基于自适应模糊理论的实时视频传输的应用研究

发布时间：2018-04-18 06:44

本文选题：视频传输 + 网络拥塞　；参考：《兰州交通大学》2014年硕士论文

【摘要】：视频实时传输过程中存在的问题随着视频点播、远程视频传输等多媒体飞速发展日益凸显，视频会议、视频点播、远程教学、远程监控等多媒体信息实时传输过程中包含大量数据，这使得用户在网络的传输容量和服务质量提出更高的要求。突发数据延迟和包丢失会严重降低客户端图像质量。因此，视频传输对网络带宽，丢包率以及端到端的延迟都有一定的要求，由于网络动态环境的大时滞性、非线性以及参数的时变特性，使得建立精确地数学模型这一问题相对比较困难，本文在国内外视频实时传输研究现状的基础上，进一步深入分析了视频实时传输过程中存在的问题及现有的控制策略，将智能控制理论算法引入到视频传输中，提出一种基于自适应变论域模糊Smith预估PID控制策略。主要研究内容如下： (1)根据流体理论对视频传输系统拥塞控制机制进行建模。 (2)模型的不精确性对视频实时传输影响的研究。引入了模糊PID控制算法对视频传输系统进行控制，，通过仿真实验验证了模糊PID控制策略控制的优越性，系统输出能够快速跟踪稳定值。若突发视频流进入系统，系统都能够维持在期望值附近，然而当参数链路容量有较大变化时，系统输出队列长度跟踪效果较差。 (3)网络环境的大时滞性对视频实时传输影响的研究。采用了一种基于Smith预估的模糊PID实时视频传输控制算法。该算法将模糊控制与Smith预估控制相结合，采取Smith预估策略补偿系统滞后，同时运用模糊控制在一定程度上克服了传统Smith预估器对模型结构与参数的精确性过于敏感、鲁棒性差的缺点，使拥塞控制性能有明显提高。仿真实验结果表明，该算法在大时滞的网络环境下能很好地将路由器队列长度收敛于期望值。 (4)自适应变论域模糊Smith预估控制在视频传输中的应用研究。采用一种自适应变论域模糊Smith预估控制策略，并应用于视频传输的拥塞控制中。最后通过仿真表明，采用变论域模糊控制方法能保证系统的全局稳定性以及控制精度，能使队列快速收敛到目标值，超调量较小，且当网络参数在较大范围内变动时，仍然保持较好的控制性能，并具有较好的稳定性和鲁棒性。
[Abstract]:With the rapid development of video on demand, remote video transmission and other multimedia, video conference, video on demand, distance teaching, etc.In the process of real-time transmission of multimedia information, such as remote monitoring, a large amount of data is included, which makes users require higher transmission capacity and quality of service in the network.Burst data delay and packet loss can seriously reduce client image quality.Therefore, video transmission has certain requirements for network bandwidth, packet loss rate and end-to-end delay, due to the large delay, nonlinear and time-varying characteristics of parameters in the dynamic network environment.It is relatively difficult to establish accurate mathematical model. Based on the current research situation of real-time video transmission at home and abroad, this paper further analyzes the existing problems and existing control strategies in the process of real-time video transmission.The intelligent control theory algorithm is introduced into video transmission, and a fuzzy Smith predictive PID control strategy based on adaptive variable theory is proposed.The main contents of the study are as follows:1) based on fluid theory, the congestion control mechanism of video transmission system is modeled.Research on the effect of imprecision of the model on real-time video transmission.The fuzzy PID control algorithm is introduced to control the video transmission system. The superiority of the fuzzy PID control strategy is verified by simulation experiments. The system output can track the stable value quickly.If the burst video stream enters the system, the system can be maintained near the expected value. However, when the parameter link capacity changes greatly, the system output queue length tracking performance is poor.3) the influence of large time delay on real-time video transmission.A fuzzy PID real-time video transmission control algorithm based on Smith prediction is adopted.The algorithm combines fuzzy control with Smith predictive control, adopts Smith prediction strategy to compensate for system lag, and uses fuzzy control to overcome the sensitivity of traditional Smith predictor to the accuracy of model structure and parameters to a certain extent.Because of the poor robustness, the congestion control performance is improved obviously.Simulation results show that the proposed algorithm can converge the router queue length to the expected value well in the network environment with large delay.Application of adaptive variable domain fuzzy Smith predictive control in video transmission.An adaptive variable domain fuzzy Smith predictive control strategy is adopted and applied to the congestion control of video transmission.Finally, the simulation results show that the variable domain fuzzy control method can guarantee the global stability and control accuracy of the system, make the queue converge quickly to the target value, reduce the overshoot, and when the network parameters change in a large range,It still maintains good control performance, and has good stability and robustness.
【学位授予单位】：兰州交通大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：TP393.06;TP273.4

【参考文献】