Wyniki wyszukiwania - BazTech

Ograniczanie wyników

Znaleziono wyników: 2

Liczba wyników na stronie

Wyniki wyszukiwania

Wyszukiwano:
w słowach kluczowych: PLS method

Sortuj według:

Ogranicz wyniki do:

A study framework for assessing the performance of the urban freight transport based on PLS approach

Moufad Imane, Jawab Fouad

Archives of Transport

2019

Vol. 49, iss. 1

69--85

The urban freight transport (UFT) is an important component of the urban logistics. It represents a driving force for the economic dynamics and attractiveness of a city. It can be treated as all movements of goods necessary for the economic activities, the institutions and the residents of urban life. Through, it faces with problems relating to the congestion, unsafety, atmospheric and noise pollution that constrain its performance and hinder its development. This performance is a complex subject on which a great deal of research has emerged in the last decade. Accordingly, this paper aims to develop a model for assessing the performance of the UFT. The objective is to identify the determinants of the performance of the urban freight transport and measure the impact of each factor. Therefore the first part of this article concerns the elaboration of the model and the formulation of the hypotheses. First, an overview of the factors that could influence the performance of the UFT was identified based on a literature review. The result of this step allowed to model, by mobilizing the GRAI grid (Graph of results and interrelated activities), the UFT system in order to release the decisional links between these factors. Then, the grid will be decomposed into hypotheses explaining the relations between the factors and the performance of the UFT. The formulation of each hypothesis will be based on all the theoretical works that have treated it. The second part includes the empirical study to test the model using the partial least squares (PLS) analysis. Therefore, we conducted a survey among managers and users of the UFT in the city of Fez. A hypothetico-deductive approach has been used with a sequential methodological complementarity between qualitative analysis for exploratory purposes and the support of quantitative analysis for confirmation. The results of the test, confirm a significant influence of the identified factors on the UFT performance. The practical scope of this paper is to provide a decision-making framework for urban management department explaining the impact of these factors on the UFT performance.

Identyfikacja lotnych związków na podstawie odpowiedzi matryc czujników gazu

Kalinowski P., Jasiński G.

Zeszyty Naukowe Wydziału ETI Politechniki Gdańskiej. Technologie Informacyjne

2012

T. 20

103--110

W pracy przedstawiono koncepcję systemu wykorzystującego wielowymiarową analizę danych, który służyć będzie do analizy odpowiedzi matrycy czujników gazu w celu identyfikacji lotnych związków występujących w powietrzu. Przedstawione podejście składa się z dwóch etapów. W pierwszym kroku wykonuje się proces kalibracji, polegający na utworzeniu w przestrzeni dwuwymiarowej grupy klas odpowiadających odpowiedziom czujników na konkretne związki. Drugi etap to identyfikacja oraz estymacja koncentracji badanej mieszaniny. Omówiono metody matematyczne wykorzystywane w procesie kalibracji i identyfikacji: metodę analizy głównych składowych (PCA), metodę cząstkowych najmniejszych kwadratów (PLS) i klasyfikatory – k najbliższych sąsiadów (k-NN) oraz Mahalanobisa. Efekt wykorzystania opisywanych metod zademonstrowano na przykładowym wielowymiarowym zbiorze danych dostępnym w literaturze.

In this work, a concept of the volatile compounds identification system based on the gas sensors matrices responses was presented. Proposed approach consists of two stages: calibration and actual measurements. In the stage of calibration, feature extraction techniques transform sensors signals into the new space in order to obtain the classes of single compounds. In the second stage the procedure of identification can be performed by using pattern-recognition techniques. Those methods allow to associate measured parameters of an unknown substance to one of the calibration classes. In this paper, mathematical methods of calibration and classification such as Principal Components Analysis, Partial least Squares and classifiers based on Euclidean or Mahalanobis distance have been described as a part of an air-pollutant identification system based on a gas sensors matrices responses. The result of described mathematical methods implementation using Matlab software has been demonstrated on the prepared for general use multivariate dataset.