State Estimation Solely Based on Prior Knowledge and Inertial Sensors - read free eBook in online reader or directly download on the web page. Select files or add your book in reader. Download and read online ebook State Estimation Solely Based on Prior Knowledge and Inertial Sensors write by Tom Lucas Koller. This book was released on 2022. State Estimation Solely Based on Prior Knowledge and Inertial Sensors available in PDF, EPUB and Kindle. How do we localize ourselves? Ever since GPS exists, it is common to know where we are and how to get to our desired location. Unfortunately, GPS is unavailable indoors. Scientists are looking for an alternative technology that can fill this localization gap. One approach is to fuse knowledge about our environment and our movement measured with inertial sensors. A particular difficulty of these sensors is that their pose (position and orientation) estimation error grows over time. This so-called drift can lead to false estimations such as passing through a wall. These false estimations could be corrected by using prior knowledge of the wall's location. In this work, I investigate how prior knowledge can be fused with inertial sensor measurements. The practical aim of this thesis is to eliminate the drift without additional sensors. I investigate three types of prior knowledge regarding the environment and the movement: The human gait pattern, terrain maps, and event-domain maps. For all three types, I follow the concept of modeling the prior knowledge as probability distributions of the system's state. This modeling enables the usage of standard probability-based algorithms to estimate the position and orientation and to fuse the knowledge with sensor measurements. The human gait is an alternating pattern of stance and swing phases. I show a new approach based on the Interacting Multiple Model Filter that can detect the phase and improve the velocity estimate of the inertial sensor. The approach automatically detects whether the sensor measurements match the probability distribution of the stance or swing phase. Simultaneously, it corrects the measurement errors of the inertial sensor by taking into account the probability distributions. The evaluation shows the potential of this method, albeit further development is required to outperform state of the art approaches. Terrain maps define the height of a vehicle or a human given its position in the horizontal plane. This can be modeled as a so-called pseudo measurement. We act like there is a sensor that measures the height above the surface but always returns zero since there is no height difference. In this way, a probability distribution is modeled that constrains the position to the surface. I investigate terrain maps with the practical example of track cycling. I show that terrain maps can yield full observability of the position and orientation; in other words, that they are able to correct the growing error of the inertial sensor. Thereby, only the curved parts of the track yield information about the position. As a result, the position can be tracked during 10km drives with an error of 1:08m (RMSE). Event-domain maps are a particular type of maps that specify where activities can be performed. For example, it is only possible to climb stairs at staircases. I investigate this type of knowledge at bouldering, where the climbers grip the holds of a route. The map represents a probability distribution of possible grip positions. I develop a two-step method where the first step estimates the transition between two holds. In a second step, the transitions are refined using the event-domain map. The estimated error improves from 0:266m (median) to 0:132m compared to an integrating solution without a map. Overall, modeling the three types of prior knowledge successfully reduces the drift in all cases. The human gait pattern can be utilized with a new kind of state estimator, which needs further investigation. The map-based types of knowledge correct the drift of the inertial sensor in the experiments. For the terrain map, it is even possible to prove the correction mathematically. This shows that prior knowledge modeled as prior distribution is effective to estimate the position solely with inertial sensors.
State Estimation for Robotics
State Estimation for Robotics - read free eBook in online reader or directly download on the web page. Select files or add your book in reader. Download and read online ebook State Estimation for Robotics write by Timothy D. Barfoot. This book was released on 2017-07-31. State Estimation for Robotics available in PDF, EPUB and Kindle. A modern look at state estimation, targeted at students and practitioners of robotics, with emphasis on three-dimensional applications.
Biomechanics of Anthropomorphic Systems
Biomechanics of Anthropomorphic Systems - read free eBook in online reader or directly download on the web page. Select files or add your book in reader. Download and read online ebook Biomechanics of Anthropomorphic Systems write by Gentiane Venture. This book was released on 2018-08-01. Biomechanics of Anthropomorphic Systems available in PDF, EPUB and Kindle. Mechanical laws of motion were applied very early for better understanding anthropomorphic action as suggested in advance by Newton «For from hence are easily deduced the forces of machines, which are compounded of wheels, pullies, levers, cords, and weights, ascending directly or obliquely, and other mechanical powers; as also the force of the tendons to move the bones of animals». In the 19th century E.J. Marey and E. Muybridge introduced chronophotography to scientifically investigate animal and human movements. They opened the field of motion analysis by being the first scientists to correlate ground reaction forces with kinetics. Despite of the apparent simplicity of a given skilled movement, the organization of the underlying neuro-musculo-skeletal system remains unknown. A reason is the redundancy of the motor system: a given action can be realized by different muscle and joint activity patterns, and the same underlying activity may give rise to several movements. After the pioneering work of N. Bernstein in the 60’s on the existence of motor synergies, numerous researchers «walking on the border» of their disciplines tend to discover laws and principles underlying the human motions and how the brain reduces the redundancy of the system. These synergies represent the fundamental building blocks composing complex movements. In robotics, researchers face the same redundancy and complexity challenges as the researchers in life sciences. This book gathers works of roboticists and researchers in biomechanics in order to promote an interdisciplinary research on anthropomorphic systems at large and on humanoid robotics in particular.
