A Novel GA-FCM Strategy for Motion Learning and Prediction: Application in Wireless Tracking of Intelligent Subjects

tang, s.h. and motlagh, o. and ismail, n. (2012) A Novel GA-FCM Strategy for Motion Learning and Prediction: Application in Wireless Tracking of Intelligent Subjects. arabian journal of science and engineering. xx-xx. ISSN xxxx-xxxx

Full text not available from this repository.

Abstract

Wireless local area network (WLAN) is used for indoor tracking of mobile terminals (MT), e.g., handheld devices carried by human subjects, mobile robots, automated guided vehicles (AGV), etc. Trilateral radiolocation is suitable for indoors using various parameters of the electromagnetic wave. However, there is always a risk of wireless disconnection between MT and access points (AP). WLAN-based tracking is therefore vulnerable to disconnection and might fail in subject tracking at times. Modification of physical network reduces risk of disconnection. Alternatively, complementary techniques based on soft-computing provide even more robustness for uninterrupted wireless tracking without physical modification. This article presents a novel path prediction approach based on decision support system (DSS) to resolve this problem. The goal is to use the knowledge of past MT trajectory to compensate for missing information during disconnections. By learning kinematical as well as decisional behaviors of an intelligent MT, its future trajectory is predictable. The system has been evaluated using trajectories of an ActivMedia Pioneer robot, and blind-folded walking locus in indoors as examples of intelligent MTs. Locational error in prediction of human motion is limited to 1m. The result is comparable with the related works with up to 10% improvement in locational accuracy. Robot motion prediction has been obtained with high accuracy (likelihood) of 80-90% which further shows the robustness of the developed system.

Item Type: Article
Subjects: Q Science > QA Mathematics > QA75 Electronic computers. Computer science
Divisions: Faculty of Manufacturing Engineering > Department of Robotics and Automation
Depositing User: Omid Motlagh
Date Deposited: 25 Jul 2013 00:19
Last Modified: 25 Jul 2013 00:19
URI: http://eprints.utem.edu.my/id/eprint/8849
Statistic Details: View Download Statistic

Actions (login required)

View Item View Item