直升机攸关动部件损伤预测及其性能测度评估研究

Helicopter (or rotorcraft)is one class of low-space vehicles, which serves as an essential part of equipment for requirement of national safety and economic development, must be used widely in army-civilian applications. Owing to the power transmission system of helicopter with single critical load paths without duplication or redundancy, the evolution and fast degradation of damage of dynamic mechanical components would cause mission interruption or even accidents.As such, dynamic mechanical component monitoring, diagnosis and prognosis is important to helicopter safety and reliability. Based on monitoring and diagnosis technique accumulation of dynamic mechanical components, this proposal focuses on novel theory and method about damage evolution prognosis,failure prognsis and performance evaluation of prognostic techniques for dynamic mechanical components. The main contents include: (1)studying damage evolution model, degradation identification method and remaining useful life prediction method under multiple variational working conditions of dynamic mechanical components by using a nonlinear dynamical systems approach. (2)studying hybrid combined prognostic models and algorithms based on data-driven prognostic methods. (3)studying metric set for evaluating performance of prognostic techniques and corresponding evaluation approach, then establishing a common ground to compare different developed prognostic techniques, so as to provide verification and validation approach for maturation and integrated application of prognostic techniques in any rotorcraft health management system. The above research work and related research production will provide credible prognostic models and algorithms for helicopter technical support and integrated logistics.

直升机是低空飞行器族中与国家安全需求、国民经济发展密切相关的装备，必将在军民兼用领域大量应用。直升机因其单通道动力传动结构与多变的服役环境，其中的动部件（动力传动和旋翼等）诱发的损伤、演化甚或快速恶化会引起任务中断甚至灾难性后果。动部件健康状态的监测、诊断与预测是直升机安全可靠飞行的重要技术保障。本项目拟在动部件监测与诊断技术积累的基础上，聚焦于研究动部件故障预测以及预测性能评估的理论与方法，创新性地以非线性动力学系统理论为工具研究适合于动部件多变工况条件下的损伤演化模型、退化状态识别与剩余寿命预测方法，进而研究数据驱动类预测方法混合的预测模型与算法。以此为基础，进一步研究预测算法的性能评估测度集及评估方法，构建动部件损伤实时预测和综合分析预测的模型库以及测度评估体系，为动部件损伤预测方法的熟化与集成应用提供验证与确认的技术途径，为直升机维护保障体系提供可信的预测模型与算法的技术支持。

直升机攸关动部件健康状态的监测、诊断与预测是直升机安全可靠飞行的重要技术保障。本项目在动部件监测与诊断技术积累的基础上，面向直升机传动系统进行了动部件退化状态识别及剩余使用寿命预测的特征提取与评估方法研究，基于直升机传动系统典型动部件进行了运行状态退化识别与故障预测方法研究，同时，对直升机关键动部件进行了剩余使用寿命预测理论、模型与方法研究，以此为基础，进一步研究了预测算法的性能评估测度集及评估方法，构建了动部件损伤实时预测和综合分析预测的模型库以及测度评估体系，为直升机动部件损伤预测方法的熟化与集成应用提供验证与确认的技术途径，能够为直升机维护保障体系提供可信的预测模型与算法的技术支持。