Using a genetic algorithm for parameter identification of transformer R-L-C-M model

The R-L-C-M model of power transformer is obtained from geometrical structure and is extremely appropriate for studying transient phenomena in a transformer and detecting mechanical faults. The precision of this model depends strongly on the precision of its parameters. The accuracy of these parameters that are calculated by analytical formulas is limited due to different reasons. In this paper a genetic algorithm (GA) is introduced as a method to identify the parameters of R-L-C-M Model without employing any analytical formulas.     

二相混合式步进电机模型参数的辨识

步进电机是一种高度非线性、强耦合的位置伺服执行元件 ,如何建立较为简单、准确的电机模型 ,为伺服系统的设计提供理论基础 ,一直是研究的热点。本文以一种二相混合式步进电动机模型参考矢量控制位置伺服系统的设计为背景 ,提出了一种比较简单的电机数学模型 ,采用最小二乘法和改进的遗传算法相结合的辨识方法获取模型参数 ,以保证其准确性。实验证明 ,模型是比较简单、准确的 ,能够较好地满足伺服系统实时性、准确性的要求     

Saturation model for squirrel-cage induction motors

An induction motor model which includes stator leakage reactance saturation, rotor leakage reactance saturation and magnetizing reactance saturation is presented. This improved model is based on experimental data from 96 motors. The power range of the motors is between 11 and 90 kW. The effects on the torque–speed and current–speed curves of each kind of saturation have been studied. In addition, the parameters of magnetizing reactance saturation and stator leakage reactance saturation have been studied for each motor, and an average value and its dispersion for each parameter are given. This model is considerably more accurate than other models. In particular, it explains the significant differences between theoretical and experimental torque–speed curves in the braking regime (s > 1).     

异步电机三电平矢量控制传动离线自整定

从异步电动机的状态方程出发,通过静止状态下对状态方程的简化和变形,采用特殊信号激励方法,对异步电动机三电平矢量控制系统进行离线参数估计,为传动系统调节器参数设计和参数在线估计做基础,仿真和实验结果验证了方法简单有效。     

Load torque identification in induction motor using neural networks technique

Induction motors are widely used in several industrial sectors. However, the dimensioning of induction motors is often inaccurate because, in most cases, the load behavior in the shaft is completely unknown. The proposal of this paper is to use artificial neural networks as a tool for dimensioning induction motors rather than conventional methods, which use classical identification techniques and mechanical load modeling. Simulation results are also presented to validate the proposed approach.     

改进粒子群算法在异步电机静态参数识别中的应用

提出一种新的粒子群协同优化算法(PSCO).应用该粒子群协同优化算法实现异步电机静态参数辨识.在有噪声的情况下,该算法明显改善了标准粒子群算法和基于遗传的粒子群算法,在解决异步电机静态参数辨识问题时,存在识别参数准确性不高、辨识成功率低的问题.     

Leakage flux saturation effects on the transient performance of wound-rotor induction motors

The prediction of induction machine performance has traditionally been based on the constant parameter models. This approach was later replaced by considering saturation in the main flux paths. However, such models have not been sufficiently accurate for certain transient conditions such as on-line starting and short-circuit. So, an accurate study of their performance should necessarily consider the magnetic saturation effect both in the main and leakage flux paths. In this paper, an experimental procedure to determine the machine parameters and saturation characteristics is adopted. The adopted experimental procedure facilitates the measurement of both stator and rotor leakage reactance saturation characteristics. Two models of saturated induction motors are developed to predict the transient performance of a laboratory wound-rotor induction motor. The results calculated by the proposed models considering and ignoring the leakage flux saturation are compared with the experimental results. The model that considers saturation in the leakage flux paths produces more accurate transient responses.     