State Estimation and Sensor Selection in Discrete Event Systems Modeled by Petri Nets
State Estimation and Sensor Selection in Discrete Event Systems Modeled by Petri Nets - read free eBook in online reader or directly download on the web page. Select files or add your book in reader. Download and read online ebook State Estimation and Sensor Selection in Discrete Event Systems Modeled by Petri Nets write by Yu Ru. This book was released on 2010. State Estimation and Sensor Selection in Discrete Event Systems Modeled by Petri Nets available in PDF, EPUB and Kindle. A discrete event system (DES) is a dynamic system that evolves in accordance with the abrupt occurrence, at possibly unknown and irregular intervals, of physical events. Such systems arise in a variety of contexts, such as energy distribution networks, computer and communication networks, automated manufacturing systems, air traffic control systems, highly integrated command, control, communication, and information (C3I) systems, advanced monitoring and control systems in automobiles or large buildings, intelligent transportation systems, and distributed software systems. Petri net models are widely used for modeling such systems, and consist of two key components: places (which typically model buffers that store system resources) and transitions (which typically model activities that move and process resources across places in the system). Sensors in Petri nets come in two major types: place sensors (i.e., sensors that indicate the number of resources in a particular place, e.g., vision sensors) and transition sensors (i.e., sensors that can detect whether a transition in a given subset of transitions has occurred, e.g., motion sensors). In this dissertation, we focus on two sensor related problems in discrete event systems modeled by Petri nets: (i) State estimation. When only transition sensors are available, sensor information can be very limited because there can be uncertainty due to unobservable events or events that generate the same sensor information. As a result, multiple states could be possible given sensing information, and we show in this dissertation that the number of possible states can grow at most polynomially (but not exponentially) as a function of the length of the observation sequence. These polynomial bounds can guide the design of systems, especially when trying to configure the sensors in order to reduce the uncertainty introduced in the state estimation stage. The polynomial bounds can also be used in analyzing algorithms in the context of state estimation, fault diagnosis, supervisory control, and even reachability checking. (ii) Sensor selection. If there are only transition sensors with uncertainty, the system state is usually not unique. If we have the freedom to configure sensors (e.g., when we design the system), we might want to add a minimal number of sensors to ensure that the current system state can be uniquely reconstructed based on the system model and the initial state. The design consideration is motivated by supervisory control applications, interface design for safety critical systems, and certain fault detection and correction settings. In its most general form, this type of sensor selection problem can involve both place sensors and transition sensors. We study how to choose a minimum number of place sensors and transition sensors (or a set of place sensors and transition sensors of minimal cost) while ensuring that the system state can be determined uniquely given sensing information and knowledge of the system model; this property is called structural observability. We show that the general sensor selection problem is computationally hard. If we are given a fixed set of transition sensors and are interested in selecting place sensors from a given set to achieve structural observability, the problem can be solved optimally by linear integer programming solvers, or suboptimally by heuristic methods we propose. On the other hand, if we have a fixed set of place sensors and then select transition sensors, the problem is solvable with complexity that is polynomial in the number of places and transitions. Among other potential applications, the heuristic methods we propose have implications for sensor selection to achieve immediate diagnosis of faults, reduct calculation in rough set theory, and approximating solutions for other NP-complete problems.
Computer Vision
Computer Vision - read free eBook in online reader or directly download on the web page. Select files or add your book in reader. Download and read online ebook Computer Vision write by Md Atiqur Rahman Ahad. This book was released on 2024-07-30. Computer Vision available in PDF, EPUB and Kindle. Computer vision has made enormous progress in recent years, and its applications are multifaceted and growing quickly, while many challenges still remain. This book brings together a range of leading researchers to examine a wide variety of research directions, challenges, and prospects for computer vision and its applications. This book highlights various core challenges as well as solutions by leading researchers in the field. It covers such important topics as data-driven AI, biometrics, digital forensics, healthcare, robotics, entertainment and XR, autonomous driving, sports analytics, and neuromorphic computing, covering both academic and industry R&D perspectives. Providing a mix of breadth and depth, this book will have an impact across the fields of computer vision, imaging, and AI. Computer Vision: Challenges, Trends, and Opportunities covers timely and important aspects of computer vision and its applications, highlighting the challenges ahead and providing a range of perspectives from top researchers around the world. A substantial compilation of ideas and state-of-the-art solutions, it will be of great benefit to students, researchers, and industry practitioners.
