Wyniki wyszukiwania - BazTech

1

Optimal configuration of standalone wind-solar-storage complementary generation system based on the GA-PSO algorithm

Sun Qian, Ma Jianwei, She Yanjie, Zhang Jingchao, Gu Bo, Zhang Zichao

Journal of Power Technologies

|

2019

|

Vol. 99, nr 4

231--236

EN

The capacity configuration of the standalone wind–solar–storage complementary power generation system (SWS system) is affected by environmental, climate condition, load and other stochastic factors. This makes the capacity configuration of the SWS system problematic when the capacity configuration method of traditional power generation is used. An optimal configuration method of the SWS system based on the hybrid genetic algorithm and particle swarm optimization (GA-PSO) algorithm is proposed in this study to improve the stability and economy of the SWS system. The constituent elements of investment, maintenance cost and various reliability constraints of the SWS system were also discussed. The optimal configuration of the SWS system based on GA-PSO was explored to achieve the optimization objective, which was to minimize investment and maintenance costs of the SWS system while maintaining power supply reliability. The investment and maintenance costs of the SWS system under different configuration methods were calculated and analyzed on the bases of the monthly mean wind speed, solar radiation and load data of Xiaoertai Village in Zhangbei County of Hebei Province in the last 10 years. Results show that the optimal configuration method based on the GA-PSO algorithm could effectively improve the economy of the system and meet the requirements of system stability. The proposed method shows great potential for guiding the optimal configuration of the SWS system in remote areas.

2

Magnetotelluric inversion of one and two dimensional synthetic data based on hybrid genetic algorithms

da Conceição Batista Joelson, Starteri Sampaio Edson Emanoel

Acta Geophysica

|

2019

|

Vol. 67, no. 5

1365--1377

EN

We applied the technique of the genetic algorithms and a local methodology integrating the Gauss–Newton and Conjugate Gradient (GNCG) techniques to test one-dimensional inverse modeling of synthetic magnetotelluric data. The result of this modeling applied to a homogeneous and isotropic five-layer model led to the development a hybrid algorithm (GAGNCG), combining the aforementioned techniques, for inverse modeling of one-dimensional magnetotelluric data. The GAGNCG modeling of the synthetic data performs more efciently than the local methodology in terms of both procedure and results. This showed that the hybridization procedure maximized the advantages of using the global search methodology and minimized the disadvantages of the local technique. Based on these results, we developed another hybrid methodology (GA2D), built from some characteristics of the genetic algorithm and the simulated annealing method, for the inverse modeling of two-dimensional magnetotelluric data. The results were satisfactory, and the GA2D algorithm was a good starting point for the inverse modeling of two-dimensional data.

3

Applying cluster analysis to feature selection for data classification

Klepaczko A.

Zeszyty Naukowe. Elektryka / Politechnika Łódzka

|

2007

|

z. 111

55-63

EN

Data classification is one of the most common tasks irwestigated within the artificial intelligence framework. Its accuracy depends on relevancy of features used to describe the classified objects. However, the fact, which features (among the all measured ones) convey significant information enabling to discriminate the data classes, is known only when feature selection is performed. The paper describes a feature selection method that is capable of solving the problem in an unsuperyised learning mode.

PL

W pracy przedstawiono nową metodę nienadzorowanej selekcji cech. Proponowane podejście zakłada, że wektory danych tworzą w przestrzeni cech znaczących klastry dobrej jakości. Jako miarę jakości klasteryzacji wybrano uśredniony współczynnik kształtu z uwagi na fakt, iż miara ta jest niezależna od położeń centrów klastrów wyznaczanych w procedurze grupowania. Drugim kluczowym elementem opracowanej metody jest tzw. hybrydowy algorytm genetyczny - stratega przeszukiwania przestrzeni cech. Algorytm ten łączy zalety strategii losowych oraz sekwencyjnych. Z jednej strony zachowuje on zdolność do ucieczki z optimum lokalnego optymalizowanej funkcji, z drugiej zaś pozwala kontrolować do pewnego stopnia kierunek poszukiwań. Mechanizm ten zapewnia dużą prędkość zbieżności w pobliżu najlepszego rozwiązania.