DEEP-LEARNING-BASED INTELLIGENT FORCE MEASUREMENT SYSTEM USING IN A SHOCK TUNNEL

Aerodynamic force measurement in high-enthalpy flow is very important for the design and optimization of hypersonic vehicles. Currently, impulse facilities are used for generating high-temperature and high-pressure driving gas to simulate the high-enthalpy flow with hypersonic flight-conditions, such as a shock tunnel. However, when force tests are conducted in an impulse facility, the inertial force has a large influence on the measuring results, which creates low-frequency vibrations of the test model and its motion cannot be addressed through digital filtering. In the case of a few milliseconds of test time, the structural design of the six-component balance is greatly challenged. Therefore, dynamic calibration becomes very important for improving the precision and accuracy of force measurement during short-duration. A new method, deep-learning-based single-vector dynamic self-calibration of the force measurement system, and intelligent force measurement system are proposed for obtaining high-accurate aerodynamic force in impulse facilities. One of the main features of this dynamic calibration method is the calibration of the overall force measurement system, not just the balance. Applying this method, the calibrated force measurement system is the wind tunnel test object, which ensures the consistency of calibration and application. In the evaluation, the test verification has achieved relatively ideal results, the large-scale low-frequency vibration interference has been basically eliminated, and the accuracy and reliability of the force measurement in impulse facility have been greatly improved.