基于回归模型的液体火箭发动机试验数据的验证与重构

Test Data Validation and Reconstruction Based on Regression Model for Liquid Propellant Rocket Engine Ground Test

由于传感器设计、制造水平和工作环境等原因,传感器工作异常、出现故障或失效现象时有发生,影响了试验的完成和试验数据的使用。在研究液体火箭发动机地面试验各测点参数相关性的基础上,建立了传感器验证网络,利用测点参数间的线性回归模型对传感器数据进行验证和重构。经过实例验证表明,利用线性回归模型建立传感器验证网络,验证与重构发动机试验数据是可行的、有效的。

The phenomena of sensors being abnormity, trouble and failure often happen because of design, manufacture ability and operation environment of sensors, which influence the test accomplishing and the test data application. On the basis of studying the relativity of measurement points parameters of liquid rocket engine ground test, to provide the reliable and authentic test data for the rocket engine designed and employed sections, the authors put forward establishing the sensor validation network to validate and reconstruct the test data with linear regression models of measurement points parameters. The examples demonstrate that it is feasible and effective.