Limits...
Aircraft aerodynamic parameter detection using micro hot-film flow sensor array and BP neural network identification.

Que R, Zhu R - Sensors (Basel) (2012)

Bottom Line: Air speed, angle of sideslip and angle of attack are fundamental aerodynamic parameters for controlling most aircraft.For small aircraft for which conventional detecting devices are too bulky and heavy to be utilized, a novel and practical methodology by which the aerodynamic parameters are inferred using a micro hot-film flow sensor array mounted on the surface of the wing is proposed.A back-propagation neural network is used to model the coupling relationship between readings of the sensor array and aerodynamic parameters.

View Article: PubMed Central - PubMed

Affiliation: State Key Laboratory of Precision Measurement Technology and Instruments, Department of Precision Instruments and Mechanology, Tsinghua University, Beijing 100084, China. katykob@163.com

ABSTRACT
Air speed, angle of sideslip and angle of attack are fundamental aerodynamic parameters for controlling most aircraft. For small aircraft for which conventional detecting devices are too bulky and heavy to be utilized, a novel and practical methodology by which the aerodynamic parameters are inferred using a micro hot-film flow sensor array mounted on the surface of the wing is proposed. A back-propagation neural network is used to model the coupling relationship between readings of the sensor array and aerodynamic parameters. Two different sensor arrangements are tested in wind tunnel experiments and dependence of the system performance on the sensor arrangement is analyzed.

No MeSH data available.


Related in: MedlinePlus

Structure of the neural network.
© Copyright Policy
Related In: Results  -  Collection

License
getmorefigures.php?uid=PMC3472866&req=5

f6-sensors-12-10920: Structure of the neural network.

Mentions: Intuitively, the relationship between the sensors' readings and the three flight parameters is a multiple input multiple output (MIMO) coupling system. As a matter of fact, f is generally a complex multivariate and nonlinear function, which plays a crucial role in the detection approach. An approximation of f can be obtained by various means, such as an analytical method using intrinsic fluid dynamic models, simulation, experiment-based identification or a mixture of the above. Theoretically, the analytical method is too complicated and inefficient (different airfoil results in different f); simulation is feasible in some sense, but is generally less valid; in contrast, the experimental identification is more reliable and effective. In this paper, we adopted experiment-based model identification method to develop f, by using a 3-layers BP neural network shown in Figure 6 as the model structure considering that it had been theoretically proved three layers of neural network could solve arbitrarily complicated nonlinear mapping problems [18]. For the application on small UAVs and MAVs, which requires both accuracy and simplicity, a 3-layer BP neural network with 9 neurons in the hidden layer was used, where the number of hidden neurons was determined through experimental testing. In Figure 6, normalization function fin, denormalization function fout, sigmoid function of the hidden layer fhid and transfer matrix wih, who constitute the model structure of the function f.


Aircraft aerodynamic parameter detection using micro hot-film flow sensor array and BP neural network identification.

Que R, Zhu R - Sensors (Basel) (2012)

Structure of the neural network.
© Copyright Policy
Related In: Results  -  Collection

License
Show All Figures
getmorefigures.php?uid=PMC3472866&req=5

f6-sensors-12-10920: Structure of the neural network.
Mentions: Intuitively, the relationship between the sensors' readings and the three flight parameters is a multiple input multiple output (MIMO) coupling system. As a matter of fact, f is generally a complex multivariate and nonlinear function, which plays a crucial role in the detection approach. An approximation of f can be obtained by various means, such as an analytical method using intrinsic fluid dynamic models, simulation, experiment-based identification or a mixture of the above. Theoretically, the analytical method is too complicated and inefficient (different airfoil results in different f); simulation is feasible in some sense, but is generally less valid; in contrast, the experimental identification is more reliable and effective. In this paper, we adopted experiment-based model identification method to develop f, by using a 3-layers BP neural network shown in Figure 6 as the model structure considering that it had been theoretically proved three layers of neural network could solve arbitrarily complicated nonlinear mapping problems [18]. For the application on small UAVs and MAVs, which requires both accuracy and simplicity, a 3-layer BP neural network with 9 neurons in the hidden layer was used, where the number of hidden neurons was determined through experimental testing. In Figure 6, normalization function fin, denormalization function fout, sigmoid function of the hidden layer fhid and transfer matrix wih, who constitute the model structure of the function f.

Bottom Line: Air speed, angle of sideslip and angle of attack are fundamental aerodynamic parameters for controlling most aircraft.For small aircraft for which conventional detecting devices are too bulky and heavy to be utilized, a novel and practical methodology by which the aerodynamic parameters are inferred using a micro hot-film flow sensor array mounted on the surface of the wing is proposed.A back-propagation neural network is used to model the coupling relationship between readings of the sensor array and aerodynamic parameters.

View Article: PubMed Central - PubMed

Affiliation: State Key Laboratory of Precision Measurement Technology and Instruments, Department of Precision Instruments and Mechanology, Tsinghua University, Beijing 100084, China. katykob@163.com

ABSTRACT
Air speed, angle of sideslip and angle of attack are fundamental aerodynamic parameters for controlling most aircraft. For small aircraft for which conventional detecting devices are too bulky and heavy to be utilized, a novel and practical methodology by which the aerodynamic parameters are inferred using a micro hot-film flow sensor array mounted on the surface of the wing is proposed. A back-propagation neural network is used to model the coupling relationship between readings of the sensor array and aerodynamic parameters. Two different sensor arrangements are tested in wind tunnel experiments and dependence of the system performance on the sensor arrangement is analyzed.

No MeSH data available.


Related in: MedlinePlus