Robot Systems for Rail Transit Applications

Cover
Robot Systems for Rail Transit Applications
Copyright
List of figures and tables
Preface
Acknowledgement
Nomenclature list
1 - Introduction
	1.1 Overview of rail transit robots
		1.1.1 Rail transit robots in manufacturing
			1.1.1.1 Assembly robots
			1.1.1.2 Collaborative robots
			1.1.1.3 Automatic guided vehicles
			1.1.1.4 Manufacturing robots
			1.1.1.5 Loading–unloading robots
		1.1.2 Rail transit robots in dispatch
		1.1.3 Rail transit robots in maintenance
			1.1.3.1 Inspection robots
				1.1.3.1.1 Inspection in traction substation
				1.1.3.1.2 Inspection along the railway
			1.1.3.2 Channel robots
			1.1.3.3 Unmanned aerial vehicles for inspection
				1.1.3.3.1 Pantograph
				1.1.3.3.2 Power lines
				1.1.3.3.3 Foreign matter invasion detection
	1.2 Fundamental key problems of rail transit robot systems
		1.2.1 Navigation
			1.2.1.1 Development progress
			1.2.1.2 Methodologies
				1.2.1.2.1 Positioning
				1.2.1.2.2 Path planning
			1.2.1.3 Applications
				1.2.1.3.1 Carrying automatic guided vehicle robots
				1.2.1.3.2 Rail transit inspection robots
		1.2.2 Human–robot interaction
			1.2.2.1 Development progress
			1.2.2.2 Methodologies
				1.2.2.2.1 Human–robot interaction based on gestures
				1.2.2.2.2 Human–robot interaction based on human skeleton
				1.2.2.2.3 Human–robot interaction based on speech recognition
			1.2.2.3 Applications
				1.2.2.3.1 Intelligent rail transit equipment manufacturing
				1.2.2.3.2 Intelligent operation and maintenance
		1.2.3 Power management
			1.2.3.1 Power forecasting
			1.2.3.2 Autonomous recharging
	1.3 Scope of this book
	References
2 - Rail transit assembly robot systems
	2.1 Overview of assembly robots
		2.1.1 Definition of assembly robots
		2.1.2 Development progress of assembly robots
		2.1.3 Types of assembly robots
		2.1.4 Key technologies of assembly robots
			2.1.4.1 Precise positioning
			2.1.4.2 Detection and sensing
			2.1.4.3 Assembly robot controller
			2.1.4.4 Graphical simulation
			2.1.4.5 Compliance wrist
	2.2 Main components of rail transit assembly robots
		2.2.1 Machinery components
			2.2.1.1 Base component
			2.2.1.2 Rotating joint
			2.2.1.3 Arm connecting component
			2.2.1.4 Wrist joint
				2.2.1.4.1 Remote center compliance wrist joint
				2.2.1.4.2 Passive remote center compliance wrist joint
				2.2.1.4.3 Active remote center compliance wrist joint
			2.2.1.5 End effector
				2.2.1.5.1 Mechanical gripper
				2.2.1.5.2 Special tools
				2.2.1.5.3 Universal hand
		2.2.2 Sensors
			2.2.2.1 Position sensor
				2.2.2.1.1 Resistive position sensor
				2.2.2.1.2 Photoelectric position sensor
				2.2.2.1.3 Magnetic position sensor
				2.2.2.1.4 Capacitive position sensor
			2.2.2.2 Angle sensor
				2.2.2.2.1 Inclination sensor
				2.2.2.2.2 Magnetic angle sensor
				2.2.2.2.3 Capacitive angular displacement sensor
		2.2.3 Controllers
			2.2.3.1 Controller structure
			2.2.3.2 Controller performance requirements
				2.2.3.2.1 Reliability
				2.2.3.2.2 Accuracy
				2.2.3.2.3 Security
		2.2.4 Actuators
			2.2.4.1 Pneumatic drive
			2.2.4.2 Hydraulic drive
			2.2.4.3 Electric drive
				2.2.4.3.1 Higher precision and repeatability
				2.2.4.3.2 Higher energy efficiency
				2.2.4.3.3 Compact design
				2.2.4.3.4 High dynamic response
				2.2.4.3.5 Longer life
	2.3 Arm dynamics of rail transit assembly robots
		2.3.1 Arm postures
			2.3.1.1 Denavit–Hartenberg coordinate transformation
			2.3.1.2 Lagrange equation
			2.3.1.3 Kinetic energy of the connecting rod
