A neural-network-based space-vector PWM controller for athree-level voltage-fed inverter induction motor drive

A neural-network-based implementation of space-vector modulation (SVM) of a three-level voltage-fed inverter is proposed in this paper that fully covers the linear undermodulation region. A neural network has the advantage of very fast implementation of an SVM algorithm, particularly when a dedicated application-specific IC chip is used instead of a digital signal processor (DSP). A three-level inverter has a large number of switching states compared to a two-level inverter and, therefore, the SVM algorithm to be implemented in a neural network is considerably more complex. In the proposed scheme, a three-layer feedforward neural network receives the command voltage and angle information at the input and generates symmetrical pulsewidth modulation waves for the three phases with the help of a single timer and simple logic circuits. The artificial-neural-network (ANN)-based modulator distributes switching states such that neutral-point voltage is balanced in an open-loop manner. The frequency and voltage can be varied from zero to full value in the whole undermodulation range. A simulated DSP-based modulator generates the data which are used to train the network by a backpropagation algorithm in the MATLAB Neural Network Toolbox. The performance of an open-loop volts/Hz speed-controlled induction motor drive has been evaluated with the ANN-based modulator and compared with that of a conventional DSP-based modulator, and shows excellent performance. The modulator can be easily applied to a vector-controlled drive, and its performance can be extended to the overmodulation region     

Closed-Loop Control of Medium-Voltage Drives Operated With Synchronous Optimal Pulsewidth Modulation

Inverters for medium voltage drives operate at reduced switching frequency so as to restrain the dynamic losses of the power semiconductor devices. The resulting current harmonics can be reduced by synchronous optimal pulsewidth modulation (PWM), provided that steady-state conditions prevail. Transient conditions, however, interfere adversely with the optimal modulation patterns. Such conditions necessarily occur when the modulator forms part of a conventional closed-loop control scheme. Trajectory tracking control is employed to achieve high dynamic control in conjunction with synchronous optimal PWM. An optimal trajectory of the stator flux linkage vector is derived from the pulse pattern in actual use. The stator flux linkage vector is forced to follow this target trajectory. Modifying the target trajectory in transient conditions enables closed-loop torque control in a deadbeat fashion while conserving optimal modulation. Experimental results obtained from a 30-kW prototype drive operated at only 200 Hz switching frequency demonstrate the effectiveness of the approach.      

Neural network based estimation of feedback signals for a vectorcontrolled induction motor drive

Neural networks are showing promise for application in power electronics and motion control systems. So far, they have been applied for a few cases, mainly in the control of converters and drives, but their application in estimation is practically new. The purpose of this paper is to demonstrate that such a technology can be applied for estimation of feedback signals in an induction motor drive with some distinct advantages when compared to DSP based implementation. A feedforward neural network receives the machine terminal signals at the input and calculates flux, torque, and unit vectors (cos &thetas;_e and sin &thetas;_e) at the output which are then used in the control of a direct vector-controlled drive system. The three-layer network has been trained extensively by Neural Works Professional II/Plus program to emulate the DSP-based computational characteristics. The performance of the estimator is good and is comparable to that of DSP-based estimation. The system has been operated in the wide torque and speed regions independently with a DSP-based estimator and a neural network-based estimator, and are shown to have comparable performance. The neural network estimator has the advantages of faster execution speed, harmonic ripple immunity, and fault tolerance characteristics compared to DSP-based estimator     

A novel startup scheme of stator-flux-oriented vector-controlled induction motor drive without torque jerk

This paper proposes a novel zero-speed startup scheme of a stator-flux-oriented speed-sensorless vector-controlled induction motor drive that does not generate any torque jerk. A programmable cascaded three-stage low-pass filter (LPF) method is used for flux vector estimation. Usually, a torque jerk is generated at the vector control transition due to time delay in developing the stator flux by the three-stage LPF. At standstill condition, an algorithm is derived that calculates the stator flux using only the stator currents. A feedforward control strategy of the stator flux is developed to eliminate the torque jerk during transition from the standstill mode to the vector control mode. The performance of the newly developed startup scheme has been verified by simulation and then experimentally on a 3-hp induction motor drive, where the control was implemented by a 32-bit TMS320C30-type digital signal processor. The performance of the drive was found to be excellent in both simulation and experiment.     

