摘要
随着道路里程不断增加,机动车辆成为了主要的交通方式。为了满足机动车更好的行驶和道路的最大化利用,首先运用现阶段的车联网大数据等高科技来获取交通流数据,对错误的数据进行删除、丢失少量数据忽略不计,得到有效的数据,并进行融合处理;其次,先用单一的支持向量机对道路交通状态进行分类识别,同时用蚁群算法对支持向量机中的各参数进行最终寻优,得到ACO-SVM状态分类识别融合模型;最后,先用单一的支持向量机对各级别的交通状态的分类精度进行仿真分析计算,结果为91.33%,然后用ACO-SVM融合模型进行分析计算得到在经过14次的迭代时可以找到最优解c=6.884,δ=0.731,将分类精度作为蚁群算法的适应度函数值,最终分类精度比单一SVM模型有所提升并达到94.6%。仿真分析结果证明了该融合模型的有效性。
As road mileage continues to increase,motor vehicles have become the main mode of transportation. In order to meet the needs of motor vehicles for better driving and maximizing the use of roads,we first use the high-tech technologies such as the Internet of Things to obtain traffic flow data,delete the wrong data,lose a small amount of data,and obtain valid data. Secondly,the road traffic state is classified and identified by a single support vector machine. At the same time,the ant colony algorithm is used to finally optimize the parameters in the support vector machine to obtain the ACO-SVM state classification recognition fusion model. Finally,the classification accuracy of each level of traffic state is 91.33% by single SVM,and then the ACO-SVM fusion model is analyzed and calculated. After 14 iterations,the optimal solution is c=6.884,δ=0.731. The classification accuracy is used as the fitness function value of the ant colony algorithm. The final classification accuracy is improved by 94.6% compared with the single SVM model. The results of simulation analysis show the effectiveness of the fusion model.
作者
运杰伦
郭元术
林欣欣
YUN Jie-lun;GUO Yuan-shu;LIN Xin-xin(School of Information Engineering,Chang’an University,Xi’an 710000,China)
出处
《计算机技术与发展》
2020年第7期17-20,共4页
Computer Technology and Development
基金
江西省交通运输科技项目(2014C0002)。
关键词
蚁群算法
SVM
融合算法
状态识别
路径优化
ant colony algorithm
SVM
fusion algorithm
state recognition
path optimization
