Nesta pesquisa são estudados aspectos relacionados à problemática da localização robótica, e um sistema de localização robótica é construído. Para determinação da localização de um robô móvel em relação a um mapa topológico do ambiente, é proposta uma solução determinística. Esta solução é empregada a fim de prover localização para problemas de rastreamento de posição, embora seja de interesse também a observação da eficácia, do método proposto, frente a problemas de localização global. O sistema proposto baseia-se no uso de vetores de atributos, compostos de medições momentâneas extraídas do ambiente através de sensoriamentos pertencentes à percepção do robô. Estimativas feitas a partir da odometria e leitura de sensores de ultra-som são utilizadas em conjunto nestes vetores de atributos, de forma a caracterizar as observações feitas pelo robô. Uma bússola magnética também é empregada na solução. O problema de localização é então resolvido como um problema de reconhecimento de padrões. A topologia do ambiente é conhecida, e a correlação entre cada local neste ambiente e seus atributos são armazenados através do uso de redes neurais artificiais. O sistema de localização foi avaliado de maneira experimental, em campo, em uma plataforma robótica real, e resultados promissores foram obtidos e são apresentados. / In this research aspects related to the robot localization problem have been studied. In order to determine the localization of a mobile robot in relation to a topological map of its environment, a deterministic solution has been proposed. This solution is applied to provide localization for position tracking problems, although it is also of interest to observe the performance of the proposed method applied to global localization problems. The proposed system is based on feature vectors, which are composed of momentaneous measures extracted from sensory data of the robots perception. Estimative made from odometry, sonars and magnetic compass readings are used together in these feature vectors, in order to characterize observed scenes by the robot. Thus, the localization problem is solved as a pattern recognition problem. The topology of the environment is known, and the correlation between each place of this environment and its features is stored using an artificial neural network. The localization system was experimentally evaluated, in a real robotic platform. The results obtained allow validation of the methodology.

This paper describes a method by which range data from a sonar or other type of rangefinder can be used to determine the 2-dimensional position and orientation of a mobile robot inside a room. The plan of the room is modeled as a list of segments indicating the positions of walls. The method works by extracting straight segments from the range data and examining all hypotheses about pairings between the segments and walls in the model of the room. Inconsistent pairings are discarded efficiently by using local constraints based on distances between walls, angles between walls, and ranges between walls along their normal vectors. These constraints are used to obtain a small set of possible positions, which is further pruned using a test for physical consistency. The approach is extremely tolerant of noise and clutter. Transient objects such as furniture and people need not be included in the room model, and very noisy, low-resolution sensors can be used. The algorithm's performance is demonstrated using Polaroid Ultrasonic Rangefinder, which is a low-resolution, high-noise sensor.

In this thesis, a method to the solution of autonomous navigation problem of a robot working in an outdoor application is sought. The robot will operate in unknown terrain where there is no a priori map present, and the robot must localize itself while simultaneously mapping the environment. This is known as Simultaneous Localization and Mapping (SLAM) problem in the literature. The SLAM problem is attempted to be solved by using the correlation between range data acquired at different poses of the robot. A robot operating outdoors will traverse unstructured terrain, therefore for localization, pitch, yaw and roll angles must also be taken into account along with the (x,y,z) coordinates of the robot. The Iterative Closest Points (ICP) algorithm is used to find this transformation between different poses of the robot and find its location. In order to collect the range data, a system composing of a laser range finder and an angular positioning system is used. During localization and mapping, odometry data is fused with range data.

In this dissertation, the mapping of outdoor environments and localization of a mobile robot in that setting is considered. It is well known that in the absence of a map or precise pose estimates, localization and mapping is a coupled problem. However, in this dissertation this problem is decoupled in to two disjoint steps / mapping and localization on the acquired map. First the images of the outdoor environment is downloaded from a website such as Google Earth and then these images are processed by utilizing several artificial neural network topologies to create maps. Once these maps are obtained, the localization is done by using Monte Carlo localization. This dissertation addresses a solution for the information which is most of the time taken for granted in most studies / a prior map of environment. Mapping is solved by using a novel approach / the map of the environment is created by processing satellite images. Several global localization techniques are developed and evaluated to be used with these map so as to localize a mobile robot globally. The outcome of this novel approach presented here may serve as a virtual GPS. Mobile phone applications can localize a user within a circle of uncertainty without GPS. This crude localization may be used to download relevant satellite images of the local environment. Once the mobile robot is localized on the map created from the satellite images by using available techniques in the literature i.e. Monte Carlo localization, it may be claimed that it is localized on Earth.

