摘要
简明介绍不停车收费系统的结构及原理,并通过将信息融合技术引进到该系统中来达到各种信息有机融合在一起。但在实际的多传感器融合目标识别中,往往会出现信息融合相冲突的问题,为了解决上述所说的问题,提出一种解决方案,利用改进的D-S证据理论来解决融合过程信息的证据冲突问题,通过引入冲突因子,最大程度地降低矛盾冲突对识别结果的影响。结合所做的计算与仿真效果,结果比较理想,能够有效地化解证据间的冲突问题,克服早期理论的不足之处,提高了融合识别的可靠性和有效性。
The paper introduces the structure and principle of the ETC,And through introducing information fusion technology in the system to achieve organic combination of all kinds of information.But in practical multi-sensor fusion target recognition,it often appears information fusion of conflict.In order to solve the problem,this paper puts forward a solution,Using the improved Dempster-Shafer Arithmetic to solve the problems of the fusion process information evidence,Through introducing the conflict,the method may reduce the influence of the conflict evidence to recognition result.Combining the calculation and simulation done the result of target recognition is ideal.It shows that the extended approach can dissolve the conflict problems and overcome the deficiencies in the early theory to improve the reliability and validity of the fusion recognition.
出处
《微计算机信息》
2010年第20期66-67,73,共3页
Control & Automation