异步电机转动惯量在线辨识与自适应控制的研究

小型异步电机在应用中受电机参数及负载扰动的影响很大。本文提出了一种基于模型参考自适应理论的电机转动惯量辨识算法,以及一种基于系统频域分析的速度调节器自整定算法。仿真和实验证实了其有效性。     

Influence of skin effect on mutual inductance of double-cage induction motors

The influence of skin effect on mutual inductance is important to determine fundamental and harmonic currents in the double-cage. A numerical example shows that significant errors arise from existing methods when losses, torque, or power are evaluated. Several methods to analyze mutual inductance are presented. In spite of skin effect, the relation, M_2,3=M_3,2, holds. For the analysis of a double-cage, a direct method not requiring mutual inductance is described.     

Adaptive non-linear speed control of induction motors

In this paper we present two adaptive non-linear speed control algorithms for induction motors. They employ input-output linearization techniques for the motor model in the stator fixed reference frame. The first control algorithm directly tracks speed and rotor flux. The second is designed using torque and rotor flux tracking and is extended for speed control. A key point is that this algorithm ensures exact current limitation in the known parameter case. The adaptation based on Lyapunov design can cope with rotor and stator resistance variations. Significant simulations using sampled controllers are presented emphasizing adaptation and current limitation effects.     

无轴承异步电机气隙磁场辨识方法与应用

针对无轴承异步电机电磁转矩与径向悬浮力这一强耦合的非线性复杂系统,依据转子磁场定向控制的特点,研制了转矩绕组采用转子磁场定向控制,径向悬浮控制所需的气隙磁场通过I-ω法适时辨识的控制系统.应用Matlab/Simulink建立了系统仿真模型.计算机仿真结果表明,实现了电磁转矩与径向悬浮力之间的完全解耦,具有良好的动、静态性能,验证了本文所提方案的有效性.     

Effects of symmetrical voltage sags on squirrel-cage induction motors

This paper analyzes the symmetrical voltage sag consequences on the induction motor behavior when single- and double-cage models are considered, namely current and torque peaks, and speed loss. These effects depend on several variables like sag type, duration and depth. Voltage sag effects are studied by using single- and double-cage models for three motors of different rated power. The double-cage model always predicts torque and current peaks higher than those of the single-cage model. The single-cage model predicts that voltage sags can produce motor instability, whereas the double-cage model is always stable. Therefore, the double-cage model must be used for the simulation of the squirrel-cage induction motor, because the single-cage model can give erroneous results in some situations.     

On-line identification of continuous time-delay systems using the genetic algorithm

This paper proposes a new kind of on-line identification method of continuous time-delay systems from sampled input-output data. In order to track the time-varying system parameters as well as time-delay, the recursive linear least squares (RLS) method is combined in a bootstrap manner with the genetic algorithm (GA) which has a high potential for global optimization. The time-delay is coded into binary bit strings and searched by the GA, while the system parameters are updated by the RLS method. Furthermore, this method (GALS method) is hybridized with the sequential nonlinear least squares method to improve the speed of convergence. Simulation results show that both the GALS and hybrid methods are efficient in the case that the system changes abruptly, and among them the hybrid method has superior convergence property to the GALs method and yields excellent estimation results in the case that the system changes with time continuously.     

传感器非线性校正的遗传支持向量机方法

针对传感器非线性校正中现有的较为常用的神经网络法的不足和支持向量机参数难确定的问题,提出了一种遗传算法和支持向量机相结合的方法,阐述了支持向量机的非线性校正原理和遗传算法优化支持向量机参数的实现过程,并分别采用BP神经网络法和遗传支持向量机方法对压力传感器进行非线性校正.实验结果表明: BP神经网络法使得传感器的最大相...     

自适应局部微调遗传算法

针对遗传算法在有限时间内难于给出高精确度解的问题,在传统遗传操作的前期与后期分别采用自适应与局部微调相结合的方法,给出了一种自适应局部微调的遗传算法,即将遗传代数划分为自适应概率搜索阶段和局部微调阶段,在交叉操作中分别采用自适应概率算术交叉和部分确定性诱导交叉;在变异操作中分别采用自适应随机扰动变异操作和最优个体诱导变异操作.应用该算法对全局最优解领域进行搜索,能在较短的时间内找到高精确度的数值解.对6个典型测试函数的优化问题实验表明,该方法具有快速、稳定和易于实现的优点.     

