Vind
| Autor: | Esten Ingar Grøtli, Jon Azpiazu, Magnus Bjerkeng |
|---|---|
| Rok vydání: | 2016 |
| Předmět: |
0209 industrial biotechnology
Bayes estimator business.industry Computer science Distributed computing 020208 electrical & electronic engineering Real-time computing 02 engineering and technology Kalman filter Extended Kalman filter 020901 industrial engineering & automation 0202 electrical engineering electronic engineering information engineering Robot System integration State (computer science) business Implementation Reusability |
| Zdroj: | ICCMA |
| Popis: | In this paper we present a framework for robot localization codenamed Vind. The framework allows to configure a multi-sensor setup by describing the configuration and entering the sensor's parameters in a series of text-based and human-readable configuration files. The framework provides, among others, distributed communication capabilities and a state estimation implementation based on the Extended Kalman Filter (EKF). Vind can also be extended to include other state estimation implementations based on clearly defined interfaces and message structures. The aim of the framework is to foster reusability, and provide developers with tools to minimize the effort required to deploy a solution for the self-localization problem. In case of researchers working on the implementation of new state estimate algorithms, it also supports them by providing high level tools for the system integration aspects. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|