执勤行车时间的KMP-RBF融合预测方法

发布时间：2018-05-22 09:23

本文选题：信息融合 + 预测　；参考：《华南理工大学学报(自然科学版)》2017年03期

【摘要】：针对执勤车辆沿某一路径行驶时长难以预测的问题,文中提出一种KMP-RBF融合方法,采用GPS定位、悉尼自适应交通控制系统(SCATS)线圈作为融合信号源,建立路况信息选择融合模型,将模糊推理知识表达、MAPSO算法寻优和RBF网络训练相结合,自适应优化系统关键参数,从训练数据库匹配适用时间、空间数据.实验中,使用交通监控系统实测实验用车行驶时长数据,并与预测数据进行对比,从误差率、算法迭代与精确度方面证明文中方法是高效而可靠的.
[Abstract]:In order to solve the problem that it is difficult to predict the length of running time of an operating vehicle along a certain path, a KMP-RBF fusion method is proposed in this paper, which adopts GPS positioning and Sydney adaptive traffic control system (SCATS) coil as the fusion signal source, and establishes the road condition information selection fusion model. By combining fuzzy reasoning knowledge expression algorithm with RBF network training, the key parameters of the system are adaptively optimized to match the applicable time and spatial data from the training database. In the experiment, the traffic monitoring system is used to measure the experimental vehicle driving time data, and compared with the prediction data, the error rate, algorithm iteration and accuracy prove that the method in this paper is efficient and reliable.
【作者单位】：天津大学电气自动化与信息工程学院;天津市公安消防局;
【基金】：国家自然科学基金资助项目(61201179) 国家博士后科学基金资助项目(2016M601265)~~
【分类号】：TP18;U491

【相似文献】