First published 2016 in Great Britain and the United States by ISTE Ltd and John Wiley & Sons, Inc.

Apart from any fair dealing for the purposes of research or private study, or criticism or review, as permitted under the Copyright, Designs and Patents Act 1988, this publication may only be reproduced, stored or transmitted, in any form or by any means, with the prior permission in writing of the publishers, or in the case of reprographic reproduction in accordance with the terms and licenses issued by the CLA. Enquiries concerning reproduction outside these terms should be sent to the publishers at the undermentioned address:

ISTE Ltd

27-37 St George's Road

London SW19 4EU

www.iste.co.uk

John Wiley & Sons, Inc.

111 River Street

Hoboken, NJ 07030

USA

www.wiley.com

The rights of Benoît Robyns, Christophe Saudemont, Daniel Hissel, Xavier Roboam, Bruno Sareni and Julien Pouget to be identified as the authors of this work have been asserted by them in accordance with the Copyright, Designs and Patents Act 1988.

Library of Congress Control Number: 2016941911

British Library Cataloguing-in-Publication Data

A CIP record for this book is available from the British Library

ISBN 978-1-84821-980-9

Table of Contents

Cover

Title

Foreword

Introduction

1 Issues in Electrical Energy Storage for Transport Systems

1.1. Storage requirements for transport systems
1.2. Difficulties of storing electrical energy
1.3. The electrical power supply of transport systems
1.4. Storage management

2 Local DC Grid with Energy Exchange for Applications in Aviation

2.1. Introduction
2.2. Onboard grid
2.3. Local DC grid
2.4. Supervisor design methodology
2.5. Specifications
2.6. Supervisor structure
2.7. Selection of design tools
2.8. Identification of different operating states: the functional graph
2.9. Tools
2.10. Membership functions
2.11. Operational graph
2.12. Fuzzy rules
2.13. Experimental validation
2.14. Fuzzy supervisor optimization
2.15. Conclusion

3 Electric and Hybrid Vehicles

3.1. Introduction
3.2. Storage technologies in hybrids and EVs
3.3. Development of EVs and interaction with electric power grids
3.4. EV charging supervision
3.5. The reversible charge of EVs
3.6. Configurations and operating principle of HV
3.7. Energy management in a hybrid vehicle
3.8. Conclusion

4 Railway System: Diesel-Electric Hybrid Power Train

4.1. Introduction
4.2. Design of an autonomous hybrid locomotive
4.3. Conclusion
4.4. Exercise: definition of the energy requirements in the railway sector and application of storage to electric traction
4.5. Appendices

5 Railway System: Hybrid Railway Power Substation

5.1. Introduction
5.2. Hybrid railway power substations
5.3. Energy management in an HRPS
5.4. Experimentation of an HRPS and sensitivity analysis
5.5. Railway smart grid perspective
5.6. Conclusion
5.7. Acknowledgments

Bibliography

Index

End User License Agreement

List of Tables

1 Issues in Electrical Energy Storage for Transport Systems

Table 1.1. Different methods for designing an energy management system, as illustrated throughout this book

2 Local DC Grid with Energy Exchange for Applications in Aviation

Table 2.1. Objectives, constraints, means of action and tools for each strategy
Table 2.2. Simplified linguistic labels
Table 2.3. The 27 fuzzy rules of the supervisor
Table 2.4. The 19 fuzzy rules of the simplified supervisor
Table 2.5. 2D table
Table 2.6. Comparison of precision and table size for different discretization steps
Table 2.7. Comparison of calculation times required for the implementation in real time, according to the selected methods: “Data tables” (DT) and fuzzy logic (FL)
Table 2.8. Sequence of the temporal power profile of the bidirectional charge
Table 2.9. Experiment and effect matrix
Table 2.10. Parameters to be optimized for level 1
Table 2.11. Parameters to be optimized for level 2
Table 2.12. Extract from the experimental matrix (SOC_init = 20, 50 and 80%)
Table 2.13. Effect matrix (SOC_init = 20, 50 and 80%)
Table 2.14. Parameters obtained by optimization
Table 2.15. Optimization of the supervisor parameters “Boolean”, “Completely fuzzy-based” and “Optimized”
Table 2.16. Indicators obtained for the different types of supervisor for initial storage values of 20 and 80%

