Wyniki wyszukiwania - BazTech

1

Supporting of manufacturing system based on demand forecasting tool

Kmiecik Mariusz, Zangana Hawre

LogForum

|

2022

|

Vol. 18, no. 1

33--48

EN

Background: Enterprises’ decision-making could be facilitated by properly creating or choosing and implementing demand forecasting systems. Currently, there are more and more advanced forecasting algorithms based on sophisticated technologies such as artificial neural networks and machine learning. The following research paper focuses on a case study of an automotive manufacturer. The main research aim is to propose the proper demand forecasting tool and show the prospects for implementing the mentioned solution. Methods: The research paper contains the statistical analysis of a chosen time series referring to the demanded quantity of the manufactured products. To create forecasts, models based on the following forecasting algorithms were created: ARIMA, ELM (Extreme Learning Machine), and NNAR (Neural Network Autoregressive). All algorithms are based on the R programming language. All algorithms are run in the same time series where the training and testing periods were established. Results: According to the forecasts ex-post errors and FVA (forecasts value-added) analysis, the best fitting algorithm is the algorithm based on ELM. It yields the most accurate predictions. All other models fail to add value to the forecast. Specifically, the ARIMA models damage the forecast dramatically. Such significant magnitudes of negative FVA values indicate that choosing not to forecast and plan based on the sales of the same period of the previous year is a better choice. However, in the case of the ELM model, the forecasts can be worth the time, finance, and human resources put into preparing them. Conclusion: The increased accuracy of ELM forecasts can contribute to optimizing the process of reaching consensus forecasts. While unconstrained statistical forecasts tend to be overridden, not only to produce constrained forecasts incorporating various variables such as calendar events, promotional activities, supply capacity, and operational abilities, they are also overridden by planners to reflect their foreseeing of demand. The proposed solution could also be easily implemented in the resource planning process to improve it. The proposition of the resource planning process supported by the proposed forecasting system is also shown in the following paper using a BPMN 2.0 (Business Process Modelling Notation 2.0) map.

2

Study on maritime logistics warehousing center model and precision marketing strategy optimization based on fuzzy method and neural network model

Xiao K., Hu X.

Polish Maritime Research

|

2017

|

S 2

30--38

EN

The bulk commodity, different with the retail goods, has a uniqueness in the location selection, the chosen of transportation program and the decision objectives. How to make optimal decisions in the facility location, requirement distribution, shipping methods and the route selection and establish an effective distribution system to reduce the cost has become a burning issue for the e-commerce logistics, which is worthy to be deeply and systematically solved. In this paper, Logistics warehousing center model and precision marketing strategy optimization based on fuzzy method and neural network model is proposed to solve this problem. In addition, we have designed principles of the fuzzy method and neural network model to solve the proposed model because of its complexity. Finally, we have solved numerous examples to compare the results of lingo and Matlab, we use Matlab and lingo just to check the result and to illustrate the numerical example, we can find from the result, the multi-objective model increases logistics costs and improves the efficiency of distribution time.

3

Black box dynamic modelling of proton exchange membrane fuel cells with artificial neural networks

Kapica J.

ECONTECHMOD : An International Quarterly Journal on Economics of Technology and Modelling Processes

|

2016

|

Vol. 5, No 4

85--89

EN

The fuel cells are energy sources which can play an important role in transition of the energy sector into broader use of renewable energy. Numerical modelling provides an easy way to investigate properties of the objects modelled. There are various ways to model dynamic behaviour of the PEM fuel cells including methods using artificial neural networks. There are no clear rules of how a neural network should be configured: how many neurons in the hidden layer and which training algorithm should be used. In a time series modelling task additional parameters including sampling frequency, learning data set duration and number of past data points used for training need to be determined. The paper presents results of research on the influence of various model parameters on the PEM fuel cell modelling accuracy.

4

Neural network modeling of the semi-active magneto-rheological fluid damper

Nabagło T., Pietraszek J.

Czasopismo Techniczne. Mechanika

|

2010

|

R. 107, z. 2-M

193-201

EN

The paper describes model of a semi-active damper based on Magneto-Rheological Fluid (MRF). This model is constructed in a neural networks (NN) system. Such a solution is used because of nonlinear character of the MRF damper elements. The result of research and calculations exposes ways to solve problems connected with this kind of modeling processes. In the final part of the paper, the authors compare results of the NN model verification process with real MRF damper force to velocity characteristics. The article is a complete description of nonlinear model construction with usage of Radial Basis neural Networks (RBN).

PL

W artykule opisano model tłumika pół-aktywnego opartego na płynach magneto-reologicznych (MR). Model ten zbudowano za pomocą sieci neuronowych. Ze względu na nieliniowy charakter tłumików MR zastosowano modelowanie neuronowe tłumików. Na podstawie badań zaprezentowano sposoby rozwiązywania problemów związanych z tego rodzaju modelowaniem. W końcowej części artykułu autorzy porównują w procesie weryfikacji charakterystykę siły do prędkości wynikłą z symulacji modelu neuronowego z charakterystyką rzeczywistego tłumika MR. Artykuł stanowi kompletny opis konstrukcji modelu tłumika MR za pomocą sieci neuronowej o podstawie radialnej.

5

Kinetics of the continuous reaction crystallization of barium sulphate in BaCl2 - (NH4)2SO4 - NaCl - H2O system - neural network model

Piotrowski K., Koralewska J., Wierzbowska B., Matynia A.

Polish Journal of Chemical Technology

|

2009

|

Vol. 11, nr 4

13-19

EN

One of the main toxic components of post quenching salts formed in large quantities during steel hardening processes is BaCl2. This dangerous ingredient can be chemically neutralized after dissolution in water by means of reaction crystallization with solid ammonium sulphate (NH4)2SO4. The resulting size distribution of the ecologically harmless crystalline product - BaSO4 - is an important criteria deciding about its further applicability. Presence of a second component of binary quenching salt mixture (BaCl2-NaCl) in water solution, NaCl, influences the reaction-crystallization process kinetics affecting the resulting product properties. The experimental 39 input-output data vectors containing the information about the continuous reaction crystallization in BaCl2 - (NH4)2SO4 - NaCl - H2O system ([BaCl2]RM = 10-24 mass %, [NaCl]RM = 0-12 mass %, T = 305-348 K and τ = 900-9000 s) created the database for the neural network training and validation. The applicability of diversified network configurations, neuron types and training strategies were verified. An optimal network structure was used for the process modeling.

6

Sieć neuronowa do predykcji własności wytrzymałościowych odlewu

Stawowy A., Macioł A., Wrona R.

Archiwum Odlewnictwa

|

2004

|

R. 4, nr 11/2

222--227

PL

W pracy zaprezentowano modele do wyznaczania zależności między składem żeliwa i temperatury zalewania a wytrzymałością Rm z wykorzystaniem regresji wielorakiej oraz z użyciem sztucznej sieci neuronowej. Ze względu na powiązania między parametrami oraz nieliniowość zjawisk efektywniejszy okazał się model stworzony przez sieć neuronową.

EN

The work describes the models for prediction of tensile strength based on temperature and chemical composition of ductile cast iron. The multiple regression and the neural networks models were employed. It is concluded that ANN model is more effective tool than the regression model.