A wearable robot should know its environment and its location in order to help its operator. Wearable robots are becoming more feasible with the development of more powerful and smaller computing devices and cameras. The main aim of this research is to build a wearable robot with a low cost stereo camera system which explores a room sized unknown environment online and automatically. To achieve 3D localization and map building for the wearable robot, a consistent visual-SLAM algorithm is implemented by using point features in the environment and Extended Kalman Filter for state estimation. The whole system includes camera models and calibration, feature extraction, depth measurement and Extended Kalman Filter algorithm. Moreover, a map management algorithm is developed. This algorithm keeps the number of features spatially uniform in the scene and adds new features when feature number decreases in a frame. Furthermore, a user-interface is presented so that the location of the camera,the features and the constructed map are visualized online. Most importantly, the system is conducted by a low-cost stereo system.

Nesta pesquisa são estudados aspectos relacionados à problemática da localização robótica, e um sistema de localização robótica é construído. Para determinação da localização de um robô móvel em relação a um mapa topológico do ambiente, é proposta uma solução determinística. Esta solução é empregada a fim de prover localização para problemas de rastreamento de posição, embora seja de interesse também a observação da eficácia, do método proposto, frente a problemas de localização global. O sistema proposto baseia-se no uso de vetores de atributos, compostos de medições momentâneas extraídas do ambiente através de sensoriamentos pertencentes à percepção do robô. Estimativas feitas a partir da odometria e leitura de sensores de ultra-som são utilizadas em conjunto nestes vetores de atributos, de forma a caracterizar as observações feitas pelo robô. Uma bússola magnética também é empregada na solução. O problema de localização é então resolvido como um problema de reconhecimento de padrões. A topologia do ambiente é conhecida, e a correlação entre cada local neste ambiente e seus atributos são armazenados através do uso de redes neurais artificiais. O sistema de localização foi avaliado de maneira experimental, em campo, em uma plataforma robótica real, e resultados promissores foram obtidos e são apresentados. / In this research aspects related to the robot localization problem have been studied. In order to determine the localization of a mobile robot in relation to a topological map of its environment, a deterministic solution has been proposed. This solution is applied to provide localization for position tracking problems, although it is also of interest to observe the performance of the proposed method applied to global localization problems. The proposed system is based on feature vectors, which are composed of momentaneous measures extracted from sensory data of the robots perception. Estimative made from odometry, sonars and magnetic compass readings are used together in these feature vectors, in order to characterize observed scenes by the robot. Thus, the localization problem is solved as a pattern recognition problem. The topology of the environment is known, and the correlation between each place of this environment and its features is stored using an artificial neural network. The localization system was experimentally evaluated, in a real robotic platform. The results obtained allow validation of the methodology.

This thesis implements solution for automatic docking for a mobile robot using visual markers. After initial survey of already implemented works, new docking solution is proposed. Feasibility of the solution is verified with tests of marker detection precision. The implementation is tested in a simulation and with a real robot. The functionality of the proposed solution is verified by long-term tests. The result of this work is robot’s ability to navigate known environment to find and dock a charging station. After charging the robot is able to safely disconnect from the station.

Monte Carlo Localization (MCL) is an algorithm that allows a robot to determine its location when provided a map of its surroundings. Particles, consisting of a location and an orientation, represent possible positions where the robot could be on the map. The probability of the robot being at each particle is calculated based on sensor input. Traditionally, MCL only utilizes the position of objects for localization. This thesis explores using wall and floor surface textures to help the algorithm determine locations more accurately. Wall textures are captured by using a laser range finder to detect patterns in the surface. Floor textures are determined by using an inertial measurement unit (IMU) to capture acceleration vectors which represent the roughness of the floor. Captured texture data is classified by an artificial neural network and used in probability calculations. The best variations of Texture MCL improved accuracy by 19.1\% and 25.1\% when all particles and the top fifty particles respectively were used to calculate the robot's estimated position. All implementations achieved comparable performance speeds when run in real-time on-board a robot.