3 Electric and Hybrid Vehicles

Table 3.1. Parameters of the stochastic mathematical model based on the home or work arrival times and the distances traveled on a daily basis [BOU 15]
Table 3.2. Calculation parameters of the energy distribution bill [BOU 15]
Table 3.3. Objectives, constraints and means of action
Table 3.4. Comparative balance sheet of the distribution costs prior and following optimization
Table 3.5. Global balance sheet of the coordination costs of renewable-powered EVs and CO₂ emissions prior and following optimization
Table 3.6. Numerical values obtained for the IT1-MF centroids

4 Railway System: Diesel-Electric Hybrid Power Train

Table 4.1. ALSTOM rolling stocks references in 2014
Table 4.2. Comparison of specific power values of power trains
Table 4.3. Kinematic characteristics of the case applications
Table 4.4. Line operation profile of the case applications
Table 4.6. Numerical application, acceleration constants of phases 1 and 3
Table 4.7. Table 4.7. Numerical application, temporal parameters
Table 4.8. Numerical application, parameters distance traveled
Table 4.9. Mechanical characteristics of the studied units
Table 4.10. Response table model
Table 4.11. Speed profile over the course of a cycle to be considered for the case applications
Table 4.12. Tramway power profile
Table 4.13. Suburban train power profile
Table 4.14. Regional train power profile
Table 4.15. High-speed train power profile
Table 4.16. Energy consumption over the course of a traction cycle
Table 4.17. Braking energy and energy consumption over the course of a full cycle
Table 4.18. Characteristics of the energy storage systems studied
Table 4.19. Response table model for the sizing of an ESS
Table 4.20. Design of an ESS of the Ni–Cd electrochemical battery type for the recovery of braking energy
Table 4.21. Design of an ESS of the supercapacitor type for the recovery of braking energy
Table 4.22. Design of an ESS of the flywheel type for the recovery of braking energy
Table 4.23. Reduction of energy consumption by using an ESS of the supercapacitor type
Table 4.24. Determination of the power and energy profiles for the dimensioning of the Ni–Cd technology
Table 4.25. Design of an ESS of the electrochemical battery type for the power supply in a non-electrified area
Table 4.26. Design of an ESS of the supercapacitor type for the power supply in a non-electrified area
Table 4.27. Design of an ESS of the flywheel type for the power supply in a non-electrified area

5 Railway System: Hybrid Railway Power Substation

Table 5.1. Extract of electrification standards for European railway power grids [CEN 05]
Table 5.2. Analysis of cost-effective services following the integration of the energy storage function in the case of the HRPS
Table 5.3. HRPS energy supervision specifications [BUZ 15]
Table 5.4. Extraction of the fuzzy rules associated with the management function, based on the operational graphs in Figures 5.23–5.26 [BUZ 15]
Table 5.5. HRPS simulation parameters [BUZ 15]
Table 5.6. HRPS simulation parameters [BUZ 15]
Table 5.7. Definition of the optimization problem at the level of the parameter supervisor [BUZ 15]
Table 5.8. Results obtained from the optimization process of the ST management parameters [BUZ 15]
Table 5.9. Results of the test conducted in the reference case of an RPS (without renewable energy and storage input) [BUZ 15]

List of Illustrations

1 Issues in Electrical Energy Storage for Transport Systems

Figure 1.1. Example of a Ragone diagram for electrochemical technologies and supercapacitors [MUL 13]
Figure 1.2. Evolution of the energy density of lead (Lead), nickel-cadmium (NiCad), nickel-metal-hydride (NiMH) and lithium-ion (Li-ion) batteries [BAS 13]
Figure 1.3. Consumption profiles with or without electric vehicles over the course of one day for the French power system as a whole [SAR 13]
Figure 1.4. Profile of power demand transmitted to a power supply substation by urban trains over the course of one week [PAN 13]
Figure 1.5. Power transmitted to and generated in a local grid on board an aircraft supplying, for example, the flying controls at the wing level [SWI 12]
Figure 1.6. Different time horizons to be considered for the management of a storage system
Figure 1.7. Supervisor structure [BOU 15]
Figure 1.8. Functional graph representing the operating modes [BUZ 15]
Figure 1.9. Membership functions of a fuzzy function (SOC = storage state of charge) [BUZ 15]
Figure 1.10. Membership functions of a fuzzy function (SOC = storage state of charge, S = Small, B= Big) [BUZ 15]
Figure 1.11. Examples of shapes of membership functions for the same variable [BOU 15]
Figure 1.12. Examples of membership functions in a type-2 fuzzy logic [MAR 12]

