Robot Systems for Rail Transit Applications

Hui Liu

• Publisher: Elsevier
• Year: 2020
• Print ISBN: 9780128229682
• eBook ISBN: 9780128229408
• Topic: Technology & Engineering
• Subtopic: Mechanical Engineering
• Index: Technology & Engineering

Robot Systems for Rail Transit Applications presents the latest advances in robotics and artificial intelligence for railway systems, giving foundational principles and running through special problems in robot systems for rail transit. State-of-the art research in robotics and railway systems is presented alongside a series of real-world examples. Eight chapters give definitions and characteristics of rail transit robot systems, describe assembly and collaborative robots in manufacturing, introduce automated guided vehicles and autonomous rail rapid transit, demonstrate inspection robots, cover trench robots, and explain unmanned aerial vehicles. This book offers an integrated and highly-practical way to approach robotics and artificial intelligence in rail-transit.- Introduces robot and artificial intelligence (AI) systems for rail transit applications- Presents research alongside step-by-step coverage of real-world cases- Gives the theoretical foundations underlying practical application- Offers solutions for high-speed railways from the latest work in robotics- Shows how robotics and AI systems afford new and efficient methods in rail transit

Chapter 1

Introduction

Abstract

The development of rail transit robots is an important way to improve the automation level of rail transit systems, which is an direction for the development of intelligent rail transit technology. The application of robot systems can improve the efficiency and economy of rail transit systems. This chapter first introduces robot systems in the manufacturing, dispatch, and maintenance stages of the rail transit system. Then, the chapter introduces the key problems of rail transit robots in positioning, path planning, human–robot interaction, and power management in detail. Finally, the chapter summarizes the scope of this book and introduces the content of each chapter.

Keywords

Dispatch; Human–robot interaction; Maintenance; Manufacturing; Navigation; Path planning; Positioning; Power management

1.1. Overview of rail transit robots

Railway transit is vital to the national economy. Research in Japan and China indicates that development of the railway can lead to economic growth in the railway field [1,2]. To stimulate economic development, governments are actively developing the railway transport industry. In such a large industrial system, the development of automation can improve efficiency and reduce costs, so the level of automation in railways should be increased. The increase in automation requirements in the rail transit system generates a pursuit for rail transit robot systems.

In worldwide, many countries have proposed similar strategic plans to encourage the development of robots in the railway transit system. Taking China as an example, “A Country With a Strong Transportation Network” and “Smart Railway” are two important plans that encourage the development of automation of the rail transit system. Driven by these plans, many Chinese rail equipment manufacturers and operators have carried out extensive research in robotics. Zhuzhou CRRC Times Electric Co., Ltd. applied an automatic generating line to produce high-speed train converters [3]. CRRC Qishuyan Institute Co., Ltd. developed an intelligent manufacturing workshop for gear transmission systems for high-speed trains. CRRC Zhuzhou Institute Co., Ltd. combined automatic guided vehicle (AGV) technology with urban transportation equipment and designed autonomous rail rapid transit (ART) [4].

A variety of robot systems are employed. The use of robot systems in the rail transit system can be divided into three aspects: manufacturing, dispatch, and maintenance. In these three parts of rail transit, different kinds of robots have completely different roles, as shown in Fig. 1.1.

1.1.1. Rail transit robots in manufacturing

“A Country With a Strong Transportation Network” points out that intelligent manufacturing is required for the rail transit system. In the modern production line, robots can greatly improve processing efficiency. Taking the production of a high-speed train gearbox as an example, in the process flow of a transmission gear, robot arms can result in the efficient transmission of gears between different machine tools; when welding the gearbox, the welding robots can improve machining efficiency; when assembling the gearbox, assembly robots can work with high-precision; when the train is assembled, AGV enables the gearbox to be transported quickly between workshops. Applying robots in manufacturing not only saves processing time and labor costs, it improves manufacturing quality.

Figure 1.1 Robots in the manufacturing, dispatch, and maintenance of rail transit.

1.1.1.1. Assembly robots

As for assembly robots, the primary mission is to achieve high-precision positioning of the workpiece. According to previous research, assembly costs account for 50% of total manufacturing costs [5]. Assembly robot systems can also be divided into rigid assembly and flexible assembly robots. Rigid assembly robots are customized processing systems for specific workpieces in the traditional industrial environment.

Rigid assembly robots have poor generalization. If the production line is replaced with processed parts, the equipment needs to be customized. Replacement of equipment will cause a great economic burden. Compared with rigid assembly robots, flexible assembly robots can design customized processing programs according to the workpiece. Flexible assembly robots are programmable, which can result in different assembly schemes for different workpieces. Flexible assembly robots are significant for a flexible assembly system. In current industrial development, flexible assembly robots are the focus of development [6]. In the following discussion, assembly robots refer to flexible assembly robots.

An assembly robot consists of four components: machinery components, sensors, controllers, and actuators. To bring about a complex workpiece track in the real assembly environment, assembly robots usually have more than four degrees of freedom (DOFs). Mainstream assembly robots can be divided into two types: selective compliant assembly robot arms (SCARAs) and six-DOF robots.

