A Research On Multi-sources Simulation Of UAV Flight Control System

With the development of UAV technology, a UAV development team is involved in more and more subjects. To complete the overall simulation of a UAV system, we need to complete interactive simulation of their responsibility subsystem models between various disciplines subsystems of UAV flight control system. In order to meet the UAV flight control system development, we need to combine the simulation models, generated by a plurality of Simulation tools, together by input and output data exchange. Based on FMI(Function mock-up interface) standard, we carried out the study of technologies related to multi-sources simulation of UAV flight control system. We studied the model input data extrapolation algorithm, multi-sources model simulation error estimation and error control algorithms, as well as multi-sources model simulation scheduling method. And on this basis we developed the UAV flight control system simulation platform and multi-sources flight control system simulation data analysis software.Based on the basic theory of UAV Flight Control Simulation, we analyzed the technical requirements for Multi-sources Simulation of UAV Flight Control System. After analyzed the futures of the existing interface to develop the co-simulation interface, HLA standard and FMI standard, we selected the third one.We analyzed multi-sources model simulation data synchronization features and multi-sources model of parallel simulation features. Based on minimum mean square error of polynomial extrapolation, we proposed multi-sources simulation data extrapolation algorithm based on changes of the previous data, which makes separation model is completed concurrent run feasible. However, this method does not guarantee parallel simulation errors will be limited within a certain range.In order to ensure the efficiency and the accuracy of the simulation, we need to carry the simulation parallel, and control the global error not exceeds a boundary. Based on analysis of the structure of multi-sources model simulation, we get the conclusion that the global errors caused by local data extrapolation will be bounded. Based on Richard Johnson extrapolation method we get a method which can improve the accuracy of the error estimation for multi-sources model simulation in parallel. Based on the error estimation method, we limited the error by adjusting the communication intervals.The scheduling method of the multi-sources UAV flight control system is divided in to three levels. System simulation scheduling will be parallel, local scheduling will be serial, a single model will be scheduled as the state transition.Finally, based on the analysis of the requirements of multi-sources simulation of UAV flight control system, we developed the platform of multi-source UAV flight control system simulation and the software of data analysis for flight control simulation. The platform supports Monte Carlo parametric simulation, and the analysis of large amounts of simulation data.