2 Local DC Grid with Energy Exchange for Applications in Aviation

Figure 2.1. Evolution of electrical power requirements in civil aviation [ROB 12]
Figure 2.2. EHA and EMA [DEL 08]
Figure 2.3. Electric power grids in conventional and “more electric” aircraft [CLE 12]. For a color version of this figure, please see www.iste.co.uk/robyns/energy.zip
Figure 2.4. Mutualization of the “AC/DC conversion” function [SAU 09]
Figure 2.5. Local DC electric power grid
Figure 2.6. Standard DC bus voltage 270 V (MIL-STD-704)
Figure 2.7. Block diagram of the hybrid DC bus storage and dissipation system
Figure 2.8. Supervisor of a hybrid system (storage + dissipation) within the framework of a local energy exchange DC grid (in relation to Figure 2.7)
Figure 2.9. Inputs and outputs according to strategies and supervision levels
Figure 2.10. General functional graph for the control of the hybrid storage–dissipation system
Figure 2.11. Functional graph of Level 1.1
Figure 2.12. Functional graph of Level 1.1.1
Figure 2.13. Functional graph of Level 1.1.2
Figure 2.14. Functional graph of Level 1.1.3
Figure 2.15. Functional graph of level 1.2
Figure 2.16. Functional graph of level 1.2.1
Figure 2.17. Functional graph of level 1.2.2
Figure 2.18. Functional graph of level 1.2.3
Figure 2.19. Functional graph of level 2
Figure 2.20. Functional graph of level 2.1
Figure 2.21. Functional graph of level 2.2
Figure 2.22. Functional graph of level 2.3
Figure 2.23. Level 1 – membership functions of the source current. For a color version of this figure, please see www.iste.co.uk/robyns/energy.zip
Figure 2.24. Level 1 – membership functions of the DC bus voltage
Figure 2.25. Level 1 – membership functions of the storage state of charge
Figure 2.26. Level 1 – membership functions of the reference power of the hybrid system
Figure 2.27. Level 2 – membership functions of the DC bus voltage
Figure 2.28. Level 2 – membership functions of the DC bus voltage derivative
Figure 2.29. Level 2 – membership functions of the distribution coefficient
Figure 2.30. General operational graph
Figure 2.31. Operational graph of level N1.1
Figure 2.32. Operational graph of level N1.1.1
Figure 2.33. Operational graph of level N1.1.2
Figure 2.34. Operational graph of level N1.1.3
Figure 2.35. Operational graph of level N1.2
Figure 2.36. Operational graph of level N1.2.1
Figure 2.37. Operational graph of level N1.2.2
Figure 2.38. Operational graph of level N1.2.3
Figure 2.39. Operational graph of level N2
Figure 2.40. Operational graph of level N2.1
Figure 2.41. Operational graph of level N2.2
Figure 2.42. Operational graph of level N2.3
Figure 2.43. Surface generated for level N1.2 when V_bus is 270 V. For a color version of this figure, please see www.iste.co.uk/robyns/energy.zip
Figure 2.44. Surface generated for level N2. For a color version of this figure, please see www.iste.co.uk/robyns/energy.zip
Figure 2.45. Simplified surface obtained using discretization (100 points. For a color version of this figure, please see www.iste.co.uk/robyns/energy.zip
Figure 2.46. Simplified surface obtained using discretization (10 points. For a color version of this figure, please see www.iste.co.uk/robyns/energy.zip
Figure 2.47. Testbed including the hybrid storage and dissipation system
Figure 2.48. Power profile (scenario) of the bidirectional charge
