Downloaded from SAE International by Brought to you by the University of Kansas (Technical reports: 1998 to Present), Saturday, August

25, 2018

Development of Energy Management Strategies and 2012-01-1609

Published
Analysis with Standard Drive Cycles for Fuel Cell 09/10/2012
Electric Vehicles

Prasada Rao Akula and Lakshmi Jandhyala

Mercedes-Benz R&D India

Frieder Herb
Daimler AG

Akash Narayana
Mercedes-Benz R&D India

Copyright © 2012 SAE International

doi:10.4271/2012-01-1609

consumption. Dynamic Programming (DP) is used for

ABSTRACT benchmark calculations with simple vehicle models. As
In order to reduce fuel consumption in Fuel Cell Electric Dynamic Programming requires lot of computation, it can't
Vehicles, effective distribution of power demand between be used for complex vehicle models. So a simple and
Fuel Cell and Battery is required. Energy management effective method is introduced for benchmark calculations
strategies can improve fuel economy by meeting power with detailed vehicle models. This new method is developed
demand efficiently. This paper explains development of based on energy and power requirements of Drive cycle. This
various energy management strategies for Fuel Cell Electric paper presents results of benchmark fuel consumption for all
Vehicle with Lithium Ion Battery. standard Drive cycles.

Drive cycles used for optimization and analysis of the Various control strategies are developed from simple to
strategies are New European Drive cycles (NEDC), Japanese complex. Different strategies like load follower strategy,
Drive cycles (JAP1015), City Drive cycles, Highway Drive equivalent consumption minimization (ECMS) strategies,
cycles (FHDS) and Federal Urban Drive cycles (FUDS). All fuzzy based strategy and neural network based strategies are
Fuel consumption and ageing calculations are done using developed. These strategies are optimized for minimizing fuel
backward model implemented in MATLAB/SIMULINK. consumption using backward model.

This work consists of following topics Detailed analysis of the performance of various control
strategies on different Drive cycles is presented. This analysis
• Benchmark calculations for fuel consumption
consists of fuel consumption per 100 km, influence on stack
• Development of Energy management strategies durability and influence on battery durability. Drive cycles
• Optimization of strategies for minimizing fuel consumption are repeated several times for simulation to get effective fuel
• Analysis of strategies for fuel consumption with different consumption values and to nullify the effect of state of charge
Drive cycles (SOC) of battery on these calculations.
• Analysis of strategies for stack and battery durability with
different Drive cycles INTRODUCTION
Fuel Cell Electric Vehicles still have enormous potential in
Initially, benchmark calculations are performed for Fuel Cell terms of costs reduction and durability. Addition of energy
Electric Vehicle to know theoretical/ultimate limits for fuel
 Downloaded from SAE International by Brought to you by the University of Kansas (Technical reports: 1998 to Present), Saturday, August 25, 2018

storage devices (ESS) to Fuel Cell Vehicles helps compared to the same vehicle to conclude best strategy
significantly to improve the stack durability, efficiency and suitable for vehicle in terms of fuel consumption and
transient response. Energy management strategies play a vital durability of power sources. Various strategies available in
role in improving performance of Fuel Cell Electric Vehicle. literature are extended to suit the requirement of Mercedes B-
So, Energy management strategies have become very class Fuel Cell Electric Vehicle. All the strategies are
important research topic nowadays. Benchmark calculations analyzed in terms of fuel economy and durability of power
are also very important to know the theoretical/ultimate limits sources. Ageing models developed by Frieder Herb [14] [15]
for fuel consumption of hybrid vehicles which helps in are used to compare stack and battery durability for different
component sizing. strategies

