【作者】 杨敏；

【导师】 王祁；

【作者基本信息】 哈尔滨工业大学 ， 仪器科学与技术， 2011， 硕士

【摘要】 我国航天事业的发展正处于需求不断增加,技术也不断进步的时期,而火箭技术是关系航天发射成败的关键。火箭最重要的组成部分,火箭系统的心脏火箭发动机,成为制约火箭技术发展的瓶颈因素。由于每一次的航天发射,都要消耗大量的人力、物力和财力,且其发射具有不可重复性,如果出现发射失败的情况,将造成巨大的灾难性后果和不可挽回的经济损失。为确保火箭工作的可靠性、有效性与经济性,在研制火箭发动机的过程中,需要进行大量的地面试验。而火箭发动机试验台系统具有规模庞大、结构复杂、零件种类多以及影响因素多的特点,实验过程中涉及到传感器、通信系统、遥测系统和控制系统等诸多方面的技术,因此,采用各种测试控制系统,使地面试车台能始终工作在正常稳定的状态成为研制火箭发动机的关键问题。为了对液体火箭发动机试验台的运行情况有一个直观的认识,必须要知道系统当前的运行状态,即系统是正常运行,还是发生轻微故障或出现强故障。为了根据原始数据确定液体火箭发动机试验台发生故障的形式、位置和特点,确定其运行状态。本文提出多种算法对传感器、试验台系统的健康状态进行评估。在本文中,主要研究了灰色综合评价方法、信息熵方法与基于模糊综合评判方法的健康度计算方法在健康评估方面的应用,并采用实际的传感器数据对方法进行了验证。灰色综合评价方法的主要思想是通过分析数据之间的关联信息,即通过灰色关联度的计算,判断数据之间的相似性与差异性。而通过计算与最优指标之间的相似程度可以评估系统与最优状态的关联信息,从而评估系统健康状态。信息熵方法将化学中的热力学熵的概念引入到信息论中,根据信息变化的剧烈程度来判断系统的运行情况。信息熵值越大,说明信息变化越复杂,系统运行越不稳定;而信息熵越小时,信息变化越小,系统越趋向于正常运行。健康度计算方法的主要思想是根据历史数据的变化特征,确定合适的模糊隶属度函数,计算当前系统的隶属度,并与相应的权重结合,得到系统的整体健康度值,来评判系统的运行状态。通过分析比较三种方法数据仿真的效果,得到了三种方法各自的特点,并验证了其可用性与有效性。从而选择合适的某一种方法,或者某几种方法的组合完成对试验台系统的健康评估。更多还原

【Abstract】 With increasing demand of aerospace industry and the development of technology, rocket technology is of great importance in space launch. The most important part of the rocket is rocket engine which limits the advance of rocket technology. Due to the vast consumption of human resources, material resources, financial recourses of the launch, a launch failure will result in catastrophic consequence and huge economic losses. To ensure the reliability, efficiency, economy in the process of developing rocket engines, extensive ground tests should be carried out. For the large scale, complex structure, part variety, multiple influential factors of the rocket engine test platform, it includes sensors, communication systems, telemetry systems, control systems etc. The stabilization and reliability of the rocket test platform is the key point in the development of the rocket.In order to obtain an intuitive understanding of the rocket test platform, the status of platform could be categorized in normal operation, slight fault and major failure based on the type, location, characteristic of the fault and the impact on the health. In this paper, the research provides a variety of algorithms to evaluate the status of sensors and test platform. The research focuses on the application of the gray comprehensive evaluation method, the information entropy, the fuzzy comprehensive evaluation in health assessment. The effectiveness is verified by experimental data.The gray comprehensive evaluation is utilized to analyze the relationship information in the data, i.e., by calculating gray relational degree to determine the similarity and diversity of data. By calculated the similarity between the obtained data and optimal index, the health status of the platform can be evaluated.The information entropy introduces the concept of entropy in the thermodynamic chemistry to information theory. The operation status can be obtained by variation intensity of the information. The larger the information entropy, indicating more complex information changes, the more unstable the system is; and the smaller the information entropy, indicating less change of information, the more the system tends to function properly.The main idea of the health degree method is: First, the appropriate fuzzy membership function is determined based on historical data variation. Then, the membership of current system is calculated. Finally, with the corresponding weight, the overall health degree is obtained to to evaluate the status of the system.By analyzing the numerical simulation results of three methods, the characteristics of the methods are obtained. The efficiency and effectiveness are validated. To select the appropriate one of the methods, or a combination of the methods can evaluate the health status of the test platform.更多还原