Figure 2.49. Definition of the variables to be observed
Figure 2.50. DC bus voltage (completely fuzzy-based supervisor, SOC_init = 20%). For a color version of this figure, please see www.iste.co.uk/robyns/energy.zip
Figure 2.51. Power levels of the DC bus (completely fuzzy-based supervisor, SOC_init = 20%). For a color version of this figure, please see www.iste.co.uk/robyns/energy.zip
Figure 2.52. Power levels of the hybrid system (completely fuzzy-based supervisor, SOC_init = 20%). For a color version of this figure, please see www.iste.co.uk/robyns/energy.zip
Figure 2.53. Storage state of charge (completely fuzzy-based supervisor, SOC_init = 20%)
Figure 2.54. DC bus voltage (non-fuzzy supervisor, SOC_init = 20%). For a color version of this figure, please see www.iste.co.uk/robyns/energy.zip
Figure 2.55. DC bus power levels (non-fuzzy supervisor, SOC_init = 20%). For a color version of this figure, please see www.iste.co.uk/robyns/energy.zip
Figure 2.56. Hybrid system power levels (non-fuzzy supervisor, SOC_init = 20%). For a color version of this figure, please see www.iste.co.uk/robyns/energy.zip
Figure 2.57. Storage state of charge (non-fuzzy supervisor, SOC_init = 20%)
Figure 2.58. Indicators of energy efficiency and voltage maintenance resulting from testing
Figure 2.59. Parameters to be optimized
Figure 2.60. a) Membership functions of the variables “HIGH” and “LOW”, Boolean case. For a color version of this figure, please see www.iste.co.uk/robyns/energy.zip
Figure 2.60. b) Membership functions of the variables “HIGH” and “LOW”, Intermediate case. For a color version of this figure, please see www.iste.co.uk/robyns/energy.zip
Figure 2.60. c) Membership functions of the variables “HIGH” and “LOW”, Completely fuzzy-based case. For a color version of this figure, please see www.iste.co.uk/robyns/energy.zip
Figure 2.61. Supervisor structure
Figure 2.62. Level 1 – membership functions of P_gen – parameter “d”. For a color version of this figure, please see www.iste.co.uk/robyns/energy.zip
Figure 2.63. Level 1 – membership functions of V_DC – parameter “a”. For a color version of this figure, please see www.iste.co.uk/robyns/energy.zip
Figure 2.64. Level 1 – membership functions of SOC – parameter “e”. For a color version of this figure, please see www.iste.co.uk/robyns/energy.zip
Figure 2.65. Level 1 – membership functions of P_ref – parameter “f”. For a color version of this figure, please see www.iste.co.uk/robyns/energy.zip
Figure 2.66. Level 2 – membership functions of dV/dt – parameter “b”. For a color version of this figure, please see www.iste.co.uk/robyns/energy.zip
Figure 2.67. Level 2 – membership functions of k – parameter “c”. For a color version of this figure, please see www.iste.co.uk/robyns/energy.zip
Figure 2.68. Charging profile
Figure 2.69. Evolution of the “loss” function value Objfun
Figure 2.70. V_DC voltage in the case of optimized and empirical supervisors (SOC 20%)
Figure 2.71. Storage power in the case of optimized and empirical supervisors (SOC 20%)
Figure 2.72. V_DC voltage in the case of optimized and empirical supervisors (SOC 50%)
Figure 2.73. Storage power in the case of optimized and empirical supervisors (SOC 50%)
Figure 2.74. V_DC voltage in the case of optimized and empirical supervisors (SOC 80%)
Figure 2.75. Storage power in the case of optimized and empirical supervisors (SOC 80%)