There are various methods available in literature for The main objectives of this work are development of vehicle
benchmark calculations of hybrid vehicle. L. Guzella and model, performing benchmark calculations, development of
A.Sciarretta [1] presents various standard methods for energy management strategies and analysis of strategies in
predicting minimum fuel consumption for a given hybrid terms of fuel economy and durability of power sources.
vehicle. Olle Sundstrom and Lino Guzella [3] developed
generic dynamic programming function in matlab and Principle
calculated fuel consumption using dynamic programming
Vehicle model is based on dynamics and energetic models for
approach. Emmanuel Vinot and Rochi Trigui [4] used
different components of vehicle. This vehicle model can be
dynamic programming to compare different configurations of
simulated with different Drive cycles to compute fuel
hybrid vehicles and component sizing.
consumption. There are two methods for solving set of
equations for each component of vehicle: Forward and
A literature survey shows that various energy management
backward methods. Each method has its own advantages and
strategies have been developed for hybrid vehicles. One of
drawbacks.
the Earlier works on energy management strategies is simple
conditions based strategy presented by Jin-Hwan Jung and
Young-Kook Lee[5] where the mode of hybrid controller is Forward Model
defined based on demanded power, state of charge of battery Forward model [2] contains driver model which estimate
(SOC) and binary rules are formed to operate Fuel Cell at power demand based on reference speed and actual speed of
efficient region. Xiangjun Li and Liangfei Xu [6] developed vehicle. Energy management strategy splits demanded power
fuzzy control based strategy for Fuel Cell hybrid vehicle between Fuel Cell system and battery. Actual power of Fuel
where demanded power and state of charge of battery are Cell and battery is computed using dynamics and limitations
taken as inputs for fuzzy controller and estimated optimal of power sources. Total power from power sources is the
load on Fuel Cell using rule base containing nine rules. input to motor model. Motor model converts input power to
Ahmad Fadel and Biao Zhou [7] developed fuzzy based torque and speed, which are inputs to transmission model.
strategy with new structure where degree of hybridization Traction force at the wheels is computed based on
(DOH) is taken as output of fuzzy controller instead of load transmission efficiency and gear ratio. Finally, velocity of
on to Fuel Cell. Zheng Chen and Chris Chunting Mi [8] vehicle is computed based on dynamics of the vehicle.
developed fuzzy based strategy for hybrid vehicles and
compared results with dynamic programming. Gino Paganelli These forward models are very complex, hence can not be
and Yann Guezennec [9] developed equivalent consumption used for optimizing parameters of energy management
minimization strategy (ECMS) and compared results with strategies. These models are not suitable for comparing the
SOC-proportional control based strategy. Cristian Musardo performance of vehicle with different energy management
and Giorgio Rizzoni [10] developed adaptive energy strategies as there always exists difference between reference
management strategy based on ECMS frame work. Xie velocity and actual velocity of the vehicle and this difference
“Chang”- jun and Quan Shu-hai [12] presented energy also depends upon energy management strategy being used.
management strategy based on Neural Network Optimization.
Xia Meng and Nicolas Langlois [13] developed energy Backward Model
management strategy based on adaptive neuro fuzzy
inference systems [NFIS]. Frieder Herb [14] [15] developed Backward model [2] does not require driver model to control
ageing models for Stack & Lithium ion battery and analyzed speed of the vehicle. Traction force at the wheel is computed
effect on stack and battery durability with simple energy directly from velocity and gradient using set of vehicle
management strategies. dynamic equations. Component to component, this
calculation approach carries backward against the tractive
Each energy management strategy mentioned previously may power flow direction to calculate power required at motor
improve the vehicle efficiency when applied to certain hybrid input. Energy management strategy splits power between
vehicles. These energy management strategies have not been Fuel Cell and battery. Finally, fuel consumption and state of
charge of battery are calculated based on respective loads.
 Downloaded from SAE International by Brought to you by the University of Kansas (Technical reports: 1998 to Present), Saturday, August 25, 2018

Figure 1. Fuel Cell Electric Vehicle backward model

Backward model is well suited for comparing performance of information i.e., velocity and gradient as inputs and computes
the vehicle with different energy management strategies as it the required power [1]. This model allows calculation of key
gives same power demand profile for a Drive cycle variables such as fuel consumption and state of charge (SOC)
irrespective of energy management strategy being used. This of battery.
model is very simple and can be used for optimization of
strategies. Longitudinal dynamics of road vehicle is given by equation
(1)
In this work, backward model is used for optimizing
parameters of energy management strategies, fuel
consumption analysis and ageing analysis of power sources
with different Drive cycles.
(1)
Fuel Cell Vehicle Configuration Traction force required at wheels is calculated by equation
This work focused on Fuel Cell Electric Vehicle (2)
configurations where lithium - Ion battery is used as energy
storage device. Main powertrain components are Fuel Cell,
battery and Electric Motor. The Vehicle used in the analysis
is comparable with Mercedes B class F-Cell. Vehicle (2)
specifications are summarized in Table 1 [16].
Aerodynamic friction force is given by equation (3)
Table 1. Fuel Cell Vehicle Specifications