大功率直线感应电动机推力脉动研究

直线电机边端效应是由定子铁心、次级开断所致的固有效应,该效应会引起电机电磁推力的不同频率的脉动。本文首先基于电磁推力公式对边端效应所致的推力脉动频率进行了理论分析。然后利用一维磁路法对动态纵向边端效应引起的次级感应涡流沿轴向的分布规律进行了理论分析和解析计算,根据次级感应涡流进而计算得到了动态纵向边端效应引起的转差频率两倍频推力脉动幅值的解析表达式,并得到了推力脉动幅值和不考虑边端效应时的电磁推力值的比例关系,最后通过峰值功率20 MW,最大出力900 k N的大功率直线感应电动机实验系统对分析结论进行了试验验证。理论分析和试验结果表明:计及边端效应时直线电机电磁推力存在供电频率两倍频和转差频率两倍频的脉动,其中转差频率两倍频的脉动由动态纵向边端效应导致,主要和次级极数以及转差频率有关。     

基于APSO算法的发电机励磁系统参数辨识

针对传统辨识方法不能辨识励磁系统非线性环节的缺点,提出了一种基于自适应粒子群优化(APSO)算法的发电机励磁系统参数辨识的方法.通过建立待辨识励磁系统的传递函数结构模型,以励磁系统的实际输入作为模型的输入,以实际励磁系统和模型的输出误差最小作为目标函数,利用APSO算法对模型参数进行优化调整,最终得到满足误差要求的励磁系统参数.该方法根据输入输出采样数据直接在时域上进行参数辨识,无需经过FFT变换,方法简便,并且有效解决了励磁系统非线性环节难以有效辨识的问题.仿真结果表明,APSO算法具有较快的收敛速度和较高的辨识精度.     

An analysis of the field acceleration method and of vector control of induction motors

The two-axis theory and vector control based on this theory are well known. On the other hand, Yamamura has proposed the phase segregation method (spiral vector method) for the analysis of induction motors along with the field acceleration method (FAM). In this paper these analytical and control methods are compared. We begin by analyzing the equations for phase segegation, which contain information derived from the three-phase stator and rotor equations. We next demonstrate that the vector control system for constant rotor flux is obtained from the type T-I type field acceleration method when the transient term is zero. By By using the equivalent circuit for a T-I transient, we readily obtain the desired vector control system, because the circuit is identical to the steady-state circuit when the rotor flux is constant. An analytical solution of the transient response for the vector control system is obtained for arbitrary initial conditions. Finally, the stability of type T-II FAM is discussed, with special emphasis on the effect of changes in stator resistance, by computing the torque transfer function. When the stator resistance is correctly estimated, pole-zero cancellation occurs on the imaginary axis.     

异步电动机矢量控制中的转速辨识

介绍了异步电动机无速度传感器矢量控制中的转速辨识方法。基于电机的理想模型和瞬时无功功率理论,采用合适的控制方法,从定子电压电流中提取转子转速信息。给出了控制系统框图,并说明了理论和实际应用中需要解决的问题。     

一种交互式模型参数自适应感应电动机参数在线辨识方法研究

在基于转子磁链模型参考自适应理论的基础上,给出了一种参考模型和可调模型交互式的模型参数自适应(MRAS)方法,使系统能在电机转速和定子电阻辨识之间相互切换.该方法消除了纯积分和电机漏感对算法的影响,电机的低速性能得到了改善;同时,对定子电阻进行在线辨识,从而得到准确的转速观测结果,也能近似地辨识转子电阻.计算机仿真结果证明了该方法的有效性.