Current regulated PWM-CSI induction motor drive system without aspeed sensor

This paper deals with a current source inverter-induction motor drive system without a speed sensor, and the speed estimation with only sense of the motor voltage is investigated. The input DC link current which is kept constant by a pulsewidth modulated (PWM) converter with sinusoidal input current is supplied to the PWM inverter. The inverter output current with sinusoidal wave is directly regulated by employing the modulation index control of the PWM strategy. The motor is driven by a slip frequency/current regulation control. The motor speed is estimated by using the measured voltage and current, or using the measured voltage and the estimated current. The filter capacitor which is installed in the inverter circuit is taken into consideration for the current estimation. A digital signal processor is employed for calculation. As the PWM pulse of the inverter is calculated by an online real-time system, the quick regulation of the sinusoidal output current is achieved. The experiments show that the high performance steady state behavior and the acceleration/deceleration characteristics with smooth drives and low noise of the motor is obtained     

A neural-network-based space-vector PWM controller for voltage-fed inverter induction motor drive

A neural-network-based implementation of space-vector modulation (SVM) of a voltage-fed inverter has been proposed in this paper that fully covers the undermodulation and overmodulation regions linearly extending operation smoothly up to square wave. A neural network has the advantage of very fast implementation of an SVM algorithm that can increase the converter switching frequency, particularly when a dedicated application-specific integrated circuit chip is used in the modulator. The scheme has been fully implemented and extensively evaluated in a V/Hz-controlled 5 hp, 60 Hz, 230 V induction motor drive. The performances of the drive with artificial-neural-network-based SVM are excellent. The scheme can be easily extended to a vector-controlled drive.     

A single-phase three-level pulsewidth modulation AC/DC converter with the function of power factor corrector and active power filter

A single-phase three-level pulsewidth modulation (PWM) AC/DC converter with the function of power factor corrector and active power filter is proposed to reduce harmonic currents flowing into the power system and to draw a nearly sinusoidal current with unity power factor. The circuit topology of the adopted three-level PWM AC/DC converter is based on a conventional two-level full-bridge rectifier and one AC power switch. The control signals of the power switches are derived from the voltage balance compensator, current controller and detected operation region of mains voltage. A three-level PWM voltage pattern on the AC side of the converter in each half cycle of mains frequency is generated. Computer simulations are implemented to confirm the operation of the adopted converter with the function of power factor corrector and active power filter.     

Interaction of drive modulation and cable parameters on AC motortransients

This paper investigates overvoltage transients on AC induction motors when connected through a cable of arbitrary length to a variable frequency drive (VFD) consisting of a pulsewidth modulation (PWM) inverter with insulated gate bipolar transistor (IGBT) power devices. Factors contributing to a motor overvoltage transient equal to a theoretical twice DC bus voltage are first described using existing transmission line analysis. A critical cable distance I_c is defined where this 2-pu overvoltage occurs. However, literature is lacking on flow motor voltage transients >2-pu bus voltage and up to 3-4 pu are generated. This phenomenon is observed on all PWM inverters with output cable lengths greater than l_c distance. Contributing factors to the >2-pu overvoltage phenomenon are investigated by exploring the complex interaction between drive modulation techniques, carrier frequency selected, cable natural frequency of oscillation, cable high-frequency damping losses, and, to a lesser extent, inverter output rise time. Theoretical calculations of cable frequency and damping are correlated with simulation and experimental results. Novel modifications to the PWM modulator, as well as external hardware apparatus, are proposed solutions to the >2-pu overvoltage problem; both are simulated and experimentally confirmed     

A High-Performance Pulsewidth Modulator for an Inverter-Fed Drive System Using a Microcomputer

An Intel 8086 microcomputer-based pulsewidth modulator (PWM) is described which receives digital voltage and frequency commands independently at the input, generates precision three- phase PWM waves at the output, and can be used to drive a transistor or thyristor inverter for ac drive systems. A computation intensive uniform sampling technique is used in the low-frequency region, whereas the higher frequency region is based on word recognition and pattern retrieval method. A laboratory breadboard of the modulator has been built and tested for the frequency range 0-250 Hz with resolution of 0.0077 Hz and smooth voltage variation up to square wave within a one-percent step in the whole range. The modulator has been extensively tested with a transistor inverter and hybrid computer simulated induction motor drive system and shows performance improvement over the currently available techniques.     

