摘要
对于固体火箭发动机羽焰温度的测量是困难的,了解其温度分布对于研究发动机内燃烧流场的特性有重要价值。考虑到固体火箭发动机羽焰高温、高压和高速燃烧流场的测试要求,使用哈尔滨工业大学研制的6目标8波长高温计进行了地面搭载试验,成功地实现了对固体火箭发动机羽焰温度的测量,并提出了多光谱测温法的新数据处理方法,即基于BP神经网络预估计固体火箭发动机羽焰的发射率与波长之间函数关系,再由逐步回归法实现真温的自动识别。实验结果表明,其真温计算值与火箭发动机设计者提供的理论值之差在±100K以内,说明多光谱法是一种测量固体火箭发动机羽焰温度的可行方法。
The temperature of a high te mperature, high pressure plume solid propellant rocket engine is difficult to me asure but is useful for determining the combustion status of the engine. This pa per presents plume temperature measurements of a mounted solid propellant rocket engine measured using a 6-target 8-wavelength pyrometer. The processing of th e multispectral thermometric data was based on the relationship between the emis sivity and the wavelength of the plume as predicted by a BP neural network. The true temperatures were then predicted using auto-regression of known emissivity data. The experimental results showed that the calculated temperatures were wit hin ±100 K of the temperatures provided by the rocket engine designer. T herefore, multispectral thermometry is a practical method for measuring the plum e temperature of solid propellant rocket engines.
出处
《清华大学学报（自然科学版）》
EI
CAS
CSCD
北大核心
2003年第7期916-918,922,共4页
Journal of Tsinghua University(Science and Technology)
基金
国家自然科学基金资助项目(69777020)
