摘要
为了能够减少交通流量错误估计造成的经济损失,准确的掌握交通流分布情况,多传感器信息融合的理论被引入到交通流的检测中,并且以此为基础,利用Bayes参数估计理论对交通流分布情况进行估计。在交通流量符合二项式分布的情况下,用Bayes分别估计了交通流中的参数无先验概率(先验概率为均匀分布)和先验概率为正态分布两种情况下的交通流量。通过仿真实验发现,在先验概率已知的情况下,运用Bayes估计理论对交通流进行检测,可以得到与实际情况相符的车流量分布情况,并且准确率较高,而且运算速度快,具很强的有实时性和实用性。
In order to reduce the economic loss, caused by the mistake in traffic volume estiamtion, and estimate the traffic volume exactly, the multi -sensor information fusion is introduced. And Bayes estimation theory is utilized in traffic flow detection based on the multi - sensor information fusion. When the traffic flow distribution conforms to binomial distribution, the traffic volume is estimated by Bayes estimation in no prior distribution （uniform distribution） and normal prior distribution. It is shown in simulation experiment that when the prior distribution is known, the traffic volume distribution accords with the actual tragic volume. The accuracy is good and the speed of operation is high. It could be completely applied in the practice.
出处
《计算机仿真》
CSCD
北大核心
2009年第2期286-288,297,共4页
Computer Simulation
基金
甘肃省教育厅硕士生导师资助计划项目(0503B-01)
甘肃省自然基金项目(0710RJZA060)
关键词
贝叶斯估计
二项式分布
正态分布
交通流检测
Bayes estimation
Binomial distribution
Normal distribution
Traffic flow detection