(3)

Rolling friction force is given by equation (4)

Fuel Cell Electric Vehicle Backward

(4)
Model
Figure 1 shows example of Fuel Cell Electric Vehicle
represented in backward model. This model takes drive cycle
 Downloaded from SAE International by Brought to you by the University of Kansas (Technical reports: 1998 to Present), Saturday, August 25, 2018

The force induced by gravity when driving on non-horizontal Energy management strategy splits required power between
road is given by equation (5) Fuel Cell and battery

Fuel consumption is calculated based on Fuel Cell efficiency

map according to equation (11)
(5)
Traction torque at wheels is given by equation (6)

(6) (11)

Assuming there is no slip in the wheels, angular speed of the

wheel is given by equation (7)

Battery current is calculated from equation (12)

(7)
Considering differential gear ratio as 1, then torque at the
wheels is equal to output torque of transmission.
(12)
Motor torque or torque input to transmission is calculated by
considering losses in transmission by equation (8)

(8)

State of charge of battery is calculated from equation (13)

Angular speed of the motor is given by equation (9)

(13)

(9)
Total power required at the input of inverter is calculated
considering efficiencies of inverter and motor by equation
(10) CALCULATING THEORITICAL
LIMITS OF FUEL CONSUMPTION
Backward model can be used to compute power required at
each time step of the Drive cycle. Once power required at
(10) each time step of the Drive cycle is known, next step is to
calculate theoretical/ultimate limits of fuel consumption of a
given Drive cycle for Fuel Cell Electric Vehicle. This paper
 Downloaded from SAE International by Brought to you by the University of Kansas (Technical reports: 1998 to Present), Saturday, August 25, 2018

presents three possible ways of predicting theoretical/ultimate

limits of fuel consumption for a given Drive cycle.

Average Operating Power Approach

In this approach [1], total energy required for a known Drive
cycle is calculated by integrating power required. Then (16)
average power required is calculated as per equation (14).
Stack is operated at average power for the duration of Drive
cycle and fuel consumption rate at each time step is
calculated. Finally, fuel consumption per 100 km is
calculated for the Drive cycle by integrating fuel
consumption rate.

Average power of the Drive cycle is calculated according to

equation (14) (17)
Total energy required for a Drive cycle has to be supplied by
Fuel Cell alone for effective calculation of fuel consumption.

(14) Average power (Pave) is calculated by formulating as

optimization problem as in equation (18)
This method is very simple and fast. Using this method
change in fuel consumption with variation of different
parameters of Fuel Cell system can be studied e.g. operating
temperature, polarization curve of stack etc. This method
(18)
gives minimum possible fuel consumption, but fuel
consumption obtained is not very near to realistic values. This S.T
method is not suitable to find change in fuel consumption
with variation of battery parameters.

Average Operating Power with Battery Fuel Cell is operated as per the equation (15) for the entire
simulation for a given Drive cycle and fuel consumption rate
Limits Approach at each time step is calculated. Finally, total fuel consumption
In this approach, stack is operated at average power (Pave). is calculated by integrating fuel consumption rate. This fuel
The average power (Pave) is calculated based on battery consumption is projected for 100 km based on distance travel
power limits namely maximum discharging power (MDP) using simple proportional rule.
and maximum charging power (MCP)
Using this method effect of battery power limits on fuel
Average operating power with battery limits are obtained by consumption can also be analyzed. This method is simple and
operating Fuel Cell as per following algorithm (15, 16, 17 & required less computations. But, this method can not be used
18) to study effect of capacity of battery on fuel consumption as
there is no constraint on state of charge (SOC) of battery.

