| With the rapid development of computer technology,new requirements have been put forward to the traditional manufacturing industry,and the "working mother machine" of the manufacturing industry-CNC machine tools represents the core competitiveness of the national manufacturing industry.For a long time,China’s traditional CNC system still uses a closed architecture and single-core processor,which is difficult to adapt to the increasingly complex manufacturing process.Therefore,the intelligentization,openness of CNC systems and the transition from single-core to multi-core systems have naturally become the top priority.In this paper,the symmetrical multi-core ARM development board is used as the experimental platform to discuss the task scheduling technology and real-time control key technology in the intelligent process of CNC system.First,using the Linux operating system and symmetric multi-core ARM card computer as the hardware and software platform,the software framework of the multicore CNC system is proposed and implemented,and the scheduling mode and detailed mechanism of tasks in the homogeneous multi-core CNC system are discussed.By comparing the characteristics of tasks with traditional computer tasks,a scheduling scheme suitable for the task execution requirements of multi-core CNC systems was formulated.Secondly,the software and hardware environment suitable for the subject is built,the real-time performance of Linux system is analyzed and a real-time transformation scheme is implemented.Thirdly,based on the research of the RCS library,the detailed implementation of the layered control system and the reference model of CMS and NML technology for modular communication and layered real-time control are discussed,and the performance test of the real-time system constructed in this paper is tested.Finally,in order to ensure the robustness and robustness of the CNC system from single-core to multi-core,the load balancing scheme suitable for multi-core CNC systems is discussed,studied and formulated.A task scheduling strategy based on sliding windows makes task processing between processor cores more stable and the load relatively balanced. |