Khurshed Alam entitled "Performance Analysis of a DTC 4S3P VSI- Fed High Performance Induction Motor Based Induction Motor" has been approved by the Board of Examiners in partial fulfillment of the requirements for the degree of Master of Science in Engineering in Department of Electrical and Electronics Engineering, Khulna University of Engineering & Technology, Bangladesh in September 2013. The work presented in this project would not have been possible without the support of the Administration of Khulna University of Engineering and Technology, under the grant of my proposal to the department of Electrical and Electronic Engineering. But overshoot and undershoot cannot be eliminated in the speed response of the drive system.
Introduction
The additional components such as the solver and position sensors not only increase costs, but also affect the reliability of the system. Today, researchers are not only working on improving the performance of the drives, but also trying to reduce the cost of the drives. The authors show a performance comparison of the 4S3P inverter-powered drive with the 6S3P inverter-powered drive in terms of speed response and total harmonic distortion of the stator current.
Literature Review
These converters have some disadvantages such as losses in the six switches, complexity of the control algorithms and the generation of six PWM logic signals [6-10]. The authors also showed a performance comparison of the 4S3P inverter-fed drive with 6S3P inverter-fed drive in terms of speed response and total harmonic distortion of the stator current. The flux observers based on recurrent neural network (RNN) methods are implemented in [20] in which mean square error (MSE) values of the rotor flux estimation are between 0.000087 and 0.000264.
Thesis Overview
Neural networks have been well established in induction motor drives for various tasks, especially flux estimation. But now no literature is available to the authors for chaotic learning based ANN for induction motor drive flux estimation. Various approaches have been proposed to solve the problem of parameter sensitivity in sensorless control of the induction motor speed [21-23].
Mathematical Model of Induction Motor Drives
Introduction
Induction Motor Model
- Stationary Two-Axis Model
- Synchronously Rotating Two-Axis Model
It appears that the coupling between the stator and rotor coils is a function of position of the rotor and is continuously variable. Taking into account the voltage drop in the stator due to resistance, self and mutual inductances, the stator circuit equations are written. Here the variables [ds qs] and [dr qr] imply the flux linkages with the stator and rotor circuits along the synchronously rotating d- and q-axis respectively.
Phase Relationship
Induction Motor Model in terms of Stator Current and Rotor Flux
Conclusion
Introduction
Induction Motor Controllers
- Voltage /Frequency Controllers
- Vector Controllers
- Field Acceleration Controllers
This method is based on avoiding electromagnetic transients in the stator current and keeping the phase continuous. As it achieves a certain computational reduction and overcomes the main problem of vector controllers, the method becomes an important alternative for vector controllers [29].
Direct Torque Controllers
- Principles of Direct Torque Control
- DTC Controller
- DTC Schematic
3.4) It is clear from equation (3.4) that the stator flux vector s lags the corresponding stator voltage vector v by 90 degrees.s. 3.1, where the positional relationship between stator voltage vector and stator flux vector is clearly indicated. Decoupled control of the stator flux modulus and torque is achieved by acting on the radial and tangential components, respectively, of the stator flux linkage space vector at its locus.
3.1, a three-phase voltage system is obtained by connecting the phase 'c' terminal of the stator windings directly to the center tap of the DC link capacitors. According to this principle, a voltage vector switching table for the DTC system powered by a 4-switch inverter can be shown in Table 3.2. The torque and flux hysteresis comparators are two valued comparators and the output of these comparators are denoted as and respectively. There are two different loops corresponding to the stator flux size and torque.
The reference values for the stator flux modulus and the torque are compared with the actual values and the resulting error values are entered into the two-level and three-level hysteresis blocks, respectively. The outputs of the stator flux error and torque error hysteresis blocks, together with the position of stator flux are used as the inputs of the lookup tables (tables 3.1 and 3.2). 3.4, the stator flux modulus and the torque errors tend to be limited within their respective hysteresis bands.
It can be proven that the flux hysteresis band basically affects the stator current distortion in terms of low order harmonics and the torque hysteresis band affects the switching frequency.
Artificial Neural Network
- Models of a Neuron
- Real Time Recurrent Learning Algorithms
- Input Variable Selection Methods for Artificial Neural Networks The development of an ANN provides the non-linear transfer function
- Proposed Stator Flux Estimation Methodology
- Flux Program
- Coordinate Transforms .1 Vector Rotator
- Hysteresis Current Control Technique
The key element of this paradigm is the new structure of the information processing system. The first subscript refers to the neuron in question and the second subscript refers to the input end of the synapse to which the weight refers. The use of bias bk results in an affine transformation being applied to the output uk of the linear combiner in the model of Figure 3.5, as shown by:.