Theoretical/ultimate limits of Fuel consumption per 100 km

is calculated with 10 kW, 20 kW, 40 kW and 80 kW
maximum battery powers using this method. Percentage of
decrease in fuel consumption compared to 10 kW battery as
(15) shown in the figure 2. Battery capacity and weight of battery
are taken as constant for this analysis.
Energy required for a Drive cycle is calculated based on
power limits of battery

Total energy required for a Drive cycle is calculated

according to (16)
 Downloaded from SAE International by Brought to you by the University of Kansas (Technical reports: 1998 to Present), Saturday, August 25, 2018

kW and 80 kW maximum battery discharge powers. Capacity

and weight of battery are considered to be constant for these
calculations. Percentage decrease in fuel consumption
compared with 10kW battery is as shown in the figure 4

Figure 2. Percentage of decrease in fuel consumption

compared to 10 kW Battery

Dynamic Programming Approach

Dynamic programming [3] [4] is used to find the optimal
trajectory of SOC of battery to minimize fuel consumption
for a known Drive cycle. The objective function to be
minimized for a Drive cycle is given by (19)
Figure 3. Optimal Trajectory of SOC of Jap1015 using
DP

(19)
S.T

SOC is taken as state variable to define power dynamics of

Fuel Cell Electric Vehicle. SOC is divided into number of
grid points and DP computes Cost-to-Go matrix at each SOC
grid point in backward direction. DP selects SOC values at
each time step whose total fuel consumption is low from
Cost-to-Go matrix in forward direction. This set of SOC
values at each time step forms optimal trajectory of SOC as
shown in Figure 3.

Dynamic Programming is a powerful approach to find fuel

consumption with different limits of battery and Fuel Cell.
Figure 4. Percentage of decrease in fuel consumption
Fuel consumption obtained is very near to realistic values.
compared to 10 kW battery using DP
But, it requires lot of computations and takes lot of time to
converge. It can not be applied to complex models.
ENERGY MANAGEMENT
Power required at each time step is obtained using backward STRATEGIES FOR FUEL CELL
model. Simple Fuel Cell and battery models are developed to
reduce computation time for dynamic programming. ELECTRIC VEHICLE
Required power is taken as input to dynamic programming Energy management strategies developed in this work for
and Cost-to-Go matrix is calculated using simple Fuel Cell Fuel Cell Electric Vehicle can be classified into three
and battery models. Theoretical/ultimate limits of Fuel categories namely heuristic energy management strategies,
consumption per 100km is calculated with 10kW, 20 kW, 40
 Downloaded from SAE International by Brought to you by the University of Kansas (Technical reports: 1998 to Present), Saturday, August 25, 2018

optimal control based energy management strategies and Table 2. Load follower energy management strategy
Neural Network/Learning based control strategies. logic

Heuristic Energy Management Strategies

Heuristic energy management strategies are developed based
on Boolean or fuzzy rules involving various vehicular
variables.

1. Proportional SOC energy management strategy

2. Load follower energy management strategy [5]

Fuzzy Based Energy Managemnet Strategy
3. Fuzzy based energy management strategy [6][7][8] Required power and SOC are converted to fuzzy variables
Proportional SOC Energy Management Very Small (VS), Small (S), Medium (M) and Big (B) using
trapezoidal membership functions. Rule base containing 16
Strategy
rules are used to define Fuel Cell power. These rules are
Load on battery is computed based on lookup table as shown defined based on experience. Weighted average
in figure5 (Battery discharge power is taken as negative). The defuzzification method is used to find fuzzy output. The
difference in demanded power and battery power will be structure of Fuzzy strategy is as shown in Figure 6.
supplied by Fuel Cell. If demanded power is negative, Fuel
Cell is turned off. Battery discharge power in this strategy is
limited to 30% of MDP to control SOC and to get good fuel Optimal Control Based Energy
economy. Management Strategies
Following energy management strategies are developed based
on principles from optimal control theory

