基于神经网络的开关磁阻电机建模与控制仿真-电气工程专业论文.docxVIP

中文摘要 开关磁阻电机 (SRM) 具有结构简单、高效率、工作可靠、制造成本低等诸 多优点。由它组成的开关磁阻电机调速系统 (SRD) 因其调速范围广，调速性能 优越而引起了国内外学者的普遍关注。由于开关磁阻电机自身的结构及运行特 点，使得其与传统电机相比具有更强的非线性特征，导致电机的运行性能和控制 方式也不同于传统电机。为了充分发挥开关磁阻电机的优势，必须研究其精确建 协 模方式和高性能控制策略。 首先，应用 MATLAB/Simlink 仿真环境建立了四相 (8/6 极)开关磁阻电机 及其驱动系统的动态仿真模型，对开关磁阻电机的电压 PWM 控制方式进行了动 态仿真分析。仿真中采用 Hall 传感器元件方式得到系统反馈的位置信号。然后， 采用模糊神经网络对电机的输入输出关系进行研究，得到开关磁阻电机的非线性 模型，该方法误差较小，实现简单。将 BP 神经网络和 PID 控制结合起来应用于 开关磁阻电机驱动系统，提出一种基于 BP 神经网络的在钱辨识与自适应 PID 控 制策略，利用非线性映射和自适应能力较强的 BP 神经网络实时观测系统输出， 在线优化 PID 控制中比例、积分、微分各分量的参数。仿真结果表明该方法具有 较好的响应速度、稳态精度和适应性强等优点。 关键词:开关磁阻电机;脉宽调制;模糊控制;神经网络 、 ABSTRACT The switched reluctance motor (SRM) possesses many specific advantages such as simple structure，high efficiency，good reliabilit弘 and low cost. The switched reluctance motor drive (SRD)，which is constructed by the SRM，has attracted ‘extensive attention in the world due 阳性s wide range and superior 阳rformances of speed tuning.ηle Switched Reluctance Motor (SRM) has strong nonlinear feature due ωits structure and operation mode，and the performances and control methods of ‘ this new kind of motors are significantly different 仕om those traditional ones. As a F resuh，the accurate modeling and high performance control of the SRM provesωbe very important and has gained widely research. Firstly，this paper establishes the dynamic simulation models ofthe 牛肉剧e (8/6 poles) SRM and its drive system based on MATLAT/Simul ink， and then the research on voltage PWM speed control str.a.tegies is proωssed on it. The position feedback signals are obtained via the Hall means. Then the input and output relationship of the SRM is researched by the fuzzy neural network method. The nonlinear model of SRM is obtained. It has low eITor and can be performed easily. W批h the BP neural network and the traditional PID controller combined，a new control method based on adaptive BP neural network PID controller and on-line identification is presented. This method uses BP neural network， which has powerful nonlinea