conceptualization. Quan-Ke Pan: Methodology, Conceptualization, Writing - review & editing, Funding acquisition. Tao Meng: Methodology, Writing - review & editing. Liang Gao: Formal analysis, Writing - review & editing. Yu-Long Wang: Formal analysis, Writing - review & editing.
Declaration of Competing Interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
Acknowledgement
This research is partially supported by the National Science Foundation of China 61973203, 51575212, and the National Natural Science Foundation for Distinguished Young Scholars of China under Grant No. 51825502, and Shanghai Key Laboratory of Power station Automation Technology.
References (51)
- et al.
Sustainable supply chain management in the digitalisation era: The impact of Automated Guided Vehicles
Journal of Cleaner Production
(2017) - et al.
The vehicle routing problem: State of the art classification and review
Computers & Industrial Engineering
(2016) Iterated local search algorithm with ejection chains for the open vehicle routing problem with time windows
Computers & Industrial Engineering
(2018)- et al.
Multi-objective real-time dispatching for integrated delivery in a Fab using GA based simulation optimization
Journal of Manufacturing Systems
(2013) - et al.
Branch-price-and-cut for the mixed capacitated general routing problem with time windows
European Journal of Operational Research
(2018) - et al.
A new memetic algorithm for optimizing the partitioning problem of tandem AGV systems
International Journal of Production Economics
(2009) - et al.
Recent advances in hybrid evolutionary algorithms for multiobjective manufacturing scheduling
Computers & Industrial Engineering
(2017) - et al.
A novel hybrid multi-objective artificial bee colony algorithm for the blocking lot-streaming flow shop scheduling problems
Knowledge-Based Systems
(2018) - et al.
An exact and a simulated annealing algorithm for simultaneously determining flow path and the location of P/D stations in bidirectional path
Journal of Manufacturing Systems
(2013) - et al.
A comparative study of artificial bee colony algorithm
Applied Mathematics and Computation
(2009)
A hybrid discrete artificial bee colony algorithm for permutation flowshop scheduling problem
Applied Soft Computing
A sustainable and conflict-free operation of AGVs in a square topology
Computers & Industrial Engineering
A multiple colonies artificial bee colony algorithm for a capacitated vehicle routing problem and re-routing strategies under time-dependent traffic congestion
Computers & Industrial Engineering
A bilevel decomposition algorithm for simultaneous production scheduling and conflict-free routing for automated guided vehicles
Computers & Operations Research
An effective co-evolutionary artificial bee colony algorithm for steelmaking-continuous casting scheduling
European Journal of Operational Research
Effective heuristics and metaheuristics to minimize total flowtime for the distributed permutation flowshop problem
Expert Systems with Applications
A discrete artificial bee colony algorithm for the lot-streaming flow shop scheduling problem
Information Sciences
A genetic algorithm for hybrid flowshops with sequence dependent setup times and machine eligibility
European Journal of Operational Research
Greedy methods for the distributed permutation flowshop scheduling problem
Omega-International Journal of Management Science
An ant colony algorithm (ACA) for solving the new integrated model of job shop scheduling and conflict-free routing of AGVs
Computers & Industrial Engineering
An artificial bee colony algorithm for the capacitated vehicle routing problem
European Journal of Operational Research
An effective krill herd algorithm with migration operator in biogeography-based optimization
Applied Mathematical Modelling
An integrated scheduling method for AGV routing in automated container terminals
Computers & Industrial Engineering
Two-level particle swarm optimization for the multi-modal team orienteering problem with time windows
Applied Soft Computing
A hybrid genetic algorithm used in vehicle dispatching for JIT distribution in NC workshop
Ifac Papersonline
Cited by (66)
HMS-RRT: A novel hybrid multi-strategy rapidly-exploring random tree algorithm for multi-robot collaborative exploration in unknown environments
2024, Expert Systems with ApplicationsAn effective collaboration evolutionary algorithm for multi-robot task allocation and scheduling in a smart farm
2024, Knowledge-Based SystemsJoint scheduling of AGVs and parallel machines in an automated electrode foil production factory
2024, Expert Systems with ApplicationsAn effective multi-objective adaptive iterated greedy algorithm for semi-open multi-AGV scheduling problem with charging and maintenance
2023, Swarm and Evolutionary Computation