1. Equivalent consumption minimization strategy (ECMS)

[9][10]
2. Equivalent consumption minimization strategy with
modes (ECMSModes)[11]
Equivalent Consumption Minimization Strategy
(ECMS)
Main idea behind this strategy is to find minimum fuel
consumption at each time step for a Fuel Cell Electric
Figure 5. Look up table used in Proportional Soc energy Vehicle. Battery power is converted to equivalent fuel
management strategy consumption by multiplying battery power with equivalent
factor and fuel consumption of Fuel Cell is calculated based
on efficiency curve of Fuel Cell system. Total equivalent fuel
Load Follower Energy Management Strategy consumption is the sum of equivalent fuel consumption from
Main objective of this strategy is to operate Fuel Cell at battery and fuel consumption from Fuel Cell system. At each
efficient region and to keep battery SOC in battery efficient time step, total equivalent fuel consumption is computed for
region where battery internal resistance won't change much different Fuel Cell powers from 0 to maximum Fuel Cell
with battery SOC. power (Pfcmax). The Fuel Cell power which gives minimum
equivalent fuel consumption is selected as best operating
This strategy is divided into four regions based on optimum power for Fuel Cell.
power of the Fuel Cell and Minimum allowed SOC of the
battery. Fuel Cell optimum power (Popt) is the power of Fuel Objective function to be minimized at each time step is give
Cell where efficiency of Fuel Cell is maximum. Load on Fuel by equation (20)
Cell is decided based on demanded power and SOC as per
Table 2

(20)
Based on whether the Battery is charging or discharging,
Equivalent fuel consumption for Battery is defined by
equations (21 & 22)
 Downloaded from SAE International by Brought to you by the University of Kansas (Technical reports: 1998 to Present), Saturday, August 25, 2018

Figure 6. Fuzzy based strategy flow from Fuzzification to defuzzification

Case: Discharging Free running mode:

(25)

(21) Neural Network/Learning Based Energy

Case: charging
Management Strategies
Following energy management strategies are developed based
on learning of neural networks to minimize fuel consumption.

(22) 1. Simple neural network based strategy (NN) [12]

2. Neuro fuzzy inference system based strategy(NFIS) [13]

Simple Neural Network Based Strategy (NN)
Optimal Fuel Cell power at each time step is obtained using
Equivalent Consumption Minimization Strategy
dynamic programming (DP) for NEDC Drive cycle. The data
with Modes (ECMSMODES) obtained from Dynamic Programming (DP) is used to train
This strategy is also similar to the ECMS strategy. To keep neural network model offline with demand power and SOC
the SOC within limits, different control modes are introduced as inputs and Fuel Cell power as output. The Neural Network
namely free running mode, SOC increasing mode and SOC (NN) model obtained by offline training is used as energy
decreasing mode. Based on the mode, Fuel Cell power search management strategy.
vector for minimum fuel consumption differs as per the
equations. 23, 24 & 25

SOC increasing mode:

(23)

SOC decreasing mode:

(24) Figure 7. Neural network strategy structure

 Downloaded from SAE International by Brought to you by the University of Kansas (Technical reports: 1998 to Present), Saturday, August 25, 2018

Neuro Fuzzy Inference System Based Strategy degradation and battery degradation. Simple proportional
(NFIS) SOC energy management strategy is taken as base strategy
and all strategies are compared with this strategy.
This strategy is also same as fuzzy based strategy, but rule
base is derived from offline training. Takagi-Sugeno (TS)
rules are used to define the fuzzy output. Weights of TS rules Fuel Economy Comparison
are obtained by optimization to minimize fuel consumption. Energy management strategies are simulated with different
Drive cycles and fuel consumption is calculated per 100 km.
The structure of the NFIS strategy is shown in figure 8. These fuel consumption values are compared with base
Inputs x and y are demanded power and SOC respectively strategy and represented in percentage increase or decrease in
and output f is load on Fuel Cell. fuel consumption as shown in Table3. Negative values mean
decrease in fuel consumption and positive values mean
Layer1: Inputs are fuzzified and output is membership increase in fuel consumption. To nullify the effect of initial
function value SOC on fuel consumption values, Drive cycles are repeated
nine times and fuel consumption is calculated as average of
Layer2: At each node product of member ship function value last two cycles. Total power supplied by both power sources
is taken is compared with demanded power to check drivability at
each time step. Energy difference is calculated between
Layer3: Membership values at each node are normalized energy demanded by Drive cycle and energy supplied by Fuel
Cell. Fuel consumption, energy difference and drivability are
Layer4: the factors to consider for selecting best strategy in terms of
fuel economy. All strategies mentioned in table 3, 4 and 5
met drivability and energy requirements of Drive cycles.

