Research On Key Technologies Of Real-time Simulation Platform Based On Kylin Operating System

The field of control systems such as aeronautics and astronautics has great demand for autonomous controllable real-time simulation platform.However,the Windows platform and Intel CPU are widely used in the bottom of the real-time simulation platform developed in China,which may bring about hidden dangers in information security.Kirin(Kylin)operating system and soar series chip have been adopted by more and more government and defense units as the result of the localization of the underlying hardware and software system.Therefore,the research on the real-time simulation platform based on Kylin operating system and FT CPU is of great theoretical significance and practical value to meet the needs of the national defense industry for the autonomous controllable simulation platform.According to the application requirements of self controlled real-time simulation platform,based on the existing research results and related analysis of Kylin operating system and the characteristics of soar CPU,the research of some key problems such as the real-time simulation platform framework,scalable memory management algorithm and graphic modeling method is carried out.The main tasks and innovations include:1.Real-time simulation platform is to ensure the real-time performance of the simulation.To make it,the first task is to build a real-time simulation framework which can support the task collaboration and real-time ensurance by using the services provided by the operating system.Based on the in-depth analysis of the functional characteristics of Kylin,a real-time simulation framework based on real-time process and non real-time process coordination scheduling is designed.The real-time process completes the task of model calculation,parameter transfer and other real-time requirements,and the non real time process completes the data display and storage task.By calling the system function provided by the Kylin system,the real-time process is automatically executed with the user provided parameters in real-time priority after the simulation starts.The processes realize the process synchronization and data transmission through the IPC mechanism,so as to ensure the real-time simulation.2.Real time simulation generates a large amount of data.In order to ensure the real-time performance of the simulation,these data are usually stored in memory,and then written to the hard disk after the simulation is finished.When the simulation time is long and the amount of data is large,it will lead to excessive memory consumption and affect the performance of the system.To solve this problem,we design a data storage management method based on shared memory,cache memory and real-time writing disk.The real time process writes the generated data to the shared storage area,and the non real time process reads the data from the shared storage area and writes it to the cyclic cache.This method can greatly reduce the consumption of memory and ensure the real-time performance of the system by the use of shared memory and cyclic cache.The test results show that,when the frame time jitter is kept below 15?s,the memory usage almost does not change with the simulation time.3.Graphical modeling method can greatly shorten the model development process and improve the efficiency of simulation.The Simulink/RTW tool can convert the graphical model created by the user into the corresponding C/C++ code,but the code lacks real time control and cannot be applied directly.Therefore,this paper designed a real-time reconstruction method based on template code.This method realizes the high precision real time simulation on Kylin system by adding the initialization,step and termination function of the model in the frame time control code.The test results show that the model can run correctly and output the result when the frame time jitter is less than 15?s.Based on the above results,a prototype system of real-time simulation platform under Kylin system is designed and implemented.The self built model test shows that the system can complete the whole process from the graphical model creation to the simulation results output when the frame jitter is less than 15?s.