Dim the lights. Problems with lamp current control using a PWM signal

It is the state of the art to drive a fluorescent lamp with a ballast that mainly consists of a high-frequency generator. High frequency, in this context, means a frequency that is substantially higher than the mains frequency. Most ballasts on the market are working at an operating frequency of about 50 kHz. It is common to realize the high-frequency generator with a half-bridge inverter. This inverter has an output waveform with a rectangular shape, and the operating frequency f/sub o/ is equal to the frequency of the oscillation of the half-bridge. The fluorescent lamp is connected to the output of the inverter via a matching network. The matching network has two main tasks. The first task is to transform the output impedance of the inverter to an impedance the lamp needs for stable operation. The second task is to cut off the DC component of the output signal that is delivered by the inverter. This task can easily be accomplished by a capacitor that is connected in series to the lamp. Of course, a lamp needs a certain value for the lamp current. One technique to control the lamp current is variation of the pulsewidth of the inverter output signal. The main purpose of this article is to discuss the problems of the pulsewidth modulation (PWM) control of the lamp current and to propose a new driving scheme for a PWM-controlled half-bridge inverter.     

Comparative investigation of PWM techniques for a new drive for electric vehicles

Results of a comparative investigation of deterministic and random pulsewidth modulation (PWM) techniques to be employed in a new ac drive for electric vehicles are presented. The study, employing the advanced simulation package SABER from Avant! Corporation, was focused on the reduction of electromagnetic interference (EMI) when a deterministic PWM strategy is replaced with an RPWM method. To assess the impact of the PWM technique on the drive performance, the drive efficiency, torque ripple, and dynamic response of the current control system were also investigated. Two RPWM techniques characterized by random variations of switching frequency of the inverter were considered: one with the sampling frequency of the modulator varying in step with the switching frequency, and one with a fixed sampling frequency equal to the average switching frequency. The study has demonstrated good EMI-mitigating potentials of RPWM techniques, whose use results in EMI reduction by 10 dB and more, while the drive performance is barely affected.     

电机磁链神经网络观测器的研究

为提高电机磁链观测器的观测性能,实现磁链的准确观测,提出一种RBF神经网络定子磁链观测器.采用RBF神经网络重构基于电压模型的带幅值和相位补偿的变截止频率定子磁链观测器,使磁链观测器的截止频率能跟随电机定子电信号频率的变化而变化.RBF神经网络磁链观测器实现了变截止频率,结构简单,自适应能力强,无直流偏移和初始相位问题...     

神经PID控制在SVPWM整流器VFDPC中的应用研究

分析了PWM整流器虚拟磁链直接功率控制(VFDPC)的工作原理,得出了基于虚拟磁链观测的瞬时功率估算式。设计了2个高通滤波器级联构成的二阶环节代替纯积分器滤除直流分量来解决虚拟磁链观测中存在的因积分初值选取不当造成的直流偏移问题。为提高PWM整流器直流电压稳定性,在电压外环设计了神经PID控制器。仿真结果表明,与传统PID控制相比,基于神经PID控制的SVPWM整流器直接功率控制系统直流侧电压稳定、动态性能更好。     

Rotor time-constant estimation approaches based on energy function and sliding mode for induction motor drive

To solve the problem of detuning due to parameter variations in the current decoupled control of a direct stator-flux-oriented induction motor (IM) drive, two approaches for the rotor time-constant estimation are presented in this study. The first approach is based on the model reference adaptive system (MRAS) using an energy function, and the second approach is based on the sliding mode technique. The estimated rotor time-constant is used in the current decoupled controller, which is designed to decouple the torque and flux in the stator flux field-oriented control. To increase the accuracy of the estimated rotor time-constant, the estimation methods are implemented using a digital signal processor (DSP). The effectiveness of the proposed estimation methods are demonstrated by some simulation and experimental results.     

Optimal Pulsewidth Modulation for Current Source Inverters

Two different pulsewidth modulation (PWM) schemes for current source inverters (CSI) are described. The first one is based on off-line optimization of individual switching angles and requires a microprocessor for implementation and the second one uses a special subharmonic modulation and could be implemented with analog and medium-scale integration (MSI) digital circuits. When CSI's are used in ac motor drives, the optimal PWM pattern depends on the performance criteria being used, which in turn depend on the drive application. In this paper four different performance criteria are considered: 1) current or torque harmonic elimination, 2) current harmonic minimization, 3) speed ripple minimization, and 4) position error minimization. As an example a self-controlled synchronous motor (SCSM) supplied by the PWM CSI is considered. The performance of the CSI-SCSM with the optimal PWM schemes proposed herein are compared with that using a conventional 120° quasi-square wave current.     

