Wyniki wyszukiwania - BazTech

1

Investigation of the stator inductances of the expanded Park model and an approach on parameter identification using the evolution strategy

Schmuelling Ch., Kreischer Ch., Gołebiowski M.

Archives of Electrical Engineering

|

2016

|

Vol. 65, nr 3

585--599

EN

Commonly, the Park model is used to calculate transients or steady-state operations of synchronous machines. The expanded Park theory derives the Park equations from the phase-domain model of the synchronous machine by the use of transformations. Thereby, several hypothesis are made, which are under investigation in this article in respect to the main inductances of two different types of synchronous machines. It is shown, that the derivation of the Park equations from the phase-domain model does not lead to constant inductances, as it is usually assumed for these equations. Nevertheless the Park model is the most common analytic model of synchronous machines. Therefore, in the second part of this article a method using the evolution strategy is shown to obtain the parameters of the Park model.

2

Discrepancies of the expanded park equations and parameter identification with evolution strategy

Schmuelling Ch., Kreischer Ch., Gołȩbiowski M.

Maszyny Elektryczne : zeszyty problemowe

|

2015

|

Nr 3 (107)

81--87

EN

The Park equations are a well known analytic model for describing synchronous machines. Over the years they were consistently extended, like the additional Canay leakage inductance in the rotor circuit. Although, there are newer and more accurate analytic models, the Park equations are still applied by all manufacturers and grid operators. The expanded Park theory [1] derives these equations from the phase-domain model. In this paper at hand, it is shown that this derivation is based on false assumptions for calculating the inductances. Therefore the transformation of the phase-domain parameter does not lead to a decoupled system as it should in the expanded Park theory. It is shown that the new system is more complex and the parameter of this system could not be derived from the phase-domain model. Two different synchronous machines, one with a salient pole rotor and one with a round rotor, are modeled by a finite-element-method (FEM) software and all stator self and mutual inductances are calculated. The model of the round rotor machine is verified by measurements. Due to common application of the Park equations a new method to determine the parameter is also presented. The results of a three-phase sudden short circuit are used in combination with the evolutionary algorithm to derive the Park parameters. All optimized parameters are totally detached from the physics. The found parameters are verified by measurements of other transients, like the twophase short circuit.

3

Zastosowanie algorytmu przeszukiwania rozproszonego do problemu marszrutyzacji z ograniczeniem pojemności środków transportu

Głuszek A., Rudziński F.

Logistyka

|

2015

|

nr 4

3417--3425, CD2

PL

Niniejszy artykuł prezentuje wyniki zastosowania algorytmu przeszukiwania rozproszonego do problemu marszrutyzacji z ograniczeniem pojemności pojazdów. Przeszukiwanie rozproszone zaliczane jest do obszaru algorytmów ewolucyjnych i znajduje wiele zastosowań w optymalizacji problemów o charakterze zarówno ciągłym jak i dyskretnym. Problem marszrutyzacji stanowi zagadnienie należące do zadań optymalizacji kombinatorycznej, a w szerszym zakresie – do badań operacyjnych. Ze względu na jego duże znaczenie praktyczne, zwłaszcza w obszarze zarządzania transportem, wciąż trwają intensywne badania w zakresie poszukiwania nowych i udoskonalania już istniejących algorytmów, umożliwiających jego efektywne rozwiązywanie. W rozdziale pierwszym niniejszego artykułu przedstawiono formalnie zadanie marszrutyzacji z ograniczeniem pojemności pojazdów. Rozdział drugi prezentuje zasadę działania algorytmu przeszukiwania rozproszonego. Rozdział trzeci przedstawia zestaw problemów testowych wykorzystywanych w niniejszej pracy oraz wyniki przeprowadzonych eksperymentów numerycznych. Rezultaty działania algorytmu przeszukiwania rozproszonego porównano z wynikami uzyskanymi przy zastosowaniu dwóch innych metod ewolucyjnych (algorytm genetyczny i strategia ewolucyjna) oraz zaawansowanego dwufazowego algorytmu heurystycznego, wykorzystującego zmodyfikowany algorytm wspinaczkowy.

EN

The paper presents application of scatter search to capacitated vehicle routing problem. Scatter search belongs to the area of evolutionary computations and it has numerous applications in continuous and discrete optimization problems. Vehicle routing problem is an important combinatorial optimization task that is related to operations research. It has great practical relevance, especially in the fields of transport management, distribution and logistics. Development of the algorithms for efficient solving of the vehicle routing problem is still very intensive. In the first section of the paper capacitated vehicle routing problem is formally presented. Next section describes in outline the scatter search algorithm. The third section presents a set of test examples, used in this study and the results of performed experiments. Proposed approach is also compared with two alternative evolutionary algorithms (genetic algorithm and evolutionary strategy) and advanced two-phase heuristic method, based on modified hill climbing algorithm.

4

Study Actuality of Immune Optimization Algorithm

Shi W., Yang H., Miao Y., Hu Y.

Przegląd Elektrotechniczny

|

2013

|

R. 89, nr 3b

98--100

EN

In this paper, an optimization algorithm base immune principle is expatiated, explain its basic theory and process. And discuss immune algorithm’s advantage than other heuristic algorithms, such as: genetic algorithm and evolution strategy. And introduce several better algorithms base immune algorithm, present application in optimization problems. At last we propose immune algorithm’s further development in optimization problems’ application.

PL

W artykule przedstawiono algorytm optymalizacji oparty na układzie odpornościowym. Opisano podstawy teoretyczne i zasady działania. Dokonano porównania z innymi algorytmami heurystycznymi, jak genetyczny i ewolucyjny. Omówiono także potencjalne zastosowanie algorytmów bazujących na proponowanym rozwiązaniu.

5

A new multi-objective optimization algorithm based on differential evolution and neighborhood exploring evolution strategy

Lobato F. S., Steffen, Jr V.

Journal of Artificial Intelligence and Soft Computing Research

|

2011

|

Vol. 1, No. 4

259--267

EN

In this paper a new optimization algorithm based on Differential Evolution, non-dominated sorting strategy and neighborhood exploration strategy for guaranteeing convergence and diversity through the generation of neighborhoods of different sizes to potential candidates in the population is presented. The performance of the algorithm proposed is validated by using standard test functions and metrics commonly adopted in the specialized literature. The sensitivity analysis of some relevant parameters of the algorithm is performed and compared with the classical DE algorithm without the strategy of neighborhood exploration and with other state-of-the-art evolutionary algorithms.