الفهرس Only 14 pages are availabe for public view
 |  | from | 153 |  |  |
| | | from | 153 | | |
Abstract
It is estimated that, more than fifty percent of the generated world electric energy is consumed by electrical motors and the majority of them are induction motors. In Egypt, Ministry of water resources and irrigation uses more than 1500 pumping stations, most of them are driven by induction motors. These pumps operate usually at constant speeds during all states of operation and mechanical control methods are used to regulate the pump flow rate. This method of control increases the energy losses and hence reduces the over all efficiency.
The advantage of using variable speed drive (VSD) control instead of mechanical control in the pump system is to increase system efficiency. Also, more energy saving can be realized by operating the motor at maximum efficiency regardless of the output torque and speed by using efficiency optimization control strategy to select proper flux level in the machine.
The efficiency optimization control techniques for the VSD can be divided into two categories, namely search control technique and model-based control technique.
The search control technique is the control principle where a search algorithm is used to minimize the drive input power which is directly measured. The main advantage of the search control is that it does not depend on the motor parameters and the point of optimal efficiency can be found accurately. On the other hand, it has the disadvantages of long convergence time to reach the optimal efficiency, significant ripples in the machine current and torque during steady state, and the hardware cost of implementation is high.
Model-based control technique is a principle where the drive control contains equations representing the drive losses and uses these equations to calculate the point of optimal efficiency. The motor parameters should be known beforehand. In a manufactured series of drives with specific motors and converters designed to operate together, parameters can be realized by off-line measurements, Otherwise parameter estimation is required. The model-based control is very fast and very cheap technique to implement. However, it has less accuracy compared with the search control method.
Accordingly, this thesis is devoted to develop a model-based efficiency optimization control technique with improved performance. In order to achieve this objective, the following research work is conducted.
A detailed efficiency model for the induction motor drive is developed. The model accounts for iron loss, copper loss, saturation effect, and harmonic losses result from PWM inverter. The effect of motor parameters variation due to magnetic circuit saturation and the operating frequency are also taken into consideration.
A numerical solution method based on the negative gradient technique is developed to solve the efficiency optimization problem. The approach used in the solution is to consider the slip as the independent or control variable. Therefore, the voltage and frequency become dependent variables. The objective is to search for the values of slip which minimize the input power. The efficiency optimization problem has been simulated using the parameters of 5.5 kW and 315 kW 3- phase induction motors. The simulation results have been obtained and discussed for different modeling phases namely; linear magnetic circuit model, nonlinear magnetic circuit model, and nonlinear magnetic circuit with PWM harmonic losses.
The fundamental aspects of modeling artificial neural networks (ANN), architectures, and learning algorithms are presented. An ANN efficiency optimizer has been developed based on the detailed efficiency model. The ANN efficiency optimizer has the advantage of reducing the computational time required to get the value of V/f ratio that gives minimum input power point. Therefore, the proposed efficiency optimization method becomes suitable for real-time implementation.
For proof of concepts, an experimental setup has been constructed. The setup consists of 3-phase PWM voltage source inverter, 5.5 kW 3-phase induction motor, centrifugal pump unit, power quality analyzers, and optical speed meter. The obtained simulation and experimental results are in good agreement and confirm the effectiveness of the introduced model-based efficiency optimization control method.