Study On Technology Information Fusion From Multi-sensor For Targetlocalization Of Mobile Robots

Recently, multi-information fusion technology of multi-sensors in the target location of mobile robot has become a hot topic, which combines the knowledge and development of computer science, sensors, signal system and error analysis. During the process of the target location of mobile robot, in order to obtain more reliable and precise information,multi-information fusion technology of multi-sensors is used for fusing the data obtained from the sensors; meanwhile, for the sake of high efficiency of mobile robot, energy management strategy should be used for lowering the energy consumption.The key point of this dissertation is multi-information fusion technology of multi-sensors during the process of the target location of mobile robot, which analysis and probe the process from two aspects: information fusion error analysis and mobile robot energy consumption management. The main works are as follows:Firstly, this dissertation gives a brief introduction of the research situation at home and abroad in the fields of mobile robot and multi-sensors fusion technology. And then we do some research on the dynamic model of mobile robot, power management model and the operating principle of the main sensor, we also describe the hierarchical structure model of the multi-sensors fusion of mobile robot and analysis the major foundation models needed by the mobile root in the process of target location, such as the coordinate system model,the motion model, the sensor model, the data correlation model and so on. We focus on the analysis of relevant algorithms of multi-sensors fusion applied in mobile robot and the power management strategy applied in the embedded system of mobile robot.Secondly, on account of the properties that unexpected errors may happen during the process of target location in the multi-sensors fusion, we use the state estimation technique of information fusion method based on the Extend Kalman Filter to analysis. This dissertation uses the subsidiary sensor for correcting the parameters and forecasting the fusion results to make up for the inadequacy of existing accumulated error control method,proposes an accuracy-oriented Kalman Filtering strategy to estimate the actual position of mobile robot. The simulation experiment shows the effectiveness of this method when dealing with the unexpected errors happened in the process of target location of mobilerobot.Thirdly, based on the analysis of embedded system of mobile robot, on account of how to improve the energy consumption effectiveness of multi-information fusion, we use the embedded power management model to analysis the energy consumption of mobile robot.Based on embedded power management model, in order to conquer the problem that peripheral sensors can not manage energy consumption according to the need of Qo S(Quality of Service), we improve a dynamic energy consumption strategy applied in mobile robot. By means of experiment, without affecting the proper functioning of the system, this strategy can reduce the energy consumption of mobile robot effectively.Consequently the working time of mobile robot can be extended.Last but not least, we summarize the researching works in this dissertation and discuss the future content and direction of the further research.