This paper describes a new approach for mobile robot navigation using an interval analysis based adaptive mechanism for an extended Kalman filter. The robot is equipped with inertial sensors, encoders and ultrasonic sensors. The map used for this study is two-dimensional and it is assumed to be a known a-priori. Multiple sensor fusion for robot localisation and navigation has attracted a lot of interest in recent years. An extended Kalman filter (EKF) is used to estimate the robots position using the inertial sensors and encoders. Since the EKF estimates are affected by bias, drift etc. we propose an adaptive mechanism using interval analysis to correct these defects in estimates. Interval analysis has been already successfully used in the past for robot localisation using time of flight ultrasonic sensors. One of the problems of the use of interval analysis sensor based navigation and localisation is that it can be applicable only in the presence of landmarks. This problem is overcome here using additional sensors such as encoders and inertial sensors, which gives an estimate of the robot position using an extended Kalman filter in the absence of landmarks. In the presence of landmarks the complementary robot position information from the interval analysis algorithm using ultrasonic sensors is used to estimate and bound the errors in the EKF robot position estimate.

/pdf/sensor-based-robot-localisation-and-navigation-using-1fa0atrw67.pdf

Sensor based robot localisation and navigation: using interval analysis and unscented Kalman filter

By the extrapolation of movement increments detected by differential encoders, the position of a mobile robot can be easily computed. However, the encoders suffer from various systematic errors resulting in an increasing error of the obtained robot position. This problem is also known from the scope of inertial navigation, but the transfer of the respective concepts of maintaining merely errors to mobile robot localization has, with few exceptions, been neglected so far. In this paper, the position error is related to the encoder errors in an augmented state-space system. As done in inertial guidance, the position error is estimated by an error-state Kalman filter making use of general complementary sensor information as, for example, absolute position measurements or redundantly detected movement increments. Specifically, the usage of a rate gyroscope as a complementary sensor is examined. This case is somehow special, as it implies a correlation of the overall error model and the observation, demanding ...

Quantification and Estimation of Differential Odometry Errors in Mobile Robotics with Redundant Sensor Information

This paper presents a novel line of sight control system for a robot vision tracking system, which uses a position feedforward controller to preposition a camera, and a vision feedback controller to compensate for the positioning error. Continuous target tracking is an important function for service robots, surveillance robots, and cooperating robot systems. However, it is difficult to track a specific target using only vision information, while a robot is in motion. This is especially true when a robot is moving fast or rotating fast. The proposed system controls the camera line of sight, using a feedforward controller based on estimated robot position and motion information. Specifically, the camera is rotated in the direction opposite to the motion of the robot. To implement the system, a disturbance compensator is developed to determine the current position of the robot, even when the robot wheels slip. The disturbance compensator is comprised of two extended Kalman filters (EKFs) and a slip detector. The inputs of the disturbance compensator are data from an accelerometer, a gyroscope, and two wheel-encoders. The vision feedback information, which is the targeting error, is used as the measurement update for the two EKFs. Using output of the disturbance compensator, an actuation module pans the camera to locate a target at the center of an image plane. This line of sight control methodology improves the recognition performance of the vision tracking system, by keeping a target image at the center of an image frame. The proposed system is implemented on a two-wheeled robot. Experiments are performed for various robot motion scenarios in dynamic situations to evaluate the tracking and recognition performance. Experimental results showed the proposed system achieves high tracking and recognition performances with a small targeting error.

A novel line of sight control system for a robot vision tracking system, using vision feedback and motion-disturbance feedforward compensation

This paper introduces a high performance vision tracking system for mobile robot using sensor data fusion. For mobile robots, it is difficult to collect continuous vision information due to robot's motion. To solve this problem, the proposed vision tracking system estimates the robot's position relative to a target and rotates the camera towards the target. This concept is derived from the human eye reflex mechanism, known as the Vestibulo-Ocular Reflex (VOR), for compensating the head motion. This concept for tracking the target results in much higher performance levels, when compared with the conventional method that rotates the camera using only vision information. The proposed system do not require heavy computing loads to process image data and can track the target continuously even during vision occlusion. The robot motion information is estimated using data from accelerometer, gyroscope, and encoders. This multi-sensor data fusion is achieved using Kalman filter. The proposed vision tracking system is implemented on a two-wheeled robot. The experimental results show that the proposed system achieves excellent tracking and recognition performance in various motion scenarios, including scenarios where camera is temporarily blocked from the target.

High performance vision tracking system for mobile robot using sensor data fusion with Kalman filter

This work describes a new approach for mobile robot navigation using interval analysis and fuzzy logic. The robot is equipped with inertial sensors, encoders and ultrasonic sensors. The map used for this study is two-dimensional and it is assumed to be known. Multiple sensor fusion for robot localisation and navigation has attracted a lot of interested in recent years. The most commonly used approach is based on Kalman filter and other stochastic filters. Here we propose an alternative approach using interval analysis with multiple sets of ultrasonic measurements. Interval analysis has been already successfully applied in the past for robot localisation. But the results obtained may be conservative. Therefore this approach is extended using multiple sets of ultrasonic measurements, which results in estimation of multiple interval robot positions. These multiple interval robot positions are then fused using fuzzy logic to give a less conservative interval robot position estimate. Also interval analysis based algorithm can be used only in the presence of land marks. This problem is overcome here using additional sensors such as encoders and inertial sensors, which gives an estimate of the robot position using fuzzy logic in the absence of land marks.

Feature based robot navigation: using fuzzy logic and interval analysis

In this paper the problem of integrated navigation for mobile robotics is addressed. The dynamics of ground robots are nonlinear, as well as the measurement devices. For the purposes of position estimation, a dead reckoning problem with a wheeled mobile robot is considered using a gyro and two wheel encoders. The extended Kalman filter (EKF) is used to provide an estimate of vehicle state, gyro error dynamics and variables describing the slip effects. The parameters of the error dynamics have been identified and validated off-line by real data. Simulations results have confirmed the efficacy of the proposed approach, by comparing the performances of the EKF with and without a wheel slip model.

A. Vinci

Papers

An application of the extended Kalman filter for integrated navigation in mobile robotics