Table 3. Percentage of Hydrogen consumption for

different standard Drive cycles

Layer5: Sum of all outputs from layer 4 to calculate final

output f

Stack Degradation Comparison

In polymer electrolyte membrane Fuel Cell, different
components have different ageing problems. Four most
influenced stress factors such as stack inlet temperature,
current dynamic, cathode humidity and idle current (max.
voltage) have been investigated to model the mathematical
function of aging mechanism in PEM Fuel Cell.[14][15].
Rate of decrease in stack cell voltage is taken as measure for
stack degradation. [14] [15] contain details of Fuel Cell stack
ageing models used in this work.

Figure 8. Neuro fuzzy inference system structure Energy management strategies are simulated with backward
model with different Drive cycles. Drive cycles are repeated
nine times for effective calculation of stack degradation.
ANALYSIS OF ENERGY Stack cell voltage is captured at each time step and rate of
MANAGEMENT STRATEGIES decrease in cell voltage for the duration of simulation is
found. The strategy which has less stack ageing for a given
Energy mangement strategies are integrated with backward Drive cycle is given Rank 1 and is the best strategy for that
model and simulated with various Drive cycles to find best Drive cycle. Next higher value in that Drive cycle is Rank 2
strategy that gives minimum fuel consumption and better and so on.
stack durability for Fuel Cell Electric Vehicles. These
strategies are compared in terms of fuel consumption, stack
 Downloaded from SAE International by Brought to you by the University of Kansas (Technical reports: 1998 to Present), Saturday, August 25, 2018

Table 4. Stack degradation analysis for different strategies which are good for stack durability are bad for
standard Drive cycles battery durability. Percentage of improvement in stack
durability is observed to be higher than the percentage of
improvement in fuel economy. Hence, there is a lot of scope
for improving stack durability with control strategies than
fuel economy. All these results are obtained by simulating
backward model. In future, all these control strategies have to
be simulated using test bench simulator with various standard
Drive cycles and results have to be verified. Development of
control strategies to improve stack durability with minimal
effect on vehicle performance is also an interesting topic for
Battery Degradation Comparison research.
In a Battery different components have different ageing
problems. Four most influenced stress factors such as REFERENCES
temperature, depth of discharge (DOD), state of charge (soc)
1. Guzzella, L., and Sciarretta, A. “Vehicle Propulsion
and current dynamic have been investigated to model an
Systems, Introduction to modeling and optimization”, second
aging mechanism of the Li-ion battery. Rate of increase in
edition, Springer
battery internal resistance is taken as measure for battery
degradation. [14] [15] contain details of Lithium-Ion battery 2. Markel, T. and Brooker, A. “ADVISOR: a systems
ageing models used in this work. analysis tool for advanced vehicle modeling”, Journal of
Power Sources 110 2002, 255-266
Rate of increase in battery internal resistance is taken as
3. Sundstrom, Olle and Guzzella, Lino. “A Generic Dynamic
measure for battery degradation [14]. Energy management
Programming Matlab Function”, IEEE Multi-conference on
strategies are simulated with standard Drive cycles. Drive
Systems and Control, July 8-10, 2009
cycles are repeated nine times for effective calculation of
battery degradation. Rate of increase in battery internal 4. Vinot, Emmanuel and Rochdi, “HEVs comparision and
resistance is captured at each time step and rate of increase in Component Sizing Using Dynamic Programming”, IEEE,
battery terminal resistance is calculated for entire duration of 2007
Drive cycle. The strategy which has less battery ageing for a
5. Jung, J., Lee, Y., Joo, J., and Kim, H., “Power Control
given Drive cycle is given Rank 1 and is the best strategy for
Strategy for Fuel Cell Hybrid Electric Vehicles,” SAE
that Drive cycle. Next higher value in that Drive cycle is
Technical Paper 2003-01-1136, 2003, doi:
Rank 2 and so on.
10.4271/2003-01-1136.

