Neural Extended Kalman Filter for State Estimation of Automated Guided Vehicle in Manufacturing Environment
Апстракт
To navigate autonomously in a manufacturing environment Automated Guided Vehicle (AGV) needs the ability to infer its pose. This paper presents the implementation of the Extended Kalman Filter (EKF) coupled with a feedforward neural network for the Visual Simultaneous Localization and Mapping (VSLAM). The neural extended Kalman filter (NEKF) is applied on-line to model error between real and estimated robot motion. Implementation of the NEKF is achieved by using mobile robot, an experimental environment and a simple camera. By introducing neural
network into the EKF estimation procedure, the quality of performance can be improved.
Кључне речи:
Automated Guided Vehicles / Monocular SLAM / Extended Kalman Filter (EKF) / Feedforward neural network / The neural extended Kalman filter (NEKF)Извор:
Proceedings of the 35th International Conference on Production Engineering (ICPE-S 2013), 2013, 331-334Издавач:
- Kraljevo : Faculty of Mechanical and Civil Engineering
Финансирање / пројекти:
Колекције
Институција/група
Mašinski fakultetTY - CONF AU - Vuković, Najdan AU - Miljković, Zoran AU - Mitić, Marko AU - Petrović, Milica AU - Husen, Mohamed A. PY - 2013 UR - https://machinery.mas.bg.ac.rs/handle/123456789/4472 AB - To navigate autonomously in a manufacturing environment Automated Guided Vehicle (AGV) needs the ability to infer its pose. This paper presents the implementation of the Extended Kalman Filter (EKF) coupled with a feedforward neural network for the Visual Simultaneous Localization and Mapping (VSLAM). The neural extended Kalman filter (NEKF) is applied on-line to model error between real and estimated robot motion. Implementation of the NEKF is achieved by using mobile robot, an experimental environment and a simple camera. By introducing neural network into the EKF estimation procedure, the quality of performance can be improved. PB - Kraljevo : Faculty of Mechanical and Civil Engineering C3 - Proceedings of the 35th International Conference on Production Engineering (ICPE-S 2013) T1 - Neural Extended Kalman Filter for State Estimation of Automated Guided Vehicle in Manufacturing Environment EP - 334 SP - 331 UR - https://hdl.handle.net/21.15107/rcub_machinery_4472 ER -
@conference{ author = "Vuković, Najdan and Miljković, Zoran and Mitić, Marko and Petrović, Milica and Husen, Mohamed A.", year = "2013", abstract = "To navigate autonomously in a manufacturing environment Automated Guided Vehicle (AGV) needs the ability to infer its pose. This paper presents the implementation of the Extended Kalman Filter (EKF) coupled with a feedforward neural network for the Visual Simultaneous Localization and Mapping (VSLAM). The neural extended Kalman filter (NEKF) is applied on-line to model error between real and estimated robot motion. Implementation of the NEKF is achieved by using mobile robot, an experimental environment and a simple camera. By introducing neural network into the EKF estimation procedure, the quality of performance can be improved.", publisher = "Kraljevo : Faculty of Mechanical and Civil Engineering", journal = "Proceedings of the 35th International Conference on Production Engineering (ICPE-S 2013)", title = "Neural Extended Kalman Filter for State Estimation of Automated Guided Vehicle in Manufacturing Environment", pages = "334-331", url = "https://hdl.handle.net/21.15107/rcub_machinery_4472" }
Vuković, N., Miljković, Z., Mitić, M., Petrović, M.,& Husen, M. A.. (2013). Neural Extended Kalman Filter for State Estimation of Automated Guided Vehicle in Manufacturing Environment. in Proceedings of the 35th International Conference on Production Engineering (ICPE-S 2013) Kraljevo : Faculty of Mechanical and Civil Engineering., 331-334. https://hdl.handle.net/21.15107/rcub_machinery_4472
Vuković N, Miljković Z, Mitić M, Petrović M, Husen MA. Neural Extended Kalman Filter for State Estimation of Automated Guided Vehicle in Manufacturing Environment. in Proceedings of the 35th International Conference on Production Engineering (ICPE-S 2013). 2013;:331-334. https://hdl.handle.net/21.15107/rcub_machinery_4472 .
Vuković, Najdan, Miljković, Zoran, Mitić, Marko, Petrović, Milica, Husen, Mohamed A., "Neural Extended Kalman Filter for State Estimation of Automated Guided Vehicle in Manufacturing Environment" in Proceedings of the 35th International Conference on Production Engineering (ICPE-S 2013) (2013):331-334, https://hdl.handle.net/21.15107/rcub_machinery_4472 .