An Exploratory Robot Controller which Adapts to Unknown Environments and Damaged Sensors

نویسندگان

  • Koren Ward
  • Alexander Zelinsky
چکیده

In this paper we describe an adaptive mobile robot control system that enables a multi-sensor robot to learn reactive behaviours by interacting with the environment. The controller is particularly suitable for exploratory robots due to its ability to adapt to unknown environments and recover from partial sensor damage. Learning is based on the robot learning a map between sensors and trajectory velocities so at any instant the robot becomes capable of realising how fast it should move along its predefined trajectories. Behaviours are performed by selecting trajectories based on their velocity and closeness to a preset behaviour criteria. Unlike reinforcement learning, the map can be obtained relatively quickly by extracting knowledge directly form the environment via the sensors thereby avoiding the credit assignment problem. We demonstrate the effectiveness of this approach to robot learning by using a Yamabico mobile robot to firstly acquire goal seeking behaviour and then recover from damage inflicted on its sensors.

برای دانلود رایگان متن کامل این مقاله و بیش از 32 میلیون مقاله دیگر ابتدا ثبت نام کنید

ثبت نام

اگر عضو سایت هستید لطفا وارد حساب کاربری خود شوید

منابع مشابه

Mobile robot wall-following control using a behavior-based fuzzy controller in unknown environments

This paper addresses a behavior-based fuzzy controller (BFC) for mobile robot wall-following control.The wall-following task is usually used to explore an unknown environment.The proposed BFC consists of three sub-fuzzy controllers, including Straight-based Fuzzy Controller (SFC),Left-based Fuzzy Controller (LFC), and Right-based Fuzzy Controller (RFC).The proposed wall-...

متن کامل

Navigation of a Mobile Robot Using Virtual Potential Field and Artificial Neural Network

Mobile robot navigation is one of the basic problems in robotics. In this paper, a new approach is proposed for autonomous mobile robot navigation in an unknown environment. The proposed approach is based on learning virtual parallel paths that propel the mobile robot toward the track using a multi-layer, feed-forward neural network. For training, a human operator navigates the mobile robot in ...

متن کامل

Basic Issues in Identification Scheme of a Self-Tuning Power System Stabilizer

Power system stabilizers have been widely used and successfully implemented for the improvement of power system damping. However, a fixed parameter power system stabilizer tends to be sensitive to variations in generator dynamics so that, for operating conditions away from those used for design, the effectiveness of the stabilizer can be greatly impaired. With the advent of microprocessor techn...

متن کامل

A New Method of Mobile Robot Navigation: Shortest Null Space

In this paper, a new method was proposed for the navigation of a mobile robot in an unknown dynamic environment. The robot could detect only a limited radius of its surrounding with its sensors and it went on the shortest null space (SNS) toward the goal. In the case of no obstacle, SNS was a direct path from the robot to goal; however, in the presence of obstacles, SNS was a space around the r...

متن کامل

AGV (automated guided vehicle) robot: Mission and obstacles in design and performance

The AGV (automated guided vehicle) was introduced in UK in 1953 for transporting. But nowadays, due to their high efficiency, flexibility, reliability, safety and system scalability, they are used in various application in industries. In brief, the AGV robot is a system which typically made up of vehicle chassis, embedded controller, motors, drivers, navigation and collision avoidance sensors, ...

متن کامل

ذخیره در منابع من


  با ذخیره ی این منبع در منابع من، دسترسی به آن را برای استفاده های بعدی آسان تر کنید

عنوان ژورنال:

دوره   شماره 

صفحات  -

تاریخ انتشار 1998