			2.3.1.4 System dynamics calculation equation
		2.3.2 Forward dynamics
		2.3.3 Inverse dynamics
		2.3.4 Arm trajectory planning
			2.3.4.1 Joint space trajectory planning algorithm
				2.3.4.1.1 Cubic polynomial interpolation method
				2.3.4.1.2 Fifth-order polynomial interpolation method
				2.3.4.1.3 B-Spline curve interpolation method
			2.3.4.2 Cartesian space trajectory planning algorithm
				2.3.4.2.1 Spatial linear interpolation algorithm
				2.3.4.2.2 Spatial arc interpolation algorithm
			2.3.4.3 Time-optimal trajectory planning algorithm
				2.3.4.3.1 Origin of particle swarm optimization algorithm
				2.3.4.3.2 Construction of 3–5–3 spline interpolation function
				2.3.4.3.3 Solution process based on particle swarm optimization algorithm
			2.3.4.4 Energy-optimal trajectory planning algorithm
	2.4 Arm inverse dynamic application of rail transit assembly robots
	2.5 Conclusion and outlook
	References
3 - Rail transit collaborative robot systems
	3.1 Overview of collaborative robots
		3.1.1 Definition of collaborative robots
		3.1.2 Development progress
			3.1.2.1 Background on the development of collaborative robots
			3.1.2.2 Early collaborative robots
			3.1.2.3 Current collaborative robots
				3.1.2.3.1 Multirobot collaboration
				3.1.2.3.2 Human–robot collaboration
		3.1.3 Application fields
			3.1.3.1 Traditional production line
			3.1.3.2 Aviation manufacturing and maintenance
			3.1.3.3 Scientific research
			3.1.3.4 Entertainment
			3.1.3.5 Other fields
		3.1.4 Key technologies
			3.1.4.1 Standards and specifications for collaborative robots
			3.1.4.2 Types of collaborative operations
				3.1.4.2.1 Safety-rated monitored stop
				3.1.4.2.2 Hand guiding
				3.1.4.2.3 Speed and separation monitoring
				3.1.4.2.4 Power and force limiting
			3.1.4.3 Robot operating system
			3.1.4.4 Programming method
			3.1.4.5 Collision detection
	3.2 Main components of rail transit collaborative robots
		3.2.1 Sensors
			3.2.1.1 Motion sensors
			3.2.1.2 Optical encoder
			3.2.1.3 Magnetic encoder
				3.2.1.3.1 Inertial measurement unit (IMU)
			3.2.1.4 Force sensors
				3.2.1.4.1 Tactile sensor
				3.2.1.4.2 Force/torque sensor
				3.2.1.4.3 Joint torque sensor
			3.2.1.5 Range sensors
				3.2.1.5.1 Ultrasonic range finder
				3.2.1.5.2 Infrared range finder
				3.2.1.5.3 Laser range finder
			3.2.1.6 Vision sensors
				3.2.1.6.1 Camera
				3.2.1.6.2 Kinect sensor
		3.2.2 Controllers
			3.2.2.1 Control system
			3.2.2.2 Human–robot interaction
				3.2.2.2.1 Speech human–robot interaction
				3.2.2.2.2 Gesture human–robot interaction
				3.2.2.2.3 Display equipment
				3.2.2.2.4 Remote control equipment
		3.2.3 Actuators
			3.2.3.1 Driving devices
				3.2.3.1.1 Servo motor
				3.2.3.1.2 Direct drive motor
				3.2.3.1.3 Harmonic gear
			3.2.3.2 End effector
	3.3 Visual perceptions of rail transit collaborative robots
		3.3.1 Feature extraction algorithms
			3.3.1.1 Image features and feature extraction
			3.3.1.2 Basic concept of image
				3.3.1.2.1 Images and graphics
				3.3.1.2.2 Digital image
				3.3.1.2.3 Pixel
				3.3.1.2.4 Color space and color images
				3.3.1.2.5 Binary image
				3.3.1.2.6 Grayscale and gray image
				3.3.1.2.7 Image graying
				3.3.1.2.8 Gray gradient
				3.3.1.2.9 Gray histogram
			3.3.1.3 Algorithms for feature extraction
				3.3.1.3.1 Histogram of oriented gradient
				3.3.1.3.2 Scale invariant feature transform
				3.3.1.3.3 Local binary pattern
		3.3.2 Target detection algorithms
			3.3.2.1 Basic task of machine vision
