Neural Networks for Robotics

An Engineering Perspective

Nancy Arana-Daniel, Alma Y. Alanis, Carlos Lopez-Franco

• Publisher: CRC Press
• Year: 2018
• Print ISBN:
  9780367733391
  9780815378686
• Edition: 1
• eBook ISBN:
  9781351231770
• Topic: Technology & Engineering
• Subtopic: Neural Networks
• Index: Technology & Engineering

The book offers an insight on artificial neural networks for giving a robot a high level of autonomous tasks, such as navigation, cost mapping, object recognition, intelligent control of ground and aerial robots, and clustering, with real-time implementations. The reader will learn various methodologies that can be used to solve each stage on autonomous navigation for robots, from object recognition, clustering of obstacles, cost mapping of environments, path planning, and vision to low level control. These methodologies include real-life scenarios to implement a wide range of artificial neural network architectures.

• Includes real-time examples for various robotic platforms.
• Discusses real-time implementation for land and aerial robots.
• Presents solutions for problems encountered in autonomous navigation.
• Explores the mathematical preliminaries needed to understand the proposed methodologies.
• Integrates computing, communications, control, sensing, planning, and other techniques by means of artificial neural networks for robotics.

Contents

Preface

Abbreviations

1 Recurrent High Order Neural Networks for Rough Terrain Cost Mapping

1.1 Introduction

1.1.1 Mapping background

1.2 Recurrent High Order Neural Networks, RHONN

1.2.1 RHONN order

1.2.2 Neural network training

1.2.2.1 Kalman filter

1.2.2.2 Kalman filter training

1.2.2.3 Extended Kalman filter-based training algorithm, EKF

1.3 Experimental Results: Identification of Costs Maps Using RHONNs

1.3.1 Synthetic dynamic environments

1.3.1.1 Synthetic dynamic random environment number 1

1.3.1.2 Synthetic dynamic random environment number 2

1.3.1.3 Synthetic dynamic random environment number 3

1.3.2 Experiments using real terrain maps

1.3.2.1 Real terrain map: grove environment

1.3.2.2 Real terrain map: golf course

1.3.2.3 Real terrain map: forest

1.3.2.4 Real terrain map: rural area

1.4 Conclusions

2 Geometric Neural Networks for Object Recognition

2.1 Object Recognition and Geometric Representations of Objects

2.1.1 Geometric representations and descriptors of realobjects

2.2 Geometric Algebra: An Overview

2.2.1 The geometric algebra of n-D space

2.2.2 The geometric algebra of 3-D space

2.2.3 Conformal geometric algebra

2.2.4 Hyperconformal geometric algebra

2.2.5 Generalization of G_6,3 into G_2n,n

2.3 Clifford SVM

2.3.1 Quaternion valued support vector classifier

2.3.2 Experimental results

2.4 Conformal Neuron and Hyper-Conformal Neuron

2.4.1 Hyperellipsoidal neuron

2.4.2 Experimental results

2.5 Conclusions

3 Non-Holonomic Mobile Robot Control Using Recurrent High Order Neural Networks

3.1 Introduction

3.2 RHONN to Identify Uncertain Discrete-Time NonlinearSystems

3.3 Neural Identification

3.4 Inverse Optimal Neural Control

3.5 IONC for Non-Holonomic Mobile Robots

3.5.1 Robot model

3.5.2 Wheeled robot

3.5.2.1 Controller design

3.5.2.2 Neural identification of a wheeled robot

3.5.2.3 Inverse optimal control of a wheeled robot

3.5.2.4 Experimental results

3.5.3 Tracked robot

3.5.3.1 Controller design

3.5.3.2 Results

3.6 Conclusions

4 Neural Networks for Autonomous Navigation on Non-Holonomic Mobile Robots

4.1 Introduction

4.2 Simultaneous Localization and Mapping

4.2.1 Prediction

4.2.2 Observations

4.2.3 Status update

4.3 Reinforcement Learning

4.4 Inverse Optimal Neural Controller

4.4.1 Planning-Identifier-C...

Table of contents

• Cover
• Halftitle Page
• Title Page
• Copyright
• Table of Contents