Data Fusion Technology Of AHRS System Simulation

Information fusion is an information processing technology which was developed gradually at the end of1990s and at the beginning of2000s. Its main effect is to perform integrated analysis and processing to the information obtained through any way, at any time and in any space, so as to lay good foundation for decision-making and control. To conduct integrated processing for various kinds of information is a conscious or unconscious behavior of mankind. In order to meet the demands of war, human beings began to study information fusion as a technology. Apart from military field, many other fields also involve research and application in this aspect.Information fusion, also called data fusion, is a critical technology during development of combined navigation system. Data fusion earlier than1960s was usually performed through method of frequency filtering or classic automatic control correction, mainly adopting form of link correction. Since1960s, Kalman filtering was beginning to be used in data fusion technology of combined navigation system gradually. Kalman filtering estimation method based on minimum variance is often called optimal filtering method of combined navigation when it is used in data fusion for combined navigation system. In order to achieve combined navigation, Kalman filtering is usually used first to estimate all the errors of the system (also known as error status), and then the system is corrected through the estimated value of error status.This dissertation performs simulation research on application of data fusion technology and Kalman filtering technique in strap-down heading and attitude reference system, which focuses the discussion on adoption of data fusion technology in order to enhance output accuracy and stability of heading&attitude and to offset drift resulted from low-accuracy inertial measurement unit. Airspeed signal of air data system is introduced as the observed and measured value in combined Kalman filtering, and the estimated error of navigation computation is compensated and corrected through Kalman filtering process, so as to improve attitude output accuracy of the system. Similarly, strap-down magnetic sensor signal is introduced and magnetic heading damping algorithm is adopted to enhance heading output accuracy of the system. As the system in previous stage was not very mature, relatively high oscillation was revealed for output data in true airspeed combined mode, which are thus not applicable; in order not to affect flight mission, and meanwhile to obtain the real flight data of system (synchronized true airspeed signal), GPS combined mode is to be used temporarily after coordination. Ground computer dynamic simulation environment is improved according to flight test data in GPS combined mode, difference between GPS ground speed and true airspeed model is analyzed through comparison by ground simulation, and furthermore, the decomposition model of true airspeed in Kalman filtering application is improved. Output data of the system in the two combined modes are compared through simulation so as to analyze error through comparison, true airspeed model and Kalman filtering parameters for true airspeed are adjusted repeatedly, and finally the feasibility of this scheme is demonstrated.