			3.3.2.2 Basic process of target detection
			3.3.2.3 Methods and algorithms of target detection
		3.3.3 Target tracking algorithms
			3.3.3.1 Differences between target tracking and target detection
			3.3.3.2 Basic process of target tracking
			3.3.3.3 Method and algorithm of target tracking
		3.3.4 Conclusion and outlook
	References
4 - Automatic guided vehicles (AGVs) in the rail transit intelligent manufacturing environment
	4.1 Overview of automatic guided vehicles
		4.1.1 Definition of automatic guided vehicles
			4.1.1.1 Development progress of automatic guided vehicles
			4.1.1.2 Types of automatic guided vehicle
			4.1.1.3 Loaded automatic guided vehicles
			4.1.1.4 Traction automatic guided vehicles
			4.1.1.5 Forklift automatic guided vehicles
			4.1.1.6 Pallet automatic guided vehicles
			4.1.1.7 Special automatic guided vehicles
	4.2 Main components of automatic guided vehicles
		4.2.1 Chassis
		4.2.2 Power devices
		4.2.3 Control devices
		4.2.4 Safety devices
	4.3 Key technologies in automatic guided vehicles
		4.3.1 Navigation
			4.3.1.1 Electromagnetic navigation
			4.3.1.2 Optical navigation
			4.3.1.3 Laser navigation
			4.3.1.4 Visual navigation
				4.3.1.4.1 Visual image pretreatment
				4.3.1.4.2 Visual image feature extraction
				4.3.1.4.3 Visual positioning method
			4.3.1.5 Simultaneous localization and mapping
				4.3.1.5.1 Visual simultaneous localization and mapping
				4.3.1.5.2 Light detection and ranging simultaneous localization and mapping
		4.3.2 Path planning
			4.3.2.1 Introduction
				4.3.2.1.1 Time-optimal path planning
				4.3.2.1.2 Minimum energy consumption
				4.3.2.1.3 Shortest path
			4.3.2.2 Global path planning algorithms
				4.3.2.2.1 Graph search algorithm
				4.3.2.2.2 Random sampling algorithm
				4.3.2.2.3 Intelligent optimization algorithm
			4.3.2.3 Local path planning algorithms
				4.3.2.3.1 Artificial potential field method
				4.3.2.3.2 Fuzzy algorithm
				4.3.2.3.3 Neural network method
				4.3.2.3.4 Reinforcement learning method
				4.3.2.3.5 Dynamic window approach
			4.3.2.4 Human–robot interaction methods
			4.3.2.5 Gesture recognition control
			4.3.2.6 Motion recognition control
			4.3.2.7 Speech recognition control
		4.3.3 Task assignment
	4.4 Automatic guided vehicle path planning application in the rail transit intelligent manufacturing environment
		4.4.1 Global static automatic guided vehicle path planning
		4.4.2 Local dynamic automatic guided vehicle obstacle avoidance
		4.4.3 Hybrid path planning application
	4.5 Conclusion and outlook
	References
5 - Autonomous Rail Rapid Transit (ART) systems
	5.1 Overview of ART
		5.1.1 Development progress of ART
		5.1.2 Advantages and characteristics of ART
	5.2 Main components of trams and ART
		5.2.1 Structures of trams
			5.2.1.1 Train body device
				5.2.1.1.1 Train driver's cab
				5.2.1.1.2 Train passenger room
				5.2.1.1.3 Train interior decoration
				5.2.1.1.4 Train roof equipment layout
				5.2.1.1.5 Train door
				5.2.1.1.6 Train coupler
				5.2.1.1.7 Train articulated device
		5.2.2 Train bogie device
		5.2.3 Traction device
			5.2.3.1 Braking device
			5.2.3.2 Electronic control device
			5.2.3.3 Auxiliary power device
			5.2.3.4 Air conditioning and ventilation device
			5.2.3.5 Vehicle multimedia terminal device
			5.2.3.6 Vehicle intelligent terminal device
			5.2.3.7 Sensors structures of ART
			5.2.3.8 Proprietary road-right for ART
				5.2.3.8.1 Camera
				5.2.3.8.2 Light detection and ranging
				5.2.3.8.3 Millimeter-wave radar
