基于双核的嵌入式数控系统速度前瞻控制研究

发布时间：2018-08-06 14:16

【摘要】：数控系统是数控机床实现高速高精加工控制的核心,其性能直接决定了机床性能。以ARM+DSP双核处理器构建的嵌入式数控系统,能满足中高档数控系统在高性能低成本方面的需求。速度前瞻控制能在插补之前对加工路径进行计算和规划,实现机床的高效平稳运行。通过将嵌入式技术与前瞻控制技术相结合,可缩小国内外在数控硬件技术和控制软件上的差距,从而为我国装备制造业的发展带来新机遇。本文在基于OMAP3530的双核数控系统平台上,分别开展了对微小线段轨迹和NURBS曲线轨迹的前瞻插补算法研究,并将此算法进行了仿真和试验验证。本文开展的主要研究工作如下:(1)加减速技术是速度前瞻控制算法中的重要组成部分。传统S型加减规划中其分段表达式比较繁琐,而且速度曲线类型众多,为了便于计算,本文研究了一种简化的以应用到后续前瞻控制算法中的S型加减速规划方法。(2)由于传统小线段插补“段内启停”会导致加工效率低下,本文基于一种复合的拐角过渡模型,对小线段衔接处的前瞻速度约束进行分析,确定了线段衔接处的最优速度,并根据该速度大小自适应选取前瞻段数,从而有效减少了前瞻计算时间。(3)为了实现曲线加工过程中精度和速度的协调控制,本文提出了一种基于前瞻控制的自适应NURBS插补算法:先根据预处理阶段的信息找出曲线上的速度突变点和突变点处进给速度,然后对突变点处速度重新规划以防止速度跃变,最后利用NURBS曲线的对称性和双向插补思想来预测减速点。通过速度规划,在实时插补阶段结合Muller法和Newton迭代法来求取下一周期的插补参数,避免NURBS曲线导数的计算。仿真结果表明,本文的NURBS插补方法能满足机床的柔性要求,减小了进给速度波动。(4)针对双核数控系统平台设计了双核通信程序和前瞻控制算法模块,并实现了嵌入式数控系统的功能软件模块。通过NURBS实时插补算法执行时间测试验证了该算法的实时性,同时速度前瞻插补算法仿真和试验结果也验证了该算法的有效性。
[Abstract]:Numerical control system is the core of NC machine tool to realize high speed and high precision machining control, and its performance directly determines the performance of the machine tool. The embedded NC system based on ARM DSP dual core processor can meet the demand of high performance and low cost for middle and high grade NC system. The speed forward control can calculate and plan the machining path before interpolation, and realize the machine tool running efficiently and smoothly. By combining embedded technology with prospective control technology, the gap between CNC hardware technology and control software at home and abroad can be narrowed, thus bringing new opportunities for the development of China's equipment manufacturing industry. On the platform of dual-core NC system based on OMAP3530, this paper studies the forward interpolation algorithm of tiny line segment trajectory and NURBS curve trajectory, and simulates and verifies the algorithm. The main work of this paper is as follows: (1) acceleration and deceleration technology is an important part of speed forward control algorithm. In the traditional S-type addition and subtraction programming, the piecewise expressions are tedious, and there are many types of velocity curves. In this paper, a simplified S-type acceleration and deceleration planning method applied to the follow-up forward control algorithm is studied. (2) because the traditional small line segment interpolation "start and stop within the segment" will lead to low processing efficiency, this paper is based on a compound corner transition model. Based on the analysis of the forward speed constraint at the junction of small segment, the optimal speed at the junction of line segment is determined, and the number of prospective segments is adaptively selected according to the size of the speed. In order to realize the coordination control of precision and speed in the process of curve processing, it can effectively reduce the time of prospective calculation. (3) in order to realize the coordinated control of precision and speed in curve processing, In this paper, an adaptive NURBS interpolation algorithm based on forward control is proposed. Firstly, the velocity abrupt point and feed velocity on the curve are found according to the information of the preprocessing stage, and then the velocity at the abrupt point is reprogrammed to prevent the velocity jump. Finally, the symmetry of NURBS curve and the idea of bidirectional interpolation are used to predict the deceleration point. Through velocity planning, the interpolation parameters of the next cycle are obtained by combining the Muller method and Newton iteration method in the real-time interpolation stage, thus avoiding the calculation of the derivative of the NURBS curve. The simulation results show that the proposed NURBS interpolation method can meet the flexible requirements of machine tools and reduce the fluctuation of feed speed. (4) Dual-core communication program and forward control algorithm module are designed for dual-core CNC system platform. The function software module of embedded NC system is realized. The real-time performance of the algorithm is verified by the execution time test of the NURBS real-time interpolation algorithm. The simulation and experimental results of the speed forward interpolation algorithm also verify the effectiveness of the algorithm.
【学位授予单位】：南京航空航天大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TG659

【参考文献】