Table 5. Battery degradation analysis for different 6. Li, Xiangjun and Xu, Liangfei, “Control Algorithm of
standard Drive cycles Fuel cell/ Battery Hybrid Vehicular Power system, IEEE,
2008
7. Fadel, A. and Zhou, B., “Power Management
Methodologies for Fuel Cell-Battery Hybrid Vehicles,” SAE
Technical Paper 2010-01-0849, 2010, doi:
10.4271/2010-01-0849.
8. chen, Zheng and Mi, Chris Chunting, “An adaptive Online
Energy management Controller for Power-Split HEV based
SUMMARY/CONCLUSIONS on Dynamic programming and Fuzzy Logic”, IEEE, 2009
In this work, backward model is developed for Fuel Cell 9. Paganelli, G., Guezennec, Y., and Rizzoni, G.,
Electric Vehicle, estimated the ultimate limits of fuel “Optimizing Control Strategy for Hybrid Fuel Cell Vehicle,”
consumption with different approaches, analyzed the effect of SAE Technical Paper 2002-01-0102, 2002, doi:
battery size on fuel consumption with different Drive cycles, 10.4271/2002-01-0102.
developed different energy management strategies from 10. Musardo, Cristian and Rizzoni, Giorgio, “A-ECMS: An
simple to complex and compared energy management adaptive Algorithm for Hybrid Electric Vehicle Energy
strategies in terms of fuel economy, stack degradation & Management”, IEEE conference on Decision and Control,
battery degradation. From the results, it is clear that Neuro December 12-15, 2005
fuzzy inference system (NFIS) based strategy, fuzzy based
strategy and ECMS based strategy are doing well in terms of 11. Sezer, Volkan and Gokasan, Metin, “A Novel ECMS and
fuel economy and stack life time. It is also clear that the Combined Cost Map Approach for High Efficiency Series
Hybrid Electric Vehicles”, IEEE, 2011
 Downloaded from SAE International by Brought to you by the University of Kansas (Technical reports: 1998 to Present), Saturday, August 25, 2018

12. Chang-jun, Xie and Shu-hai, Quan, “Control Strategy of

Hybrid Power System for Fuel Cell Electric Vehicle based on
DEFINITIONS/ABBREVIATIONS
Neural Network Optimization”, IEEE, 2008 FCEV - Fuel cell electric vehicle
13. Meng, Xia and Langlois, Nicolas, “Optimized Fuzzy NEDC - New European Drive cycle
Control Strategy of Hybrid Vehicle Using ADVISOR”,IEEE, FHDS - Federal Highway Drive cycle
2010
JAP1015 - Japanese Drive cycle
14. Herb, Frieder, “Aging mechanisms in Lithium-Ion- DP - Dynamic programming
Battery and PEM-Fuel Cell and their influence on Hybrid-
SOCll - Lower limit of Soc
Systems”, Dissertation, Ulm, Mai 2010, http://vts.uni-ulm.de/
doc.asp?id=7404 SOChh - Higher limit of Soc
NN - Neural Network
15. Herb, Frieder and Trivedi, Kavin, “Aging mechanisms in
Lithium-Ion-Battery and PEM-Fuel Cell and their influence NFIS - Neuro fuzzy inference system
on Hybrid - Systems”, European Fuel Cell Forum 2011. ECMS - Eqvivalent consumption minimization strategy
16. http://www.autowandel.de/cms/servlet/Query? MDP - Maximum Discharge power
node=14190&language=1 MCP - Maximum Charge power
Effdis_batt - Efficiency of battery during discharge
CONTACT INFORMATION Effch_batt - Efficiency of battery during charging
Prasada Rao Akula Pbatt - Battery demanded power
Mercedes-Benz Research and Development Preq - Total required power
Bangalore
India Pfc - Fuel cell demanded power
[email protected]

Lakshmi Jandhyala
Mercedes-Benz Research and Development India
Bangalore
India
[email protected]

Frieder Herb
Daimler AG
GR/AFF
Nabern, Stuttgart
Germany
[email protected]

Akash Narayana
Mercedes-Benz Research and Development India
Bangalore
India
[email protected]

The Engineering Meetings Board has approved this paper for publication. It has Positions and opinions advanced in this paper are those of the author(s) and not
successfully completed SAE's peer review process under the supervision of the session necessarily those of SAE. The author is solely responsible for the content of the paper.
organizer. This process requires a minimum of three (3) reviews by industry experts. SAE Customer Service:
Tel: 877-606-7323 (inside USA and Canada)
All rights reserved. No part of this publication may be reproduced, stored in a
Tel: 724-776-4970 (outside USA)
retrieval system, or transmitted, in any form or by any means, electronic, mechanical, Fax: 724-776-0790
photocopying, recording, or otherwise, without the prior written permission of SAE. Email: [email protected]
ISSN 0148-7191 SAE Web Address: http://www.sae.org
Printed in USA