				5.2.3.8.4 Ultrasonic radar
		5.2.4 Key technologies in ART
			5.2.4.1 Road traffic interaction
			5.2.4.2 Navigation
			5.2.4.3 Communication
			5.2.4.4 Serial communication
			5.2.4.5 TCP communication
			5.2.4.6 UDP communication
			5.2.4.7 Scheduling management
		5.2.5 Pedestrian detection algorithms for ART
		5.2.6 Traditional pedestrian detection
		5.2.7 Smart pedestrian detection based on deep learning
			5.2.7.1 Convolutional neural networks
				5.2.7.1.1 Convolution layer
				5.2.7.1.2 Pooling layer
				5.2.7.1.3 Full connection layer
			5.2.7.2 Feature extraction of deep learning
			5.2.7.3 Detection algorithm based on region proposal
			5.2.7.4 End-to-end detection algorithm based on deep learning
			5.2.7.5 ART pedestrian detection application
			5.2.7.6 Pedestrian keyframe extraction HOG+SVM
				5.2.7.6.1 HOG feature extraction
				5.2.7.6.2 SVM classifier classification
			5.2.7.7 Edge detection and background subtraction fusion for pedestrian contour extraction
			5.2.7.8 Establishment of pedestrian pose behavior parameter model and extraction of characteristic parameters
				5.2.7.8.1 Pedestrian posture behavior parameter calculation
				5.2.7.8.2 Feature selection
			5.2.7.9 Pedestrian posture prediction
		5.2.8 Conclusion and outlook
	References
6 - Rail transit inspection robot systems
	6.1 Overview of rail transit inspection robots
		6.1.1 Background of rail transit inspection robots
		6.1.2 Development progress of inspection robots
	6.2 Main components of rail transit inspection robots
		6.2.1 Driving devices
		6.2.2 Sensors
			6.2.2.1 LiDAR sensor
			6.2.2.2 Ultrasonic sensor
				6.2.2.2.1 Ultrasonic ranging principle
				6.2.2.2.2 Determination of the position of obstacles based on ultrasonic ranging
			6.2.2.3 RFID sensor
				6.2.2.3.1 RFID composition
				6.2.2.3.2 How an RFID sensor works
				6.2.2.3.3 Principle of RFID measurement
			6.2.2.4 Sound collection device
		6.2.3 Pan/tilt
			6.2.3.1 Infrared thermal imager sensor
			6.2.3.2 Binocular vision sensor
		6.2.4 Wireless recharging devices
	6.3 Key technologies in rail transit inspection robots
		6.3.1 Navigation
			6.3.1.1 Positioning methods
			6.3.1.2 Navigation methods
			6.3.1.3 Path-planning methods
				6.3.1.3.1 Global path-planning method
				6.3.1.3.2 Local path-planning method
		6.3.2 Hand-eye systems
			6.3.2.1 The role of the hand-eye system
			6.3.2.2 Hand-eye system
				6.3.2.2.1 Eye-to-hand
				6.3.2.2.2 Eye-in-hand
	6.4 Conclusion and outlook
	References
7 - Rail transit channel robot systems
	7.1 Overview of rail transit channel robots
		7.1.1 Development progress of rail transit equipment
			7.1.1.1 Background
			7.1.1.2 Development of the rail transit industry
			7.1.1.3 Significance of vigorously developing rail transit equipment
				7.1.1.3.1 Helping the Chinese economy grow steadily
				7.1.1.3.2 Improving the internationalization of the Chinese rail transit equipment industry
				7.1.1.3.3 Promoting the development of the industrial chain of rail transit equipment industry
				7.1.1.3.4 Promoting the optimization of export trade structure
		7.1.2 Development progress of rail transit channel robots
		7.1.3 Definition of rail transit channel robots
		7.1.4 Main components of rail transit channel robots
		7.1.5 Ground track
		7.1.6 Sensors
			7.1.6.1 Laser sensor
				7.1.6.1.1 Operating principle
				7.1.6.1.2 Advantages and disadvantages of laser sensor
			7.1.6.2 Infrared thermal imager
				7.1.6.2.1 Basic composition and function
				7.1.6.2.2 Working principle
