Can Ulas Dogruer

Global urban localization of outdoor mobile robots using satellite images

Localization is one of the major research fields in mobile robotics. With the utilization of satellite images and Monte Carlo localization technique, the global localization of an outdoor mobile robot is studied in this paper. The proposed method employs satellite images downloaded from the Internet to localize the robot iteratively. To accomplish this, the proposed method matches the local laser scanner data with the segmented satellite images. Initial test results conducted on the METU campus are found to be quite promising. Further improvement of this approach has the potential of cutting down not only the operational costs but also the preparation period of the mobile robot enabling researchers to operate their robots in diverse outdoor settings.

Global Urban Localization of an Outdoor Mobile Robot with Genetic Algorithms

The localization of mobile robots has been studied rigorously in the past. However, only a few studies have focused on developing specific Genetic Algorithms (GAs) to address the localization problem effectively. In this study; the global urban localization of an outdoor mobile platform is considered with the utilization of the odometer, the laser-rangeq finder measurements and the digital maps created from the relevant satellite images on the Internet. The localization issue is formulated as a constrained optimization problem. The study proposes a GA-based technique to solve the problem at hand efficiently.

Online Identification of Odometer Parameters of a Mobile Robot

In this paper, odometer parameters of a differential drive mobile robot are learned in an online fashion. EKF which is designed using nominal values of odometer, estimates pose of the mobile robot. A second open-loop model that tracks EKF filter is designed. This open-loop tracking system updates its parameter so as to track the states of the system estimated by EKF. As the parameters of the open-loop system is learned, nominal values of parameters of the EKF plant model are updated with these learned values. Hence a cascaded closed-loop system is proposed. In order to verify the results, a simulink model is developed and performance of the proposed adaptive learning system is investigated. It is seen that regular EKF filter diverges even under mild parameter uncertainty whereas the cascaded closed loop system is stable against severe parameter uncertainty.

Novel solutions for Global Urban Localization

In this study, novel solutions to Global Urban Localization problem is proposed and examined rigorously. Classical approaches including Particle Filter, mixture of Gaussians, as well as novel solutions like Viterbi Algorithm and differential evolution are evaluated. The contribution of this paper is twofold: The Viterbi algorithm is extended by exploiting the structure of the problem at hand that is the states are partially connected temporally. Differential evolution is modified by taking into account the covariance matrix of states. Thus states encoded in genes are only allowed to interact locally within the region described by covariance matrix. This prevents the differential evolution from getting trapped into false maxima in the early stages of optimization. Finally, it is demonstrated with extensive experiments that solution of Global Urban Localization problem is possible.

A novel soft-computing technique to segment satellite images for mobile robot localization and navigation

Localization of mobile robots has been studied rigorously in the last decade. A number of successful approaches such as Extended Kalman Filter, Markov Localization, and Monte Carlo Localization assume that the map of the environment is originally presented to the robot. However, an important information package like the map of the environment could not be taken for granted in most real- world problems. In this study, a novel technique composed of a combination of Fuzzy C-Means and Fuzzy Neural Network methods is proposed to segment and convert a satellite image into a digital map for outdoor mobile robot applications.

Optimal trajectory tracking under parametric uncertainty

The kinematic model of a wheeled mobile robot is considered to be not very accurate; this imprecision is due to several uncertain geometric parameters expressed in the form of scale factors. Naturally, these geometric uncertainties have an adverse influence on the output computed using this kinematic model. In this paper, the imprecise inverse kinematic model of a wheeled mobile robot is studied in discrete-time; an optimal controller is designed to enhance the tracking performance of a mobile robot. In the final part of this paper, the analytical solution of this problem is tested with computer simulations.

Outdoor mapping and localization using satellite images

Recently, satellite images of most urban settings has become available on the internet. In this study, a novel mapping and global localization approach, which uses these images, is proposed for outdoor mobile robots operating in urban environment. The mapping of large-scale outdoor environments is done by employing the satellite images acquired by remote sensing technology, and then a map-based approach, that is, Monte Carlo localization is used for localization. The novelty of proposed method is that it uses standard equipment present on almost all autonomous robots and satellite images thus it acts as an alternative to GPS data in urban environments. Extensive field tests are presented to demonstrate the effectiveness of proposed approach.

Multiple model adaptive estimator with model set update

Multiple Model Adaptive Estimator (MMAE) which is the simplest robust estimator, relies on a finite number of time-invariant models which approximate the true mode of a system. Filters in MMAE update relative importance of their estimate, when new measurements are available. However, MMAE has got a basic drawback; each filter in MMAE is associated with a time-invariant model of the system. Furthermore, the prototype models are created with few information available about the system, before MMAE is started. In this paper, a method is proposed to update parameters of these models and move them in the domain of interest where they can approximate the true mode of a system better. To this end, an error model is designed and least square estimation is employed to update those models used by MMAE. This robust estimator which is proposed in this paper, was tested on a typical mobile robotic problem, i.e. estimation of the odometry model parameters.

Estimation of odometer parameters with MMAE and LSE

Extended Kalman filter is used intensively to achieve optimal sensor fusion to estimate the states of plant. In general, parameters of sensor and plant models are inaccurate so biased and random errors are inevitable unless they are calibrated accurately. In this paper, biased parameters of plant are estimated with Multiple-Model-Adaptive-Estimation algorithm (MMAE) and Least Square Estimation (LSE). It is shown that proposed method can learn the parameters of a differential-drive mobile robot odometer e.g. scale factors of left and right wheel radii and distance between wheels, accurately.