SCARAs have four DOFs, which are commonly used in electronic assembly, screw assembly, and so on [7]. SCARAs are specially designed for assembly applications by Yamanashi University. SCARAs contain two parallel joints, which can assemble a workpiece in a specified plane. Compared with six-DOF robots, advantages of SCARAs are a higher assembly speed and precision; disadvantages are limited workspace. Commonly used control strategies for SCARAs contain adaptive control, force control, robust control, and so forth [8]. In state-of-the-art research on robot control, intelligent algorithms are employed to improve control performance [9]. Dulger et al. applied a neural network to control the SCARA [10]. The neural network was optimized by particle swarm optimization to improve performance. Son et al. adopted an optimized inverse neural network for feedback control [11]. To deal with disturbances in running, the parameters of the inverse neural network are updated by a back propagation algorithm. Luan et al. used the radial basis function (RBF) neural network to achieve dynamic control of the SCARA [12].

Six-DOF robots can locate the workpiece at almost any point. Thus, six-DOF robots can handle the assembly task of complex three-dimensional (3D) workpieces. The dynamics of six-DOF robots are basic for operating the robots. Zhang et al. considered the friction of the robots and used a hybrid optimization method to model the dynamics of the six-DOF robot [13]. After optimization, dynamic accuracy increased significantly. Yang et al. proposed a simulator for the dynamics of the six-DOF robot [14]. Robots with a large degree of freedom have large feasibility. However, too much freedom is uneconomical. To handle the trade-off between economy and feasibility, the DOF can be optimized for specific tasks. Yang et al. proposed an optimization method to minimize the DOF [15]. This optimization method can reduce the DOF and improve the use of the DOF.

Assembly robots should cooperate with the ancillary equipment. The fixtures are vital equipment to ensure cooperation in performance. The fixtures can fix the relative position between the workpiece and the robot under load. If the precision of the location of the fixtures is low, no matter how accurate the positioning precision of the robot is, it cannot achieve high-precision assembly. Currently, the flexible fixture is a future development [16]. Lowth et al. proposed a unique fixture that can adjust the radial and angular adaptively [17]. Although auxiliary devices are applied for assembly robots, the results of assembly robots may still be unsuccessful. Avoiding unsuccessful assembly is particularly important in electric connector assembly, because the electric connector is not a rigid component. To detect the unsuccessful assembly of the electric connector assembly, Di et al. proposed a hybrid detection system with a force sensor and camera [18].

The fault diagnosis and prognosis system of assembly robots guarantees assembly accuracy. There are many studies about fault diagnosis and prognosis systems. Huang et al. designed a classifier for the wiring harness robot [19]. That study modeled the manufacturing process was and calculated the fault with a fuzzy model. Baydar et al. introduced a diagnosis model with error prediction [20]. The proposed model integrated the Monte Carlo simulation, genetic algorithm, and so forth. The functions of the fault diagnosis and prognosis system for assembly robots should contain the main aspects as given in Choo et al. [21]:

• (a) The health states of the assembly robots are monitored in real time. The monitored data are logged into the dataset. The health features are extracted from the health states of the assembly robots. The faults and remaining useful life can be calculated according to the features.
• (b) According to the fault diagnosis and prognosis results, the assembly tasks are reassigned to make sure the failed assembly robots are replaced by the fully functioning robots. The maintenance plans can be made to repair the failed robots.

1.1.1.2. Collaborative robots

When applying the resulting classical industrial robot systems, interaction between robots and humans is limited. There are three reasons for this phenomenon:

• (a) Traditional industrial robots do not consider moving humans. If a collision occurs along a certain trajectory, it may cause great damage to humans. Therefore, the working area of the industrial robot is mostly separated from the working area of the human.
• (b) The weight and volume of traditional industrial robots are large, and it is difficult for humans to operate robots.
• (c) Reprogramming of robots is difficult and requires special programming tools for tuning.

However, human–robot collaboration can combine human creativity with the efficiency of robots and can amplify the flexibility of robots and further improve work efficiency. Under this demand, collaborative robots are born. Compared with traditional robots, collaborative robots have three main advantages: safety, ease of operation, and ease of teaching. Some robot manufacturers have launched collaborative robotics products. The robot company Universal Robot launched the UR3 collaborative robot [22]. This robot is the first truly collaborative robot. The UR3 collaborative robot is based on a six-DOF robotic arm and is flexible enough to achieve complex motion trajectori...

Table of contents

• Cover image
• Title page
• Table of Contents
• Copyright
• List of figures and tables
• Preface
• Acknowledgement
• Nomenclature list
• Chapter 1. Introduction
• Chapter 2. Rail transit assembly robot systems
• Chapter 3. Rail transit collaborative robot systems
• Chapter 4. Automatic guided vehicles (AGVs) in the rail transit intelligent manufacturing environment
• Chapter 5. Autonomous Rail Rapid Transit (ART) systems
• Chapter 6. Rail transit inspection robot systems
• Chapter 7. Rail transit channel robot systems
• Chapter 8. Rail transit inspection unmanned aerial vehicle (UAV) systems
• Index