论文部分内容阅读
针对液体火箭发动机试验台故障子样少,故障预测精度低,故障维修保障困难等问题,在分析标准RVM优缺点的基础之上,提出了一种自适应能力较强的故障预测模型——ARVM(Adaptive Relevance Vector Machine)。为测试该模型,以某型轨控发动机高空模拟试验台管路流量、燃烧室压力为输入参量对推力矢量进行了预测,预测结果表明,ARVM方法能够有效跟踪推力矢量参数的变化趋势,并且获得了较高的预测精度和模型稀疏性。该方法对于复杂系统的故障预测和维修保障具有一定的理论价值和工程应用意义。
Aiming at the problems such as less fault sub-sample, low fault prediction accuracy and difficult fault maintenance support for liquid propellant rocket engine test bench, an adaptive failure prediction model based on the advantages and disadvantages of standard RVM is proposed. (Adaptive Relevance Vector Machine). In order to test this model, the thrust vector is predicted by the pipeline flow rate and combustion chamber pressure of a track-mounted engine at high altitude simulation test bench. The prediction results show that the ARVM method can effectively track the trend of thrust vector parameters and obtain A higher prediction accuracy and model sparsity. The method has certain theoretical value and engineering application significance for the fault prediction and maintenance of complex systems.