模糊神经网络控制的机床进给伺服系统研究

发布时间：2018-02-23 21:14

本文关键词： 进给伺服永磁同步电机模糊神经网络位置环　出处：《辽宁工业大学》2016年硕士论文　论文类型：学位论文

【摘要】：伺服产品在机床行业被广泛应用,所以机床行业的前景依赖于伺服技术的发展。目前,永磁同步电机伺服系统的技术愈发成熟,能够适应高性能的伺服驱动要求,以永磁同步电机作为伺服驱动装置被广泛应用于数控机床进给系统。进给伺服系统需要很高的位置跟踪精度和定位精度,它在数控机床伺服系统中要求最高,对高性能的永磁同步电机伺服系统控制策略的研究,将在很大程度上影响进给系统的性能和控制效果的好坏。基于此,本文提出了模糊神经网络伺服控制方法并用于机床进给伺服系统中。首先,论文详细介绍了机床进给伺服系统的结构组成、分类以及对伺服电机的要求。接着从永磁同步电机的结构出发,得出了旋转坐标系下的数学模型,建立PMSM本体仿真模型,并分析了空间矢量脉宽调制技术的原理与算法。在此基础上,进行了PMSM磁场定向矢量控制仿真。然后深入研究了模糊逻辑和神经网络的控制理论,给出了模糊控制器的设计过程以及神经网络的建立步骤。针对永磁同步电机伺服系统非线性、强耦合、多变量和时变的特性,在永磁同步电机矢量控制的基础上,提出了模糊神经网络控制策略运用在位置环上,它是一种集模糊逻辑推理、自适应、自学习能力一体的混合智能控制器,并给出了模糊神经网络的结构及学习算法,然后利用MATLAB仿真工具进行不同控制策略下的位置环仿真,得出PMSM伺服系统位置环采用模糊神经网络控制器,将具有更好的响应速度和跟随性能。通过理论研究和仿真结果分析后,最后进行了基于DSP的系统软硬件设计,选择采用TI公司生产的TMS320F2812芯片作为整个系统的控制核心。其中硬件设计主要包括主电路、控制电路、位置控制模式设计和按键及显示电路设计;而软件设计包括主程序、中断子程序、SVPWM子程序和功率保护中断子程序。在结尾部分,对通篇文章所做的工作进行了总结,进一步给出了该课题的后续工作安排和对未来的展望。总之,基于模糊神经网络控制的机床进给伺服系统响应迅速、控制精度高、稳定性好,实现了高性能的电机驱动系统。
[Abstract]:Servo products are widely used in the machine tool industry, so the future of the machine tool industry depends on the development of servo technology. At present, the technology of PMSM servo system is more mature and can adapt to the requirement of high performance servo drive. The permanent magnet synchronous motor (PMSM) is widely used in the feed system of NC machine tools, which requires high position tracking accuracy and positioning accuracy, which is the highest requirement in the CNC machine tool servo system. The research on the control strategy of high performance PMSM servo system will greatly affect the performance and control effect of the feed system. In this paper, a fuzzy neural network servo control method is proposed and used in the machine tool feed servo system. Firstly, the structure of the machine tool feed servo system is introduced in detail. From the structure of permanent magnet synchronous motor (PMSM), the mathematical model in rotating coordinate system is obtained, and the simulation model of PMSM ontology is established. The principle and algorithm of space vector pulse width modulation (SVPWM) are analyzed. On this basis, the simulation of PMSM field-oriented vector control is carried out, and the control theory of fuzzy logic and neural network is deeply studied. The design process of fuzzy controller and the steps of establishing neural network are given. In view of the nonlinear, strong coupling, multivariable and time-varying characteristics of PMSM servo system, based on the vector control of PMSM, The fuzzy neural network control strategy is applied to the position loop. It is a hybrid intelligent controller with fuzzy logic reasoning, self-adaptation and self-learning ability. The structure and learning algorithm of the fuzzy neural network are also given. Then the position loop of PMSM servo system is simulated with MATLAB simulation tool under different control strategies, and the fuzzy neural network controller is used in the position loop of PMSM servo system. It will have better response speed and follow performance. After theoretical research and simulation results analysis, the software and hardware design of the system based on DSP is carried out. The TMS320F2812 chip produced by TI company is chosen as the control core of the whole system. The hardware design mainly includes main circuit, control circuit, position control mode design, key and display circuit design, and software design includes main program. Interrupt subroutine SVPWM subroutine and power protection interrupt subroutine. At the end, the work done in the whole article is summarized, and the follow-up work arrangement and future prospect of this topic are given. The machine tool feed servo system based on fuzzy neural network has the advantages of rapid response, high control precision and good stability, and the high performance motor drive system is realized.
【学位授予单位】：辽宁工业大学
【学位级别】：硕士
【学位授予年份】：2016
【分类号】：TG659

【相似文献】