Energy-efficient path planning for a single-load automated guided vehicle in a manufacturing workshop

Highlights

• Energy consumption can be an independent optimization objective in AGV path planning.

• An energy-efficient path planning model is formulated for a single-load AGV.

• AGV energy consumption characteristics are analyzed by motion state and vehicle structure.

• A two-stage solution method and a PSO-based method are proposed and verified.

• Transport task execution order has an impact on AGV transport energy consumption.

Abstract

With the aggravation of the global greenhouse effect and environmental pollution, energy saving and emission reduction have already become the consensus of the manufacturing industry to enhance sustainability. A material handling system is an essential component of a modern manufacturing system, and its energy consumption (EC) is a non-negligible part when evaluating the total production EC. As typical transport equipment, automated guided vehicles (AGVs) have been widely applied in various types of manufacturing workshops. Correspondingly, AGV path planning is usually a multi-objective optimization problem, and closely related to the workshop logistics efficiency and the smoothness of the whole manufacturing process. However, the optimization objectives that current AGV path planning research mostly focuses on are transport distance, time, and cost, while EC or EC-related environmental impact indicators are seldom touched on. To address this, an investigation into the energy-saving oriented path planning is executed for a single-load AGV in a discrete manufacturing workshop environment. Based on the analysis of AGV EC characteristics from the perspective of motion state and vehicle structure, transport distance and EC are selected as two optimization objectives, and an energy-efficient AGV path planning (EAPP) model is formulated. Further, two solution methods, i.e., the two-stage solution method and the particle swarm optimization-based method, are put forward to solve the established model. Moreover, the experimental study verifies the effectiveness of the proposed model and its solution methods and indicates that transport task execution order has a significant impact on AGV transport EC.

Introduction

An automated guided vehicle (AGV) is a typical driverless material handling device, which has been widely applied in inside or outside environments, such as manufacturing workshops, automated storage and retrieval systems, and port terminals (Fazlollahtabar and Saidi-Mehrabad, 2013). In a modern manufacturing setting, flexibility is a significant characteristic. Since an AGV can execute material handling tasks automatically and accurately, and adjust its transport route in time with the change of production process, a material handling system composed of multiple AGVs can greatly improve the production flexibility and the competitiveness of manufacturing enterprises. Correspondingly, AGV path planning is an issue worthy of attention in the application of AGVs in manufacturing workshops.

Nowadays, green manufacturing has already become a development trend of the manufacturing industry. However, the focus of energy saving and emission reduction is mainly on machine tools, and the attention paid to material handling systems is not enough on the shop floor. AGVs are usually powered by batteries, and will continuously consume electrical energy during travelling. Due to the limited battery capacity, the battery of an AGV needs to be changed or charged when the power is insufficient. On this condition, the AGV will be temporarily unavailable and influence AGV dispatching, routing, and scheduling. Hence, the consideration of energy consumption (EC) in AGV path planning can not only save energy for AGVs but also reduce the disturbance to the material handling system. Moreover, it is beneficial to supporting energy-efficient manufacturing systems. AGVs are the physical equipment consuming energy in a manufacturing system (Vichare et al., 2009). Generally, the total production EC in a manufacturing workshop can be divided into direct EC and indirect EC. The former involves the EC of machine tools and auxiliary facilities like robots, conveyors, and AGVs, and the latter is related to the EC maintaining the environment in which the manufacturing process is executed (e.g., lighting, ventilation, dehumidification, and heating), and the embodied energy of materials like workpiece, cutting tool, and cutting fluid (Seow and Rahimifard, 2011, Wang et al., 2014). The EC of material handling equipment is an essential part of direct EC, and the consideration of transport EC can improve the EC prediction accuracy for the whole manufacturing process. In addition, it has been indicated by current research that scheduling is an effective measure to realize energy-efficient production, which is suitable for all types of workshops. Nevertheless, there are usually various paths and transport modes (e.g., batch transport and sequential transport) available in a manufacturing workshop. Consequently, the workpiece transfer time between various machines may vary greatly when choosing diverse paths, which may have an impact on the job waiting time of the relevant machines, and then affect the generation and implementation of energy-efficient scheduling schemes (Ji et al., 2018, Zhang et al., 2019).

So far, extensive achievements have been reported on the AGV path planning on the shop floor. However, the most concerning objectives are transport distance, time, and cost (Fazlollahtabar and Saidi-Mehrabad, 2015, Han, Wang, Liu, Zhao, & Hu, 2017, Umar et al., 2015), while EC or other environmental impact indicators (e.g., carbon emission) are rarely touched on. According to the number of AGVs involved, AGV path planning can be classified into two categories, namely single AGV path planning and multi-AGV path planning (Bae and Chung, 2018, Fazlollahtabar and Hassanli, 2018). The former focuses on searching for an optimal path for a single AGV from its current location to the destination while making a trade-off among several transport objectives, and satisfying obstacle avoidance requirements. Meanwhile, the latter concentrates on how to assign various transport tasks to multiple AGVs rationally and deal with the possible conflict and deadlock, thereby improving the overall efficiency of the AGV system (Yoo et al., 2005, Lyu et al., 2019). Generally, the planned path for each AGV, i.e., the output of single AGV path planning, is utilized as the input information of multi-AGV path planning. Therefore, considering the increasing attention to the EC of manufacturing processes and the potential benefits of introducing EC into the workshop logistics transport process as an optimization objective, we investigated energy-efficient AGV path planning. As single AGV path planning is the basis of multi-AGV path planning, a single-load AGV was selected as the research object to facilitate study. Correspondingly, based on reasonable research assumptions and the quantitative analysis of AGV transport EC, an energy-efficient AGV path planning (EAPP) model with two performance criteria, transport distance and EC, was proposed. Then, by solving this model and performing the experimental study, we hope to reveal the relationship between EC and other traditional path planning objectives.

To achieve this objective, a literature review on the green vehicle route problem and AGV path planning in Section 2 is followed by Section 3 analyzing AGV transport EC and formulating the EAPP model. Further, two solution methods for the EAPP model, i.e., the two-stage solution method and the particle swarm optimization (PSO)-based method, are put forward and illustrated in Section 4. Then, Section 5 describes the experiments conducted to verify the effectiveness of the EAPP model and its solution methods and explores the factors affecting the optimization objectives of EAPP. Finally, Section 6 presents conclusions and future research prospects.