The necessity of accurate localization in mobile robotics is obvious - if a robot does not know where it is, it cannot navigate accurately to reach goal locations. Robots learn about their environment via sensors. Small robots require small, efficient, and, if they are to be deployed in large numbers, inexpensive sensors. The sensors used by robots to perceive the world are inherently inaccurate, providing noisy, erroneous data or even no data at all. Combined with estimation error due to imperfect modeling of the robot, there are many obstacles to successfully localizing in the world. Sensor fusion is used to overcome these difficulties - combining the available sensor data in order to derive a more accurate pose estimation for the robot. In this thesis, we dissect and analyze a wide variety of sensor fusion algorithms, with the goal of using a set of inexpensive sensors in a suite to provide real-time localization for a robot given unknown sensor errors and malfunctions. The sensor fusion algorithms will fuse GPS, INS, compass and control inputs into a more accurate position. The filters discussed include a SPKF-PF (Sigma-Point Kalman Filter - Particle Filter), a MHSPKF (Multi-hypothesis Sigma-Point Kalman Filter), a FSPKF (Fuzzy Sigma-Point Kalman Filter), a DFSPKF (Double Fuzzy Sigma-Point Kalman Filter), an EKF (Extended Kalman Filter), a MHEKF (Multi-hypothesis Extended Kalman Filter), a FEKF (Fuzzy Extended Kalman Filter), and a standard SIS PF (Sequential Importance Sampling Particle Filter). Our goal in this thesis is to provide a toolbox of algorithms for a researcher, presented in a concise manner. I will also simultaneously provide a solution to a difficult sensor fusion problem - an algorithm that is of low computational complexity (< O(n³)), real-time, accurate (equal in or more accurate than a DGPS (differential GPS) given lower quality sensors), and robust - able to provide a useful localization solution even when sensors are faulty or inaccurate. The goal is to find a locus between power requirements, computational complexity and chip requirements and accuracy/robustness that provides the best of breed for small robots with inaccurate sensors. While other fusion algorithms work well, the Sigma Point Kalman filter solves this problem best, providing accurate localization and fast response, while the Fuzzy EKF is a close second in the shorter sample with less error, and the Sigma-Point Kalman Particle Filter does very well in a longer example with more error. Fuzzy control is also discussed, especially the reason for its applicability and its use in sensor fusion.

Orientador: Jacques Wainer / Dissertação (mestrado) - Universidade Estadual de Campinas, Instituto de Computação / Made available in DSpace on 2018-09-11T21:14:07Z (GMT). No. of bitstreams: 1 Pinheiro_PauloGurgel_M.pdf: 1259816 bytes, checksum: a4783df9aa3755becb68ee233ad43e3c (MD5) Previous issue date: 2009 / Resumo: Em um problema de localização multirrobô cooperativa, um grupo de robôs encontra-se em um determinado ambiente, cuja localização exata de cada um dos robôs é desconhecida. Neste cenário, uma distribuição de probabilidades aponta as chances de um robô estar em um determinado estado. É necessário então, que os robôs se movimentem pelo ambiente e gerem novas observações que serão compartilhadas, para calcular novas estimativas. Nos últimos anos, muitos trabalhos têm focado no estudo de técnicas probabilísticas, modelos de comunicação e modelos de detecções, para resolver o problema de localização. No entanto, a movimentação dos robôs é, em geral, definida por ações aleatórias. Ações aleatórias geram observações que podem ser inúteis para a melhoria da estimativa. Este trabalho apresenta uma proposta de localização com suporte a planejamento de ações. O objetivo é apresentar um modelo cujas ações realizadas pelos robôs são definidas por políticas. Escolhendo a melhor ação a ser realizada, é possível receber informações mais úteis dos sensores internos e externos e estimar as posturas mais rapidamente. O modelo proposto, denominado Modelo de Localização Planejada - MLP, utiliza POMDPs para modelar os problemas de localização e algoritmos específicos de geração de políticas. Foi utilizada a localização de Markov como técnica probabilística de localização e implementadas versões de modelos de detecção e propagação de informação. Neste trabalho, um simulador de problemas de localização multirrobô foi desenvolvido, no qual foram realizados experimentos em que o modelo proposto foi comparado a um modelo que não faz uso de planejamento de ações. Os resultados obtidos apontam que o modelo proposto é capaz de estimar as posturas dos robôs com uma menor quantidade de passos, sendo significativamente mais e ciente do que o modelo comparado sem planejamento. / Abstract: In a cooperative multi-robot localization problem, a group of robots is in a certain environment, where the exact location of each robot is unknown. In this scenario, there is only a distribution of probabilities indicating the chance of a robot to be in a particular state. It is necessary for the robots to move in the environment generating new observations, which will be shared to calculate new estimates. Currently, many studies have focused on the study of probabilistic techniques, models of communication and models of detection to solve the localization problem. However, the movement of robots is generally defined by random actions. Random actions generate observations that can be useless for improving the estimate. This work describes a proposal for multi-robot localization with support planning of actions. The objective is to describe a model whose actions performed by robots are defined by policies. Choosing the best action to be performed, the robot gets more useful information from internal and external sensors and estimates the posture more quickly. The proposed model, called Model of Planned Localization - MPL, uses POMDPs to model the problems of location and specific algorithms to generate policies. The Markov localization was used as probabilistic technique of localization and implemented versions of detection models and information propagation model. In this work, a simulator to multi-robot localization problems was developed, in which experiments were performed. The proposed model was compared to a model that does not make use of planning actions. The results showed that the proposed model is able to estimate the positions of robots with lower number of steps, being more e-cient than model compared. / Mestrado / Inteligencia Artificial / Mestre em Ciência da Computação