Abstract
The aim of this thesis is to estimate the perspectives of integrated switched reluctance drives (I-SRDs), i.e. reluctance machines integrated with converters. It is assumed that such drive series can be manufactured in the power range of 0.75...7.5 kW and speed ranges of 300...3000 rpm and 600...6000 rpm for applications like pumps, fans, conveyors, compressors, extruders and mixers. Based on the performed research and design work it is stated that the new drives have to be developed according to their applications, which determine objective functions and constraints, and that the best possible design should be found as a solution of a synthesis task. Sizing equations are not applied at all. The approach used in the thesis is based on the virtual prototyping concept, i.e. the new I-SRD series is designed in a virtual environment. Therefore, mathematical models and the ways to verify them have to be elaborated. The concepts of multidisciplinary and multilevel modeling are applied. The multidisciplinary model is a combination of interconnected electromagnetic, thermal and noise models. The multilevel concept is the approach when different elements of the drive are described using different languages, i.e. on different levels. Several original solutions are introduced, like the electromagnetic model comprising SIMULINK block-diagrams and MATLAB script, expressions for the correction of the flux linkage due to end-effects, an original equivalent circuit for thermal analysis, which allows using a very simple and fast method to solve the circuit, together with the concept of a multi-layer equivalent cylinder for modeling the motor winding. For verification of the multidisciplinary model a database of test results has been collected using both testing of several reluctance machines in the laboratory and analyzing of test results published by other researchers. After verification the model can be considered as a virtual prototype and can be used in the synthesis process. Several methods of solving the synthesis task were tested. The method, proved to be best suited for solving this task in the proposed form, is the genetic algorithm in the vector form with alphabetic encoding. The genetic algorithm should be coupled with the experimental design method or with the Monte-Carlo method.
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have