In this figure, the effect of bias is taken into account by performing two arbitrary views: The effectiveness of the proposed flux estimator along α-β axes needs to be verified before it is implemented in the propulsion system. The weights of the RNN are obtained by training the neural network using CRTRL algorithm and are given in the appendix.
Improvements to the performance of the steering systems are given in the last part of the chapter. The flux loop then generates the magnetizing component of the stator current (i*m) and helps maintain the desired λ regardless of the effect of changing parameters. The main task of the current is to force the current vector in the three-phase load according to a reference trajectory.
In addition, the application of PI controller and hysteresis current controller in vector control of IM drive is briefly described.
Starting Performance of the IM Drive
The phase voltages, and resulting from the hysteresis currents controller are transformed into d- and q-axis quantities and.
Performance under Different Operating Conditions
- Sudden Change of Load Torque
- Variation of Stator Parameters
- Speed Reversal
- Ramp Speed Change
- Sudden Disturbance Torque
The estimated rotor angle and the speed responses for the change of the load torque are given in Fig. 4.7(c) and (d) show that the developed torque and motor currents are also insensitive to stator parameter variations. 4.8(c) and (d) show that the developed torque and motor currents are also insensitive to stator parameter variations.
By using a PI controller, the change in the speed response curve of the stator resistance is taken into account at the same time. The effect on the developed electromagnetic torque and motor current due to speed reversal is shown in the figure. The estimated rotor position, speed response, developed electromagnetic torque and motor currents are shown in the figure.
The response curve to the set speed of the regulator gradually increased from zero and at time t=1.0 seconds remained constant up to the value of 1500 rpm. Simulation studies show that the proposed control scheme is still able to accurately estimate the rotor angle. A sudden load disturbance was applied to the IM drive for a very short period of 0.1 seconds.
Discussion
Conclusion
Discussion
Conclusion
The estimated flow components used to estimate rotor flow axis positions also provide accurate results under steady-state and transient conditions. The proposed control methodology derives from the handling of speed and torque errors through controllers instead of instantaneous switching. This work presents a direct torque control methodology for driving an induction motor with 4 switches and a 3-phase inverter.
The control system needs a very simple structure and is able to estimate the torque and stator flux acceptably. The proposed method generates the required voltage components which were used to find out the required voltage magnitude and its position from the inverter. The results of the proposed control system were compared with the control system with PI controller. Very fast speed response, fewer torque pulsations and able to work under different operating conditions indicate the effectiveness of the proposed control method.
It was observed that the proposed control method was robust against parameter changes and perturbations. Based on the results of the previous chapters and the above discussion, it can be concluded that the proposed control scheme provides high performance control for induction motor drives. The proposed control scheme provides good speed response of the motor when the load torque varies from its initial value.
Based on the results from previous chapters and the above discussion, it can be concluded that the proposed CRTRL-based 4S3P control scheme was suitable for the current industrial applications.
Proposed for Future Research
The artificial neural network-based rotor flux estimator presented in this thesis has been shown to be highly accurate and robust to parameter changes. It uses evolutionary algorithm learning-based correlated α- and β-axis flux estimation and has been shown to perform satisfactorily under transient and steady-state conditions. The IM drive follows the reference track very quickly during speed reversal and ramp change of speed.
Bose, and J.O.P Pinto, "Recurrent Neural Network Based Implementation of a Programmable Cascaded Low Pass Filter Used in Stator Flux Synthesis of Vector Controlled Induction Motor Machine", IEEE Transactions on Industrial Electronics, Vol. Ghosh, "An Improved Induction Motor Rotor Flux Estimator Based on Real-time Cross-correlated Iterative Learning Algorithm". Finney, "Chaotic SVPWM-Based Vector Controlled Induction Motor Drives," Transactions of the Electrotechnical Society of China, 2009-2011.
Ashari, "RNN Based Rotor Flux and Speed Estimation of Induction Motor", International Journal of Power Electronics and Drive System (IJPEDS), vol. 21] En Derdiyok, "Speed sensorless control of induction motor using a sliding mode continuous control approach and flux observer,” IEEE Trans. 22] C Lascu, I Boldea, and F Blaabjerg, “Direct torque control of sensorless induction motor drives: A sliding approach,” IEEE Trans.
24] P Vaclavek and P Blaha, “Lyapunov function-based flux and speed observer for sensorless control and parameter estimation of AC induction motors”, IEEE Trans. 25] MJ Duran, JL Duran F Perez and J Fermandez, “Induction Motor Sensorless Vector Control with In-line Estimated Parameters and Overcurrent Protection,” IEEE Trans Ind. Janye, “New direct torque control (DTC) scheme with fuzzy adaptive torque ripple reduction”, IEEE Transaction on Industrial Electronics, June, Vol.
A New Adaptive Integration Methodology for Flow Evaluation in Induction Machine Drives. IEEE Transactions on Power Electronics.
