IPC분류정보
국가/구분 |
United States(US) Patent
등록
|
국제특허분류(IPC7판) |
|
출원번호 |
US-0449372
(2003-05-30)
|
발명자
/ 주소 |
|
출원인 / 주소 |
|
대리인 / 주소 |
Lanier Ford Shaver & Payne P.C.
|
인용정보 |
피인용 횟수 :
12 인용 특허 :
5 |
초록
▼
In one embodiment, a method for controlling an aircraft comprises providing an attitude error as a first input into a neural controller and an attitude rate as a second input into the neural controller. The attitude error is calculated from a commanded attitude and a current measured attitude, and t
In one embodiment, a method for controlling an aircraft comprises providing an attitude error as a first input into a neural controller and an attitude rate as a second input into the neural controller. The attitude error is calculated from a commanded attitude and a current measured attitude, and the attitude rate is derived from the current measured attitude. The method also comprises processing the first input and the second input to generate a commanded servo actuator rate as an output of the neural controller. The method further comprises generating a commanded actuator position from the commanded servo actuator rate and a current servo position, and inputting the commanded actuator position to a servo motor configured to drive an attitude actuator to the commanded actuator position. The neural controller is developed from a neural network, wherein the neural network is designed without using conventional control laws, and the neural network is trained to eliminate the attitude error.
대표청구항
▼
1. A method for controlling an aircraft comprising: providing an attitude error as a first input into a neural controller, the attitude error calculated from a commanded attitude and a current measured attitude; providing an attitude rate as a second input into a neural controller, the attitude
1. A method for controlling an aircraft comprising: providing an attitude error as a first input into a neural controller, the attitude error calculated from a commanded attitude and a current measured attitude; providing an attitude rate as a second input into a neural controller, the attitude rate derived from the current measured attitude; processing the first input and the second input to generate a commanded servo actuator rate as an output of the neural controller; generating a commanded actuator position from the commanded servo actuator rate and a current servo position; and inputting the commanded actuator position to a servo motor configured to drive an attitude actuator to the commanded actuator position; wherein, the neural controller is developed from a neural network, the neural network designed without using conventional control laws, the neural network trained to eliminate the attitude error. 2. The method of claim 1, wherein the commanded attitude, current measured attitude, attitude error, and attitude rate is a commanded roll attitude, a current measured roll attitude, a roll attitude error, and a roll attitude rate, respectively.3. The method of claim 1, wherein the commanded attitude, current measured attitude, attitude error, and attitude rate is a commanded pitch attitude, a current measured pitch attitude, a pitch attitude error, and a pitch attitude rate, respectively.4. The method of claim 1, wherein the commanded attitude, current measured attitude, attitude error, and attitude rate is a commanded yaw attitude, a current measured yaw attitude, a yaw attitude error, and a yaw attitude rate, respectively.5. The method of claim 1, wherein the current measured attitude is provided by an attitude sensor on-board the aircraft.6. The method of claim 1, wherein the aircraft is a remote controlled aircraft.7. The method of claim 1, wherein the aircraft is a fixed-wing aircraft.8. The method of claim 1, wherein the aircraft is a rotary-winged aircraft.9. The method of claim 1, wherein the neural network is trained to eliminate the attitude error via a method comprising: providing an open-loop stimulus to the aircraft, the open-loop stimulus causes the aircraft to oscillate about a free axis; capturing data indicative of the aircraft's response to the open-loop stimulus; selecting a training region from the captured data; and using the training region to train the neural network to eliminate the attitude error. 10. The method of claim 9, wherein the open-loop stimulus is an exponentially decaying sinusoidal waveform.11. The method of claim 9, wherein the open-loop stimulus is provided by an operator.12. The method of claim 9 further comprising mounting the aircraft on a test stand.13. The method of claim 9 further comprising providing a computer coupled to the aircraft, the computer operable to capture data indicative of the aircraft's response to the open-loop stimulus.14. The method of claim 9, wherein the training region starts substantially at a beginning of a sinusoidal waveform and ends substantially at a point where an attitude and a commanded servo profile have very low rates.15. The method of claim 9, wherein the training region comprises at least two regions of overshoot.16. The method of claim 9 further comprising tuning the neural controller by providing a first performance-shaping constant as a third input to the neural controller and a second performance-shaping constant as a fourth input to the neural controller, the first performance-shaping constant determined from an upper performance-shaping line and the second performance-shaping constant determined from a lower performance-shaping line, wherein the upper and lower performance-shaping lines envelope the training region.17. The method of claim 1 further comprising: calculating an attitude error input bias; and adding the attitude error input bias to the attitude error; wherein adding the attitude error inpu t bias to the attitude error causes the output of the neural controller to converge to zero when the attitude error is zero. 18. The method of claim 17, wherein the attitude error input bias is calculated using the Newton-Raphson bisection method.19. An apparatus for controlling an aircraft comprising: an attitude sensor operable to provide a current attitude; a differentiator operable to receive as input the current attitude and derive an attitude rate; a neural controller operable to receive a plurality of inputs comprising an attitude error and the attitude rate, the attitude error calculated from a commanded attitude and the current attitude, the neural controller also operable to generate a commanded servo rate from the plurality of inputs, the commanded servo rate applied to a current actuator position to generate a commanded actuator position; and a servo motor operable receive the commanded actuator position, the servo motor further operable to drive an attitude actuator to the commanded actuator position; wherein the neural controller is developed from a neural network designed without using conventional control laws. 20. The apparatus of claim 19, wherein the current attitude is a roll attitude.21. The apparatus of claim 19, wherein the current attitude is a pitch attitude.22. The apparatus of claim 19, wherein the current attitude is a yaw attitude.23. The apparatus of claim 19, wherein the neural network is trained via a method comprising: mounting the aircraft on a test stand; providing an open-loop stimulus to the aircraft, the open-loop stimulus causes the aircraft to oscillate about a free axis; capturing data indicative of the aircraft's response to the open-loop stimulus; selecting a training region from the captured data; and using the training region to train the neural network to eliminate the attitude error. 24. The apparatus of claim 23, wherein the open-loop stimulus is an exponentially decaying sinusoidal waveform.25. The apparatus of claim 23, wherein the training region starts substantially at a beginning of a sinusoidal waveform and ends substantially at a point where an attitude and a commanded servo profile have very low rates.26. The apparatus of claim 19, wherein the plurality of inputs to the neural controller further comprise a first constant and a second constant, wherein the first constant and the second constant affect an oscillatory behavior of the aircraft and are used to tune the neural controller.27. The apparatus of claim 26, wherein the first constant is a value on an upper performance-shaping line and the second constant is a value on a lower performance-shaping line, the upper and lower performance-shaping lines envelope an exponentially decaying sinusoidal waveform used to train the neural network.28. The apparatus of claim 19 further comprising an attitude error input bias, wherein the neural controller is further operable to receive as input a sum of the attitude error input bias and the attitude error to generate the commanded servo rate.
※ AI-Helper는 부적절한 답변을 할 수 있습니다.