Study On The Control Of Spacecraft Attitude Based On BP Neural Network

With the development of the astronautics technology unceasing, the spacecraft obtained the more and more widespread application in the correspondence, meteorology, resource prospecting, environmental monitoring, navigation localization, scientific research, military and so on many domains. The spacecraft attitude control system is in the spacecraft control important constituent, after whether is it relating the injection spacecraft to the inertia is or other reference systems, the attraction centrosome maintains by certain precision at the predetermined position or the direction. Therefore, correct, real-time controls the spacecraft attitude to have the vital practical significance.The spacecraft attitude control system is multivariable, non-linearity, time-variable and complex system.Unifies the traditional control and the neural network technology, directly in the non-linear model foundation, designs the controller using the computer online optimized operation, enhances the control system the anti-jamming, improves the spacecraft the flight quality.This article discussed the superiority which the neural network technology development present situation and uses in controlling, has analyzed the spatial spacecraft attitude control research present situation.In view of the spacecraft attitude angle equation, used the neural network forecast model to obtain the spacecraft attitude angle forecast output, the use had forecast the control the non-linear optimizer and the expectation output carried on the comparison, obtained the optimization control sequence function in the spacecraft attitude, obtained the attitude angle forecast output.This article first discussed the spacecraft attitude forecast control issue, uses the BP algorithm in the non-linearity forecast model, obtained the system forecast output, in view of the expectation output path, the use non-linearity optimization method minimizing two performance indices, obtained the optimization control sequence.Then has realized using the spacecraft attitude angle equation group to the attitude forecast.Next, has studied the spacecraft attitude control question, adjusts the spacecraft using the BP algorithm the angular speed in the coordinate system projection, and aims at the BP network convergence rate slow characteristic, has introduced the inertia factor. In trains the spacecraft the angular speed in the coordinate system projection and in the establishment spacecraft attitude angle forecast model, has used the BP algorithm separately, thus enabled the spacecraft the attitude to achieve the ideal control effect, has realized the spacecraft operation quality.In order to confirm the above conclusion, this article in view of the spacecraft single axle to the place direction detection roll angle attitude mobile MATLAB simulation, the simulation result has indicated this method accuracy and validity.