基于多信息融合模型的SLAM算法研究及控制器设计

The Research of SLAM Algorithm Based on Multi-information Integrated Cognitive Model and the Design for Its Controller

针对无人驾驶车辆SLAM算法中全局定位感知系统存在的问题,提出一种融合多信息认知模型的广义SLAM算法来设计控制器结构,并且利用道路结构化环境下智能传感器节点作为环境地图的一部分,快速构建道路环境地图,采用基于局部视觉的辅助定位导航方法进行车辆辅助定位,从而提高控制器全局定位的感知能力和系统定位响应速度。仿真计算表明,控制器可以得到车辆位置似然度的最佳值。

Aiming at the problems existing in the whole Localization apperceive system of Simultaneous Localization and Mapping（SLAM）algorithm for unmanned vehicle, this paper puts forward a generalized SLAM algorithm based on a fusion multi-information integrated cognitive model to design the controller structure, utilizes the intelligent sensor node under the road strueturalized environment as a part of environmental map, builds the road environmental map fastly, and adopts an auxiliary location and navigation method based on partial visual sense to assisting localization, thereby improves the control- ler＇s ability of whole localization apperceive and the system＇s localization response speeds. Simulation calculation indicates that the controller could gain the best value of unmanned vehicle position likelihood.