Sharif Digital Repository

In this paper, a radial basis function neural network-based (RBFNN) adaptive inverse controller for real-time position tracking control of a two phase hybrid stepper motor in microstep mode is presented and compared with a conventional fixed gain stand-alone PID controller. Also it is compared to the same neuro-controller using feed forward backpropagation neural networks (FFBP) presented by the same author showing superior performance of RBFNN to FFBP neural networks. A radial basis neural network is used as a controller in the adaptive inverse control method combining with a closed loop PID controller to ensure system's initial robustness and closed loop stability and improves the... 

A new switching table is proposed for decreasing the torque ripple in DTC especially in low speed. Each sector is divided to two half sector and a proper vector set is defined for each. Through this new switching table, the torque response is smooth even in the beginning of each sector and it leads to the reduction of torque ripple. None of motor parameters are used in this switching strategy and there is no need to shorten the sampling period. Simulation and experimental results confirm the ability of this method in reducing the torque ripple. ©2004 IEEE  

In this study, we have used a single link system with a pair of muscles that are excited with alpha and gamma signals to achieve an oscillatory movement with variable amplitude and frequency. This paper proposes a reinforcement learning method with an Actor-Critic architecture instead of middle and low level of central nervous system (CNS). The Actor in this structure is a two layer feedforward neural network and the Critic is a model of the cerebellum. The Critic is trained by State-Action-Reward-State-Action (SARSA) method. The system showed excellent tracking capability and after 280 epochs the RMS error for position and velocity profiles were 0.02, 0.04 radian and radian/sec,... 

The notable effeets of time delay aid time discretization on the torque and current ripple amplitude of the direct torque controlled induction motor scheme are investigated. A simply modified torque hysteresis comparator structure is proposed. It minimizes the undesirable effects of the time delay and time discretization on the torque ripple and allows a sensible reduction of torque and current ripple amplitudes and switching frequency in the whole rotor speed range. Numerical simulations and experimental tests have been carried out to validate the proposed torque comparator effect. © 2004 IEEE  

A flux search controller is proposed to achieve the best efficiency in a direct torque controlled induction motor with light load. The reference flux value is determined through a minimization algorithm with stator current as objective function. Experimental results show that using motor current instead of motor input power as the objective function causes better performance. © 2004 IEEE  

Real-time linear robust control of a two phase hybrid stepper motor with Quantitative Feedback Theory method in micro-stepping operation is considered in this paper. Utilizing the phase currents as inputs, linear robust controller is derived for a Hybrid Stepper Motor that achieves robustness to parametric and dynamic uncertainties such as viscous friction, load torque, flux linkage and other uncertainties. Simulation and experimental studies are presented to show the efficiency of the control design approach  

In Direct Torque Control, selection of each voltage vector is based on the measured parameters in the beginning of the sampling period. Due to the fast dynamic of torque, this delay causes an extra torque ripple. In this paper, a new predictive controller is proposed for compensating this delay and reducing the torque ripple. The stator current is measured in the beginning of each sampling period and its expected value in the end of period is predicted according to an extrapolation algorithm. This algorithm is performed by calculating the slope of stator current variation according to the stator flux variation. Thus, there is no need to solve the high order equations of machine. The... 

In the classical DTC scheme, due to the limited number of the VSI voltage vectors, large and small errors can not be distinguished. In fact, the switching vectors chosen for large errors are the same as those chosen for small errors. In this paper, in order to overcome this problem, after considering the primary ideas of the DSVM-DTC technique, a new approach for the realization of the DSVM-DTC methodology of the induction machine is proposed. The proposed method, along with its simplicity, has an improved performance in terms of the torque and the current ripple and can be easily extended. In addition, it minimizes the steady state torque error. By using the proposed DSVM-DTC methodology,... 

In Direct Torque Control, selection of each voltage vector is based on the measured parameters in the beginning of the sampling period. Due to the fast dynamic of torque, this delay causes an extra torque ripple. In this paper, a new predictive controller is proposed for compensating this delay and reducing the torque ripple. The stator current is measured two times at the beginning of each sampling period and its expected value at the end of period is predicted according to an extrapolation algorithm. Thus, there is no need to machine parameters and the algorithm is quite robust against machine parameters variations. The calculation of the electromagnetic torque is performed using the... 

Torque ripple is a major difficulty in direct torque controlled induction motors. The reference value of the motor flux greatly influences the torque ripple. On the other hand, many motors work in an operating cycle, which includes load intervals of less than their nominal power. In these situations, it is not necessary to set their flux value as the nominal value. In this paper, the effect of the reference value of the motor flux on the motor torque ripple is investigated. A new control algorithm is proposed to determine the reference value of the flux according to the value of the motor torque. Simulation and experimental results justify the validity of the proposed method to reduce the... 

In this paper, a linear piezoelectric stepper (LPS) motor drive using a resonant converter is proposed. The mechanism of operation of the LPS motor is first described. Then, a single-phase equivalent model of the LPS motor is introduced. In addition, design theory of the driving circuit using a current source inverter for the LPS motor is explained in detail. At the end, simulation and experimental results based on the resonant converter are presented and analyzed to verify the validity of the proposed system  

Induction motor parameter estimation is generally needed for such purposes as fault detection and the achievement of high dynamic performance drives. This paper is an attempt to use variable structure control (VSC) methodology for the on-line estimation of several significant mechanical and electrical induction motor parameters. The estimated parameters are rotor resistance, magnetizing inductance, stator resistance and viscous damping coefficient In this combined control/estimation method, we propose to apply field-oriented control to the non-linear model of induction motor, and then transform the model by Input-state linearization into canonical form. Application of variable structure... 

The flux and torque hysteresis bands are the only gains to be adjusted in direct torque control (DTC). The torque ripple and harmonic loss of motor and switching loss of inverter greatly Influence them. Hence, these variables must be observed in the control process to maximize performance of the system. In this paper, the effects of flux and torque hysteresis bands on inverter switching loss, harmonic loss, and torque ripple of induction motor are investigated. In order to find their optimum values, a cost function consisting motor harmonic losses, torque ripples, and inverter switching losses is defined. Minimizing this cost function leads to the optimum value of the hysteresis bands of... 

Flux spectrum is greatly influenced by flux and torque controller hysteresis loops (dead bands) in direct torque control of AC motors. On the other hand iron loss in induction motor is closely related to flux harmonics. Hence flux spectrum must be observed in control process to minimize the motor losses. In this paper the relation between the magnetic-flux spectrum and electromagnetic-torque and magnetic-flux controller hysteresis loops (dead bands) as well as machine parameters is investigated. Simulation results are presented to justify the results  

A new control law for direct yaw moment control of an electric vehicle is developed. Although this study is considered as part of a global control system for the traction control of a four motorized wheel electric vehicle, but the results of this study is quite general and can be applied to other types of vehicles. The dynamic model of the system has been analyzed and, in accordance with the optimal control theory, an optimal controller is designed. Two different versions of the control law have been considered and the performance of each version has been separately studied and compared with each other. Finally, the numerical simulation of the vehicle-handling model with and without the use... 

A linear piezoelectric step motor (LPS) driver using current source inverter is proposed in this study. The mechanism of operation of an LPS motor is first described. Then, a single-phase equivalent model of the LPS motor is introduced. A detailed theory of driving circuit design using current source inverter for the LPS motor is explained. At the end the simulation and experimental results are presented and compared