无线列控系统中轨旁基站布局优化方法研究

Train control systems are not only the guarantee of the operational safety of trains but also provide sufficient support to enable a high-efficiency railway traffic management. In order to help the train control systems to meet the increasing demands on a safer operation, more efficient train scheduling and lower system cost, wireless communication-based train control systems have been widely applied in both mainline railway and urban mass rail transit. As a result, in this proposal, based on the combinatorial optimization theory, and by making use of the modern telecommunication technology, artificial neural network and recursive filtering in modern control theory, we aim to propose a novel wayside base station deployment optimization method: Building the wayside base station deployment optimization objective functions, extracting the wireless channel joint parameters and measuring the channel characteristic changes, screening the feasible deployments by using neural network. The main scientific problems are: Establishing the coupling relations between different wireless network performance indexes, collecting the joint parameters in the wireless channel, real-time estimation on the slowly-varying parameters of the neural network-based channel model.

轨道交通列车控制系统，不仅是列车安全运行的保证，也为高效的运营管理提供支持。为使其满足对运营安全、运行效率和降低成本等性能需求的日益提高，基于无线通讯的列控系统已在高速铁路系统与城市公共轨道交通系统中被广泛应用。作为列控系统中车地无线通讯的重要组成部分，轨旁无线基站的合理布局，直接关系到列控系统能否最大限度地实现所需求的各类性能指标。因此，本项目以组合优化理论为基础，针对轨道交通典型场景环境下的轨旁无线基站布局优化问题，综合利用无线通讯技术、人工神经网络技术和现代控住理论中的递归滤波技术，建立一种新的轨旁无线基站布局优化方法：建立轨旁无线基站布局优化问题的目标函数，无线信道联合参数提取与特征变化度量，基于神经网络的可行布局筛选，试验验证与性能评估。重点解决所遇到的关键科学问题：无线网络性能指标耦合关系建立，信道特征联合参数提取，神经网络信道中缓变参数实时估计。

轨道交通列车控制系统，不仅是列车安全运行的保证，也为高效的运营管理提供支持。为使其满足对运营安全、运行效率和降低成本等性能需求的日益提高，基于无线通讯的列控系统已在高速铁路系统与城市公共轨道交通系统中被广泛应用。本项目通过研究基于最大信息系数的两级特征提取方法，提出了一种精确度高且可以大大减少相似样本数量的特征提取方法；通过研究一种基于深度神经网络的无线信道建模方法，可以对无线信号传输过程进行快速准确建模；通过研究轨旁无线基站布局优化问题，提出可运用基于分解的多目标进化算法对基站布局问题进行准确求解。本项目的研究成果可以较好解决轨旁基站的布局优化问题。