中文摘要：

为了提高基于GNSS的深组合列车定位系统的故障容错性能，本文结合模糊系统与神经网络的基本原理，提出基于FRBFNN的GPS／SINS深耦合列车组合定位系统故障诊断算法，建立FRBFNN列车组合定位故障诊断模型，应用免疫差分进化理论设计其相应的网络学习算法。针对列车组合定位故障隔离及系统重构问题，定义组合定位系统结构状态迁移规则，设计深耦合列车组合定位系统故障容错控制处理流程，使深组合列车定位系统的容错控制性能得到进一步提高。以青藏线某区间列车实际运行数据为基础，通过计算机仿真验证了所提模型及算法的有效性与实用性。

英文摘要：

With the aim to improve the fault‐tolerant performance of GNSS based deeply integrated train positio‐ning system ,a FRBFNN based fault detection algorithm was proposed referring to the basic principles of fuzzy system and neural network .The FRBFNN based fault detection model for the integrated train positioning sys‐tem was established ,while the corresponding training algorithm was designed based on the thought of immuni‐ty differential evolution .With respect to the issues of fault isolation of integrated train positioning and system reconfiguration ,the rules for state transition of integrated system structure were defined ,and the fault‐tolerant control process was also designed aiming at increasing the fault tolerant control performance of the integrated train positioning system .The effectiveness and practicality of the proposed model and solution were verified by computer simulation ,using the real operation data of a train in a section of the Qinghai‐Tibet railway .