A New Neutral-Point-Clamped PWM Inverter

A new neutral-point-clamped pulsewidth modulation (PWM) inverter composed of main switching devices which operate as switches for PWM and auxiliary switching devices to clamp the output terminal potential to the neutral point potential has been developed. This inverter output contains less harmonic content as compared with that of a conventional type. Two inverters are compared analytically and experimentally. In addition, a new PWM technique suitable for an ac drive system is applied to this inverter. The neutral-point-clamped PWM inverter adopting the new PWM technique shows an excellent drive system efficiency, including motor efficiency, and is appropriate for a wide-range variable-speed drive system.     

A simplified approach to continuous on-line tuning offield-oriented induction machine drives

A simplified approach to the continuous online tuning of rotor flux feedforward field-oriented induction machine drives is presented. The procedure offers the advantage of not requiring a special test signal or special test conditions. The approach takes advantage of the stator voltage equations which allow robust parameter-insensitive estimation of the electromagnetic torque while operating at nominal speeds for which the stator IR voltage drop is negligible (≈5 Hz and above). This torque computation is independent of the rotor flux field-oriented control. In essence, the approach uses a stator flux controller to detect improper tuning of the feedforward rotor flux controller. The theoretical and practical implementations of such a continuous self-tuning system are presented. Experimental results are based on a rotor flux feedforward field-oriented induction machine servo drive     

Multilevel inverter modulation schemes to eliminate common-mode voltages

It is well known that conventional two-level pulsewidth modulated (PWM) inverters generate high-frequency common-mode voltages with high dv/dt. Similarly, commonly used multilevel inverter modulation schemes generate common-mode voltages. Common-mode voltages may cause motor shaft voltages and bearing currents and conducted electromagnetic interference (EMI). Premature motor bearing failures and electronic equipment malfunctions have been reported to be directly related to bearing currents and EMI. In this paper, approaches to eliminating common-mode voltage when using multilevel PWM inverters are presented. It is shown that inverters, which have an odd number of levels, will generate zero common-mode voltage by switching among certain states. Therefore, motor bearing currents will be eliminated and conducted EMI will be reduced. Both sinusoidal PWM and space-vector modulation (SVM) schemes are discussed and detailed comparative simulation results between conventional and novel modulation schemes are provided. The value of the proposed technique is demonstrated experimentally by applying the novel SVM approach to a conventional multilevel inverter.     

An improved full-bridge zero-voltage-transition PWM DC/DC converterwith zero-voltage/zero-current switching of the auxiliary switches

In this paper, a novel full-bridge (FB) zero-voltage-transition (ZVT) pulsewidth modulation (PWM) DC/DC converter topology is proposed. The proposed FB-ZVT PWM DC/DC converter has zero-voltage switching (ZVS) of main switches for the entire line and load range and an advantage of achieving ZVS and zero-current switching of the auxiliary switches using the auxiliary network. The auxiliary network is simply composed of a saturable inductor, auxiliary capacitors, and auxiliary diodes. With the help of the proposed auxiliary network, the improved FB-ZVT PWM DC/DC converter has such characteristics as higher overall system efficiency and better utilization of the auxiliary switches compared with the conventional FB-ZVT PWM DC/DC converter. The operation principles are explained in detail and the several interesting simulation and experimental results verify the validity of the proposed circuit with an 83 kHz 1 kW prototype converter using insulated gate bipolar transistors     

Comparison of PWM and PFM induction drives regarding audible noise and vibration for household applications

This paper is aimed at comparing the performance of pulse frequency modulation (PFM) and pulsewidth modulation (PWM) techniques regarding audible noise generated from inverter-driven induction motors. For the purpose of illustrating the performance of the two modulation techniques, a drive developed for washing machine applications is considered. First, the measured and simulated harmonic content of this inverter is compared with the measured harmonic spectrum of a three-phase input-output commercial variable-frequency inverter. It is found that despite the 8-bit processor inexpensive implementation, the total harmonic distortion (THD) of PFM is comparable to the THD of PWM in more sophisticated applications. It is discovered that the harmonic spectrum of PFM exhibits harmonics of smaller magnitude distributed over 0.9f/sub o/--2.1f/sub o/ (base switching frequency) range, as opposed to a smaller number of dominant harmonics of PWM. Next, the prototype drive is programmed to generate both sine-PWM and PFM waveforms. The motor is driven at several frequencies and the audible noise and vibration level of the motor is measured. It is observed that the type of current harmonic content of PFM leads to reduced noise and vibration.