基于字典学习的车辆交通真实性评价模型

发布时间：2018-01-06 16:37

本文关键词：基于字典学习的车辆交通真实性评价模型　出处：《浙江大学》2017年硕士论文　论文类型：学位论文

【摘要】：近些年,车辆交通群组动画在各行各业得以广泛应用。关于交通模拟的研究也越来越引人注目,包括基于宏观特征和微观模型在内的许多新颖的模拟方法被提了出来。在某些实际应用中,如城市交通管理,对动画的真实性要求非常之高。然而,关于交通模拟的真实性评价的工作却鲜有人研究。本文提出了一个完整的车辆交通群组动画的真实性评价模型。本模型参考了交通模拟中纹理合成的方法,将车流数据统一进行图像信号化处理。首先,从大量真实车流数据中提取一些低维特征信号。然后,通过字典学习算法从真实信号中提出车流结构性特征。最后,这些特征与测试信号进行比较,通过测试信号的重构残差计算出一个评价值,并以此来实现对车辆交通动画真实性的评价。本文为车流数据的模拟提供了一套完善客观的评价系统和反馈纠正机制,以促进这一研究方向的发展。并为车辆交通群组动画的应用如城市交通的治理起到积极的作用。此外,本文使用字典学习算法来评价车流数据的思想,也为整个群组动画方向的研究开辟了一条崭新的思路。
[Abstract]:In recent years, traffic group animation has been widely used in all walks of life. The research on traffic simulation is more and more noticeable, including many novel features and micro macro simulation method, based on the model is proposed. In some practical applications, such as city traffic management, real demand for animation is very high however. True, evaluation about traffic simulation work but few research. This paper proposes a complete evaluation of real vehicle traffic group animation model. The model reference texture synthesis method in traffic simulation, the traffic data unified signal processing image. Firstly, some low dimensional feature extraction from the signal a large number of real traffic data. Then, proposed traffic structural features from the real signal through dictionary learning. Finally, the characteristics and the test signals are compared through An evaluation value calculation of residual test signal reconstruction, in order to achieve the evaluation of vehicle traffic animation authenticity. This paper provides a complete and objective evaluation system for traffic flow data and feedback mechanism, to promote the development of the study and application. For group animation such as vehicle traffic city traffic management play a positive role. In addition, this paper use a dictionary to evaluate traffic data the idea of learning algorithm, but also opens up a new way to study the whole group animation direction.

【学位授予单位】：浙江大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TP391.41

【参考文献】