Wyniki wyszukiwania - BazTech

1

Representation of The Pairwise Comparisons in AHP Using Hesitant Cloud Linguistic Term Sets

Huang H.-C., Yang X.

Fundamenta Informaticae

|

2016

|

Vol. 144, nr 3/4

349--362

EN

The analytic hierarchy process (AHP) is the most popular extension to the pairwise comparisons method which is based on the observation that it is much easier to rank several objects when restricted to two objects at one time. As the pairwise comparisons are subjective, the use of linguistic expressions rather than numerical values is straightforward and friendlier due to the uncertainties that are inherent in human judgments. In this paper, to handle the uncertainty and hesitancy in practical decisionmaking situations, we represent pairwise comparisons in AHP using hesitant cloud linguistic term sets (HCLTSs) which are proposed based on hesitant fuzzy linguistic term sets (HFLTSs) and normal cloud models. Then, the synthetic cloud model aggregation algorithm is proposed to transform the HCLTS pairwise comparison matrix into the positive reciprocal synthetic cloud matrix. A prioritization method using the geometric mean technique is adopted, and the ranking method based on comparing of the parameters of normal cloud models is proposed. Thus, we extend the traditional AHP method in hesitant and uncertain environment, and we call it HCLTS-AHP method. The comparative linguistic expressions of preferences become more flexible and richer and are more similar to human beings’ cognitive models. Furthermore, the synthetic cloud model is consistent with objectivity and the calculations are easy to implement. An illustrated example is applied to the ranking of four alternatives to show the usefulness of the proposed HCLTS-AHP method.

2

Using Artificial Neural Networks to Establish a Customer-cancellation Prediction Model

Huang H.-C., Chang A. Y., Ho C.-C.

Przegląd Elektrotechniczny

|

2013

|

R. 89, nr 1b

178--180

EN

In the past, judgments concerning customer cancellations relied primarily on managers’ experience. Prediction errors can cause surpluses or insufficient service capacity. Data mining technology can improve prediction and judgment accuracy. This study applies back propagation neural networks and general regression neural networks to establish a customer-cancellation prediction model. The empirical results showed that both prediction models possessed good predictive abilities and can aid in service capacity scheduling.

PL

W artykule opisano zastosowanie sieci neuronowych o propagacji wstecznej (ang. BPNN) oraz regresji generalnej (ang. GRNN) w budowie modelu anulowania klientów. Działanie to zwykle opiera się na doświadczeniu manager’a, co może doprowadzić do błędnych decyzji. Rezultaty badań empirycznych dowodzą dobrych własności przewidywania i możliwej użyteczności w określaniu potencjalnych działań z klientem opracowanych modeli.

3

Development of Practical Smart House Scenario Control System

Chang A. Y., Huang H.-C., Tai D.-R.

Przegląd Elektrotechniczny

|

2013

|

R. 89, nr 1b

159--161

EN

Smart houses have received significant attention in recent years because they are considered to be an ideal living environment. The key point of smart space is that it is self-adjustable to an optimal state through interactions between people and electronic devices. Object detection technology was applied to efficiently calculate the exact number and location of people. The concurrent RFID authentication mechanisms were examined to identify their security threats, and a two-factor RFID security authentication framework is proposed to be integrated into the central controls. The proposed system also combines heterogeneous appliances so that they could adjust themselves correspondingly to various scenarios.

PL

W artykule przedstawiono projekt systemu kontroli inteligentnego domu, opartego na wykorzystaniu czujników, określających ilość i rozmieszczenie ludzi w pomieszczeniach. Wykorzystano także radiowy system zabezpieczeń RFID w celu uwierzytelnienia lokatorów, który w trybie dwu-parametrowym proponowany jest do jednostki sterującej. Zastosowana dodatkowo, niejednorodna struktura urządzenia pozwala mu dopasowywać się do zmieniających się warunków.

4

FPGA-Based Hybrid GA-PSO Algorithm and Its Application to Global Path Planning for Mobile Robots

Huang H.-C.

Przegląd Elektrotechniczny

|

2012

|

R. 88, nr 7b

281--284

EN

This paper presents an FPGA-based (field-programmable gate array) hybrid metaheuristic GA (genetic algorithm)-PSO (particle swarm optimization) algorithm for mobile robots to find an optimal path between a starting and ending point in a grid environment. GA has been combined with PSO in evolving new solutions by applying crossover and mutation operators on solutions constructed by particles. This hybrid algorithm avoids the premature convergence and time complexity in conventional GA and PSO algorithms. The initial feasible path generated from the hybrid GAPSO planner is then smoothed using the cubic B-spline technique, in order to construct a near-optimal collision-free continuous path. Experimental results are conducted to show the merit of the proposed hybrid GA-PSO path planner for global path planning for mobile robots.

PL

W artykule zaprezentowano algorytm dla mobilnych robotów poszukujący optymalnej ścieżki między punktem startu i końcowym. Algorytm wykorzystuje układy FPGA i bazuje na algorytmach genetycznych i mrówkowych.