PreprintPDF Available

An application of Hybrid Bayesian Network (HBN) in Hybrid Electric Vehicle (HEV) manufacturing

Authors:
  • Jeddah College of Technology
Preprints and early-stage research may not have been peer reviewed yet.
HALId:hal-02102982
Preprint submitted on 7 Jan 2021
HAL is a multi-disciplinary open access
archive for the deposit and dissemination of sci-
entic research documents, whether they are pub-
lished or not. The documents may come from
teaching and research institutions in France or
abroad, or from public or private research centers.
L’archive ouverte pluridisciplinaire HAL, est
destinée au dépôt et à la diusion de documents
scientiques de niveau recherche, publiés ou non,
émanant des établissements d’enseignement et de
recherche français ou étrangers, des laboratoires
publics ou privés.
To cite this version:
An application of Hybrid Bayesian Network (HBN)
in Hybrid Electric Vehicle (HEV) manufacturing
Muneer Mujahed Lyati
Lyati Muneer Mujahed, An application of Hybrid Bayesian Network (HBN) in Hybrid Electric Vehicle (HEV) manufacturing
2021
Working paper on Artifical Intelligence
An application of Hybrid Bayesian Network
(HBN) in Hybrid Electric Vehicle (HEV)
manufacturing
Muneer Mujahed Lyati
January, 2021
Abstract:
A network is a hybrid Bayesian network if it has both discrete and continuous variables. In
this research, we discuss how the hybrid Bayesian network can utilized to further understand
the network from subsidies, manufacturing to the environmental quality in the context of
Hybrid electric vehicles.
Keywords: HBN, HEV, manufacturing, artificial intelligence.
Author:
Muneer Mujahed Lyati is a Graduate from
College of Technology in mechanical
engineering as a bachelor of science in
mechanical engineering with major
Engines and Vehicles. His research mainly
focuses on automotive factors, hybrid
cars, electrical cars, engines, and artificial
intelligence https://muneerlyati.com/
Background
Global demand for electric vehicles in 2019 was USD 160.34 trillion and is
estimated to hit USD 793.24 trillion by 2027 for a CAGR of 22.2 per cent. The
demand is largely driven by increased government policies and programs to
promote the adoption of electric vehicles (EVs). Increasing investment in R&D
for the development of advanced technologies is also fueling the growth of the
industry. Compared to conventional vehicles, emission levels are much lower in
electric vehicles, which has led the government to encourage the production of
EVs all over the world. [1] [2] [3] .
[4], [5] [6], [6] [5] [7]
Asia Pacific had the biggest market share in the electric vehicles market in 2019.
The growing demand for electricity supplies in countries like China, Malaysia,
India and Indonesia is due to a growing urban population. Moreover, government
action to reduce pollution has led to the increasing adoption of EVs. Given the
growing environmental awareness and increasing investment in new technology
by key manufacturers, North America is expected to report significant growth
over the forecast p eriod [1] [8].
HEV can also help producers: by increasing the worldwide CO2 emission targets,
HEV sales can reduce the overall CO2 output of a manufacturer's fleet and help
to prevent the related fines. In fact, in Europe, th e report finds that several
automakers will not meet the emission goals and will either have to buy credits
from other manufacturers or face heavy fines. The advantages of CO2 reduction
from HEVs are nowhere near those of BEV and PHEV drivetrains. The
technology is however maturer and can thus serve as a short-term stopover to
achieve these objectives.
Bayesian networks
Bayesian networks are a type of probabilistic graphic model which uses the
Bayesian inference to calculate probability. The Bayesian networks have the
objective of modeling conditional dependency and thus cause, by representing
conditional dependence in a directed graph by edges. Through these relationships,
the random variables of the graph can be efficiently deduced using factors.
Hybrid Bayesian network and Hybrid electric vehicle
A network is called a hybrid Bayesian network if it has both discrete and
continuous variables. We need to specify two new types of distribution in order
to specify a hybrid network: the conditions for a constant variable given to
discerning or continuous parents, and the conditional distribution for a discrete
variable given to continuous parents [9] [10]
Figure 1: A network with discrete variables (Subsidy and environmental quality)
and continuous variables (manufacturing and emission).
The linear Gaussian distribution, in which the child has a Gaussian distribution
whose mean μ varies linearly with the value of the parent and whose standard
deviation σ is fixed, is the most fundamental choice in our case. Two distributions
are necessary, one for subsidies and one for -subsidies.
Figure 2 demonstrates the distribution of probability over emissions as a function
of the size of production, with true and false subsidies, respectively. The last panel
in Figure 2 shows the P(Emission | Manufacturing) distribution, obtained by
summarizing the two subsidy cases.
Note that the slope is negative in each instance, because emissions decrease as
supply increases. (The assumption of linearity, of course, implies that at s ome
point the emission becomes negative; the linear model is reasonable only if the
production size is limited to a narrow range.) The last panel in Figure 1 shows the
P(c | h) distribution, averaging over the two p ossible subsidy values and the two
possible subsidy values.
Figure 2 the distribution OF variables in network
We now discuss the distributions with continuous parents for discrete variables.
The " Environmental quality " node in Figure 1., for instance, It seems reasonable
to assume that the quality of the environment will increase if the emissions are
low and decrease if they are high. The conditional distribution, in other words, is
like a "soft" threshold function. The use of the Integral is one way to create soft
thresholds
󰇛󰇜 󰇛󰇜󰇛󰇜