				7.1.6.2.3 Selection of infrared thermal imager
		7.1.7 Autonomous recharging
	7.2 Channel robot TEDS
		7.2.1 Visible light image processing
			7.2.1.1 Image registration
			7.2.1.2 Change detection
		7.2.2 Infrared thermal images processing
			7.2.2.1 Bogie fault location and image segmentation
				7.2.2.1.1 Threshold segmentation
				7.2.2.1.2 Edge detection method
				7.2.2.1.3 Region growing method
			7.2.2.2 Image recognition and classification of bogie fault parts
				7.2.2.2.1 Backpropagation neural network
				7.2.2.2.2 Extreme learning machine
				7.2.2.2.3 Adaptive network-based fuzzy inference system neural network
				7.2.2.2.4 Elman neural network
				7.2.2.2.5 Support vector machine
		7.2.3 Location detection
		7.2.4 Big data analysis
			7.2.4.1 Data collection
			7.2.4.2 Data storage
			7.2.4.3 Data analysis
			7.2.4.4 Model software
	7.3 Bogie fault diagnosis based on deep learning
		7.3.1 Vibration signal acquisition and preprocessing
		7.3.2 Vibration signal feature extraction based on wavelet analysis
			7.3.2.1 Fast Fourier transform
			7.3.2.2 Wavelet transform
		7.3.3 Feature extraction of vibration signal based on empirical mode decomposition
			7.3.3.1 Instantaneous frequency
			7.3.3.2 Intrinsic mode function
			7.3.3.3 Empirical mode decomposition
			7.3.3.4 Advantages and defects of empirical mode decomposition
		7.3.4 Bogie fault diagnosis based on convolutional neural network
			7.3.4.1 Convolutional neural network architecture
			7.3.4.2 Activation function of convolutional neural network
				7.3.4.2.1 Sigmoid activation function
				7.3.4.2.2 Tanh function
				7.3.4.2.3 Rectified linear unit
				7.3.4.2.4 Fault diagnosis result acquisition
	7.4 Conclusion and outlook
	References
8 - Rail transit inspection unmanned aerial vehicle (UAV) systems
	8.1 Overview of inspection unmanned aerial vehicles
	8.2 Main components of rail transit inspection unmanned aerial vehicles
		8.2.1 Structures
			8.2.1.1 Fixed-wing unmanned aerial vehicle
			8.2.1.2 Unmanned helicopter
			8.2.1.3 Rotor unmanned aerial vehicle
		8.2.2 Sensors
			8.2.2.1 Sensors for pan/tilt camera
			8.2.2.2 Sensors for unmanned aerial vehicles
	8.3 Key technologies in rail transit inspection unmanned aerial vehicles
		8.3.1 Communication methods
			8.3.1.1 Communication key technologies
			8.3.1.2 Wireless communication frequency band
			8.3.1.3 Wireless transmission methods
		8.3.2 Data collection methods
			8.3.2.1 Components of the aircraft condition monitoring system
			8.3.2.2 Functions of aircraft condition monitoring system
			8.3.2.3 Demand of aircraft condition monitoring system
		8.3.3 Scheduling methods
			8.3.3.1 Positioning
				8.3.3.1.1 Inertial navigation system/global positioning system integrated positioning
				8.3.3.1.2 Vision positioning
			8.3.3.2 Unmanned aerial vehicle path planning
				8.3.3.2.1 Classical path planning algorithms
				8.3.3.2.2 Reinforcement learning algorithms
			8.3.3.3 Scheduling of unmanned aerial vehicles
				8.3.3.3.1 Problem statement
				8.3.3.3.2 Control methods
	8.4 Rail transit intruding detection based on inspection unmanned aerial vehicles
		8.4.1 Image stabilization
			8.4.1.1 Global motion estimation
			8.4.1.2 Motion compensation
			8.4.1.3 Image generation
		8.4.2 Extraction of region of interest
		8.4.3 Saliency detection
		8.4.4 Unmanned aerial vehicle intruding detection application in rail transit
	8.5 Conclusion and outlook
	References
Index
	A
	B
	C
	D
	E
	F
	G
	H
	I
	K
	L
	M
	N
	O
	P
	Q
	R
	S
	T
	U
	V
	W
	Z
Back Cover