轨道交通全自动无人驾驶多专业系统耦合模型分析

Coupling Model Analysis of Multi-disciplinary Systems for Fully Automatic Rail Transit

在轨道交通全自动无人驾驶技术迅猛发展的背景下,多专业系统间的耦合协作至关重要。为了在应急、故障等降级场景下提供有效的运营辅助决策,从多系统耦合模型的角度,创新性地对轨道交通多专业系统进行耦合关系分析及模型研究,充分考虑各专业子系统之间的耦合关系及权重变化。首先,分析自动化等级最高的GoA4级下各专业子系统之间的耦合权重指数,采用层次分析法构建耦合判断矩阵并判断了耦合权重的合理性。其次,研究在轨道交通自动化等级降级运营情况下,采用马尔可夫状态空间转移模型对多专业系统的耦合关系和状态转移进行分析与建模。研究发现,随着GoA自动化等级的逐步降级,轨道交通各专业子系统之间的耦合关系逐步降低且权重趋于平均化,证实了GoA自动化等级越高,各专业子系统之间的耦合关系越密切。这为未来轨道交通全自动无人驾驶系统的数字化平行推演神经网络等大模型预测与应急辅助决策提供了理论支撑。

In the context of the rapid development of fully automated unmanned driving technology in rail transit,the coupling and collaboration between multiple specialized systems have become increasingly crucial.To provide effective operational decision support in emergency and fault scenarios under a fully automated unmanned system,The coupling relationships and models of various specialized systems are innovatively analyzed from a multi-system coupling perspective,with the coupling relationships and weight changes among the subsystems being considered.Firstly,the coupling weight indices among subsystems at the highest automation level,GoA4,were analyzed.The Analytic Hierarchy Process was employed to construct a coupling judgment matrix and assess the rationality of the coupling weights.Secondly,in scenarios where the automation level degrades,a Markov state-space transition model was used to analyze and model the coupling relationships and state transitions of these systems.The study found that as the GoA automation level decreases,the coupling relationships among the specialized subsystems gradually weaken and the weights become more balanced.This confirms that higher GoA automation levels result in closer coupling relationships among subsystems.Theoretical support for the prediction and decision-making processes in future fully automated unmanned rail transit systems is provided using digital parallel simulation neural networks and other large models.