Then the p robability of ‘‘Environmental quality’’ given Emission is:
󰇛 󰇜
󰇛
󰇜
This means that the emission threshold is around μ, the width of the threshold
region is proportional to σ, and as emissions decrease, the probability of the
environment increases.
References
[1] F. Che Jamil and A. Shariff Adli Aminuddin, Preliminary study of Malaysian eco-friendly car
selection by using analytic hierarchy process, in Journal of Physics: Conference Series, 2019.
[2] G. Rizzo, C. Pisanti, M. DAgostino, and M. Naddeo, Driver intention analysis for a through-
the-road solar hybridized car, in SAE Technical Papers, 2013.
[3] M. Tengiz, NEW METHODOLOGY OF EVALUATING THE EFFICIENCY OF MANAGERIAL
DECISION-MAKING, in Colloquium-journal, 2020, no. 8 (60).
[4] M. Tengiz, APPLICATION OF GAME THEORY SIMULATION IN ENTERPRISE MANAGEMENT, in
Colloquium-journal, 2020, no. 8 (60).
[5] T. Magradze, MATHEMATICAL MO DELING IN THE ENTERPRISE MANAGEMENT, in
Colloquium-journal, 2020, no. 7 (59).
[6] T. Magradze, TAX ADMINISTRATION IN THE RUSSIAN FEDERATION: CURRENT PROBLEMS
AND DEVELOPMENT PROSPECTS, in Colloquium-journal, 2020, no. 5 (57).
[7] R. Sanghvi and A. Gordon, The plant-based-diet and obesity: The process and literature
survey, 2021.
[8] L. James, Impact of Digital Marketing on SME Growth in South Asia: A Case Study on Faheem
Haydar Dealzmag.
[9] T. Choi, Bayesian networks with examples in R, Biometrics, 2015.
[10] L. M. De Campos, J. M. Fernández-Luna, J. F. Huete, and M. A. Rueda-Morales, Combining
content-based and collaborative recommendations: A hybrid approach based on Bayesian
networks, Int. J. Approx. Reason., 2010.
ResearchGate has not been able to resolve any citations for this publication.
Article
Full-text available
Recommender systems enable users to access products or articles that they would otherwise not be aware of due to the wealth of information to be found on the Internet. The two traditional recommendation techniques are content-based and collaborative filtering. While both methods have their advantages, they also have certain disadvantages, some of which can be solved by combining both techniques to improve the quality of the recommendation. The resulting system is known as a hybrid recommender system.In the context of artificial intelligence, Bayesian networks have been widely and successfully applied to problems with a high level of uncertainty. The field of recommendation represents a very interesting testing ground to put these probabilistic tools into practice.This paper therefore presents a new Bayesian network model to deal with the problem of hybrid recommendation by combining content-based and collaborative features. It has been tailored to the problem in hand and is equipped with a flexible topology and efficient mechanisms to estimate the required probability distributions so that probabilistic inference may be performed. The effectiveness of the model is demonstrated using the MovieLens and IMDB data sets.
Article
In last decade, Hybrid Electric Vehicles (HEV) have emerged as real alternatives to engine-driven vehicles, in order to reduce fuel consumption and emissions. But their market share is still limited, as their impact on global fossil fuel demand and CO2 production. In parallel, the possibility of upgrading conventional vehicles to HEV is gaining interest. A research work on the development of a kit for converting a conventional vehicle into a Through-The-Road (TTR) Hybrid Solar Vehicle (HSV) has been recently performed at the University of Salerno, where flexible solar cells, an additional Lithium-Ion battery and two electrically driven wheel-motors have been mounted on a FIAT Punto. Preliminary studies performed by simulation have shown the technical and economic feasibility of this solution. In the proposed vehicle, the control of wheel motors is performed via a Vehicle Management Unit (VMU), which in turn reads data from the OBD port. In order to develop an effective and safe control strategy for wheel-motors, a precise real-time knowledge of the Driver Intention is required. In the paper, a set of mathematical models using data measured only by the OBD port are developed and integrated into a fuzzy logic model. A first release of the control system has been implemented and successfully validated on real driving cycles on a FIAT Punto.
Article
EDITOR: TAESUNG PARKBayesian Networks With Examples in R (Marco Scutari and Jean-Baptiste Denis)Taeryon ChoiApplied Meta-Analysis with R (Ding-Geng Chen and Karl E. Peace)Mira Park
Preliminary study of Malaysian eco-friendly car selection by using analytic hierarchy process
  • Che Jamil
  • A Shariff Adli Aminuddin
F. Che Jamil and A. Shariff Adli Aminuddin, "Preliminary study of Malaysian eco-friendly car selection by using analytic hierarchy process," in Journal of Physics: Conference Series, 2019.
  • M Tengiz
M. Tengiz, "NEW METHODOLOGY OF EVALUATING THE EFFICIENCY OF MANAGERIAL DECISION-MAKING," in Colloquium-journal, 2020, no. 8 (60).
APPLICATION OF GAME THEORY SIMULATION IN ENTERPRISE MANAGEMENT
  • M Tengiz
M. Tengiz, "APPLICATION OF GAME THEORY SIMULATION IN ENTERPRISE MANAGEMENT," in Colloquium-journal, 2020, no. 8 (60).
TAX ADMINISTRATION IN THE RUSSIAN FEDERATION: CURRENT PROBLEMS AND DEVELOPMENT PROSPECTS
  • T Magradze
T. Magradze, "TAX ADMINISTRATION IN THE RUSSIAN FEDERATION: CURRENT PROBLEMS AND DEVELOPMENT PROSPECTS," in Colloquium-journal, 2020, no. 5 (57).
The plant-based-diet and obesity: The process and literature survey
  • R Sanghvi
  • A Gordon
R. Sanghvi and A. Gordon, "The plant-based-diet and obesity: The process and literature survey," 2021.
Impact of Digital Marketing on SME Growth in South Asia: A Case Study on Faheem Haydar Dealzmag
  • L James
L. James, "Impact of Digital Marketing on SME Growth in South Asia: A Case Study on Faheem Haydar Dealzmag."