基于视觉感知系统和模糊逻辑控制器的车辆辅助泊车设计  

Parking Assist System Based on Visual Perception System and Fuzzy Logic Controller

作　　者：马少华[1] 王超[1] 

机构地区：[1]沈阳建筑大学信息与控制工程学院,辽宁沈阳100168

出　　处：《沈阳建筑大学学报（自然科学版）》2011年第5期1000-1004,共5页Journal of Shenyang Jianzhu University：Natural Science

基　　金：国家863计划项目(2009AA011908)

摘　　要：目的解决车辆自动泊车问题.方法建立一种基于多相机和模糊逻辑控制器的车辆辅助泊车系统,利用广角相机采集车辆周围及预泊车位置的实时信息,通过视觉感知系统将现实环境的三维的图像转换为俯视图,并在实时监控的同时为驾驶员提供泊车控制.利用改进的模糊系统自行学习最优模糊算法,提高对非线性、时变性和非确定性的适应能力,降低参数的确定难度.结果达到有效地解决智能车辆泊车控制中的非线性、参数时变和不确定性问题,完成车辆辅助泊车任务.结论视觉感知系统和模糊逻辑控制器的车辆辅助泊车设计系统能有效地提高泊车过程的准确性和安全性.With the purpose of achieving automatic parking, this research develops a parking assist system which is based on the four-camera and the fuzzy logic controller. The real-time information around vehicle and of pre-parking space is obtained by wide-angle cameras. And the three-dimensional images of the actual environment are transformed to top views by visual perception system, which offers the driver parking-con- trolling while proceeding real-time monitoring. The optimal fuzzy algorithm is learned through the improved fuzzy system, which enhances the adaptability of nonlinearity, time-varying property and nondeterminacy and also decreases the difficulties of determining parameters. This design can effectively solve problems of the parking controlling, nonlinearitiy, time varying parameters and uncertainty problems, which achieves the task of parking assist and makes the progress of parking more accurate and safer.

关 键 词：视觉感知系统 辅助泊车 模糊逻辑控制 特征点检测与跟踪 

分 类 号：TP181[自动化与计算机技术—控制理论与控制工程]