永磁调速器隔热模型分析及智能控制策略研究

发布时间：2018-07-01 12:34

本文选题：永磁调速器 + CAD　；参考：《东北石油大学》2017年硕士论文

【摘要】：目前永磁调速器的主要研究方向是对电磁场、温度场和结构场等的仿真分析以及对永磁调速器的传动性能的优化计算。在各种场的分析过程中忽略了永磁调速器的结构及永磁体失磁的影响。通过阅读大量文献,发现永磁调速器的结构建模方面可以进一步优化,即在传统的永磁调速器的结构基础上加上隔热结构。传统的永磁调速器系统中,为了防止永磁体温度过高引起永磁体磁性能衰减和削弱有效工作磁场以及降低传递转矩等,在系统中加入散热结构,同时避免温度升高到居里温度点时,永磁体永久失效。传统结构的散热方式主要是以空气对流及液冷散热为主,其弊端是降低了经济效益,且散热结构需要定期检查修理,给生产带来一定的不便。本论文提出的隔热模型可以很好地处理因永磁体过热而产生失磁问题。本论文的永磁调速器的调速控制系统选用的是自抗扰算法控制系统,因为该控制方法的优点是对被控对象不要求有精确的数学模型,将系统建模中的不确切性因素与系统中的外部扰动一起作为系统中的同一扰动而进行跟踪补偿。在测量过程中发现传感器没法提供理想反馈信号,传感器自身也有很大系统误差,还有一些噪声、测量误差和信号滞后等缺陷,在很大程度上降低了系统可靠性与准确性。而且,对于调速系统本身也会有一些相关的摄动,运行时还有负载侧带来的不确定干扰。虽然自抗扰控制方法具有较强的鲁棒性,能够很好地观测出系统中的变化量而及时地做出补偿,但是自抗扰控制算法的参数较多,对参数调节难度大,从而很难实现快速调节参数提高设计效率。在此基础上,本论文用了量子粒子群算法对自抗扰算法进行了改进。首先,本文通过CAD和ANSYS软件联合对永磁调速器的铜盘、永磁体盘及初始气隙长度分别作了有限元的磁密分析,设置不同参数观察铜盘、钢盘及初始气隙磁密的变化,从而确定永磁调速器的模型参数。之后对永磁调速器的永磁体进行有限元热分析,通过数值计算及仿真得出永磁体隔热膜厚度,确定永磁调速器加隔热膜之后的三维模型。其次,本论文对永磁调速器的控制策略进行选择设计。选用双直流力矩电机作为永磁调速器的控制电机,选择自抗扰控制器作为永磁调速器的控制系统,用量子粒子群算法对永磁调速器控制器进行优化,并通过QPSO(量子粒子群算法)和PSO(粒子群算法)对比分析,利用MATLAB软件搭建两种算法的永磁调速器的仿真模型,分别对自抗扰控制器的TD、NLSEF、ESO参数作了优化,在s函数模块中编入程序,输出仿真结果。对比得出量子粒子群算法优化后的控制器具有很好的控制性能,提高了永磁调速器的运行效率及调速性能。
[Abstract]:At present, the main research direction of permanent magnet governor is the simulation analysis of electromagnetic field, temperature field and structure field, and the optimization calculation of transmission performance of permanent magnet governor. In the analysis of various fields, the structure of permanent magnet governor and the influence of permanent magnet loss of magnetic field are neglected. By reading a large number of literatures, it is found that the structure modeling of permanent magnet governor can be further optimized, that is to say, the structure of permanent magnet governor can be further optimized by adding heat insulation structure to the structure of traditional permanent magnet governor. In the traditional permanent magnet governor system, in order to prevent the permanent magnet temperature from decaying the magnetic performance of the permanent magnet, weaken the effective working magnetic field and reduce the transfer torque, the heat dissipation structure is added to the system. At the same time, avoid the permanent invalidation of permanent magnets when the temperature rises to Curie temperature. The traditional structure of heat dissipation is mainly air convection and liquid cooling heat dissipation, its disadvantage is to reduce the economic benefits, and the heat dissipation structure needs regular inspection and repair, which brings some inconvenience to production. The thermal insulation model proposed in this paper can deal with the loss of magnetic field due to overheating of permanent magnets. In this paper, the speed control system of the permanent magnet governor is the ADRC control system, because the advantage of the control method is that there is no need for accurate mathematical model for the controlled object. The imprecise factors in the system modeling and the external disturbances in the system are used to track and compensate for the same disturbance in the system. In the process of measurement, it is found that the sensor can not provide the ideal feedback signal, the sensor itself also has a large system error, and there are some defects such as noise, measurement error and signal lag, which greatly reduce the reliability and accuracy of the system. Moreover, there are some related perturbations for the speed control system itself, and the uncertain disturbance caused by the load side. Although the active disturbance rejection control method has strong robustness, it can observe the variation of the system well and compensate in time, but the ADRC algorithm has more parameters and is difficult to adjust the parameters. Therefore, it is difficult to quickly adjust the parameters to improve the design efficiency. On this basis, this paper uses quantum particle swarm optimization algorithm to improve the active disturbance rejection algorithm. Firstly, the magnetic density of copper disk, permanent magnet disk and initial air gap length of permanent magnet governor are analyzed by CAD and ANSYS software respectively. Different parameters are set up to observe the change of magnetic density of copper disk, steel disk and initial air gap. Thus, the model parameters of permanent magnet governor are determined. Then the permanent magnet of the permanent magnet governor is analyzed by finite element method. The thickness of the permanent magnet insulation film is obtained by numerical calculation and simulation, and the three-dimensional model of the permanent magnet governor with the thermal insulation film is determined. Secondly, the control strategy of permanent magnet governor is selected and designed in this paper. The double DC torque motor is chosen as the control motor of the permanent magnet governor, and the auto disturbance rejection controller is chosen as the control system of the permanent magnet governor. The quantum particle swarm optimization algorithm is used to optimize the controller of the permanent magnet governor. Through the comparative analysis of QPSO (Quantum Particle Swarm Optimization) and PSO (Particle Swarm Optimization), the simulation model of the two algorithms is built by MATLAB software, and the TDNLSEFESO parameters of the ADRC are optimized, respectively. The program is programmed in the s function module, and the simulation results are output. It is concluded that the controller optimized by quantum particle swarm optimization (QPSO) has good control performance and improves the operation efficiency and speed regulation performance of the permanent magnet governor.
【学位授予单位】：东北石油大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TM301.2;TM921.51

【参考文献】