Chun-Hua Zhou;Bang-Feng Wang
. 2012, 43(08): 1079-1083.
Abstract (
)
Download PDF (
)
Knowledge map
Save
In the research of variant skinned wing, a type of SMARC rippled skin structure with active deformation was proposed at first for meeting requirements of continuous smooth and large deformation. And then, two kinds of skin testing samples were made and tested for stretch and driving ability in different environment temperature. Results of tensile test show that SMARC rippled skin test sample has better stretching deformation performance than the ordinary rippled skin type. However, the temperature has more influence on SMARC rippled skin test samples. When the temperature is higher, structure stiffness of SMARC rippled skin test sample is smaller. In the driving test, the initial displacement of SMARC rippled skin test sample has linear increase. When the drive electric current is more than 3.4A, matrices of composite material are heated to soften. As the result of that, the displacement has a nonlinear variation trend. At last, on the basis of driving test data, a type of RBF neural network model was established where current strength was input parameter and the driver displacement was output parameter. The approximation curve of this neural network model is more accurate and close to the practical characteristics, and the biggest driver relative prediction error is less than 6%, which provides for active control of intelligent skin with a certain theoretical and experimental basis.