作 者: (朱美印); (张松); (但志宏); (裴希同); (王信); (王曦);
机构地区: 北京航空航天大学能源与动力工程学院,北京100191 先进航空发动机协同创新中心,北京100191
出 处: 《燃气涡轮试验与研究》 2017年第4期39-44,共6页
摘 要: 大口径蝶阀是高空模拟试车台的一个重要设备,其数学模型建立依赖于大口径蝶阀的流量特性。为获得大口径蝶阀的流量特性,提出了一种基于特征坐标,从大稀疏分散试验数据中提取连续特征信息的大口径蝶阀流量特性的坐标定位回归算法。即先对无规律的试验数据过滤预处理,再基于特征坐标对预处理数据在特征坐标轴上进行压缩、合并。在此基础上提出了迭代扩充回归流量特性系数的算法,从而获得所需的大口径蝶阀的流量特性。与高空模拟试车台的试验数据对比表明,利用建立的大口径蝶阀流量特性所计算的蝶阀流量与测量流量之间的稳态误差均在5%以内。 Large butterfly valve is an important equipment in the new Altitude Test Facilities(ATF).The establishment of the mathematical model of ATF depends on the flow characteristics of the large butterfly valve.In order to acquire the flow characteristics of a large butterfly valve,a coordinate positioning and regression algorithm which extracts continuous feature information from the large sparse experiment data based on the feature coordinate was proposed.First of all,the irregular data was preprocessed by filtering.Then based on the feature coordinates,the preprocessed data was compressed and merged on the feature axis.Finally,an iterative algorithm was proposed to extend the flow coefficient and then the expected flow characteristics of large butterfly valve were acquired.Compared with the experimental data of ATF,the results show that the steady state error between the flow calculated by the method mentioned above and the measurement flow is less than 5%.