3 Electric and Hybrid Vehicles

Figure 3.1. Estimation of the well-to-wheel CO₂ emissions of FLUENCE, calculated based on the standard European cycle of the NEDC type (new European driving cycle created to reproduce the conditions encountered on European roads) [BAS 13]
Figure 3.2. Specific power and energy values of batteries (Johnson Control – SAFT 2007)
Figure 3.3. Projected increase in the number of charging infrastructures and EVs in France (ERDF)
Figure 3.4. Temporal charging profile (voltage, current and state of charge) of a Li-ion battery cell [ROU 09]
Figure 3.5. Temporal charging profile (power and state of charge) of the “Nissan Altra” EV (Li-ion battery) [QIA 10]
Figure 3.6. Simplified models of the charging profiles of an EV for different power values [BOU 15]
Figure 3.7. EV/PHEV development scenario (http://www.voiture-electrique-populaire.fr/actualites/analyse-persperctives-vehicule-electrique-france-022)
Figure 3.8. Normal distribution and cumulative function for the variable home “arrival time” [BOU 15]
Figure 3.9. Configuration of the electric system [BOU 15]
Figure 3.10. Profile of consumption, wind and photovoltaic energy production and charge of EVs over the course of 1 week [BOU 15]
Figure 3.11. Structure of the supervision strategy [BOU 15]
Figure 3.12. Functional graph [BOU 15]
Figure 3.13. FM system configuration [BOU 15]
Figure 3.14. BM system configuration [BOU 15]
Figure 3.15. Characteristic parameters of the membership functions of variables P_SS a), B_S b) and P_{ref_SM} c) [BOU 15]. For a color version of this figure, please see www.iste.co.uk/robyns/energy.zip
Figure 3.16. Operational graph [BOU 15]
Figure 3.17. Characteristic surface of the fuzzy system [BOU 15]. For a color version of this figure, please see www.iste.co.uk/robyns/energy.zip
Figure 3.18. Example of an overlap between FM and P_{ref_FM} [BOU 15]. For a color version of this figure, please see www.iste.co.uk/robyns/energy.zip
Figure 3.19. Power at the SS: without EV, EV with and without supervision [BOU 15]. For a color version of this figure, please see www.iste.co.uk/robyns/energy.zip
Figure 3.20. Evolution of the energy levels stored in the EVs with and without supervision [BOU 15]. For a color version of this figure, please see www.iste.co.uk/robyns/energy.zip
Figure 3.21. Charging profiles of the EVs with and without supervision [BOU 15]. For a color version of this figure, please see www.iste.co.uk/robyns/energy.zip
Figure 3.22. Series hybrid configuration [ROY 14]
Figure 3.23. Parallel hybrid configuration [ROY 14]
Figure 3.24. Series–parallel hybrid Toyota Prius configuration [NIS 16]
Figure 3.25. Different membership functions representing an average speed: a) Boolean logic, b) type-1 fuzzy logic, c) type-2 fuzzy logic
Figure 3.26. Configuration of a type-1 fuzzy system [SOL 12b]
Figure 3.27. Configuration of a type-2 fuzzy system [SOL 12b]
Figure 3.28. Primary and secondary membership functions [SOL 12b]
Figure 3.29. An IT2-MF membership function discretized in p type-1 membership functions [SOL 12b]
Figure 3.30. Illustrative example of the centroid calculation
Figure 3.31. Illustrative example of the inference calculation
Figure 3.32. ECCE vehicle and studied configuration
Figure 3.33. Schematic representation of the energy management function onboard the ECCE vehicle
Figure 3.34. Energy management principle using type-2 fuzzy logic
Figure 3.35. Type-2 membership functions: a) membership functions (type 2) for the state of charge variation of the supercapacitors; b) membership functions (type 2) for the difference in power dP; c) output membership functions (type 2) for the power variation required at the level of the fuel cell system. For a color version of this figure, please see www.iste.co.uk/robyns/energy.zip
Figure 3.36. ECCE running tests using the fuel cell/SCap/Batteries configuration: a) speed profile; b) reference power, battery, supercapacitor and fuel cell profiles; c) DC bus voltage profile. For a color version of this figure, please see www.iste.co.uk/robyns/energy.zip

4 Railway System: Diesel-Electric Hybrid Power Train

Figure 4.1. Specific power and speed of railway rolling stock applications [JEU 13]
Figure 4.2. Railway electric traction system
Figure 4.3. Examples of series hybrid configurations
Figure 4.4. Original BB63000 diesel locomotive and PLATHEE hybrid locomotive
Figure 4.5. “Macroscopic” view of the PLATHEE configuration to be designed
Figure 4.6. Energy management strategies by means of power supply filtering: a) amplitude limitation and b) frequency dimensioning
Figure 4.7. Qualitative example of a power assignment: traction phase and battery charge
Figure 4.8. Characterization of sources by projecting the Ragone plane on the frequency axis
Figure 4.9. Implementation of the double-hybridization frequency management (electrochemical battery–supercapacitors)
Figure 4.10. Division of assignment parts in the double-hybridization frequency management strategy
Figure 4.11. Example of conventional charging power profiles for the locomotive (traction + auxiliaries): a) yard assignment and b) local service road assignment
Figure 4.12. Two qualitative examples of assignments characterized by different PHP values
Figure 4.13. Qualitative examples of two different assignments despite having identical PHP values
Figure 4.14. Examples of assignment classifications pertaining to the three railway systems
Figure 4.15. Synthesis examples of a compact dimensioning (15 hours) and representative assignment for a set of 10 assignments with a duration of 4 hours [JAA 13]
Figure 4.16. Sequential design process: (top) general process and (bottom) particular case of the synthesis process of PLATHEE equipment
Figure 4.17. Representation of the variable granularity level and associated CPU time cost
Figure 4.18. Power flow models of the diesel generator a), the accumulator b) and the supercapacitor c)
Figure 4.19. Specific dimensionless consumption with reference to the 215 kW generator
Figure 4.20. Power flow model of the management strategy
Figure 4.21. Dimensional synthesis of the supercapacitor pack according to the assignment and the filtering frequency used to determine the contribution of the pack
Figure 4.22. Calculation of required energy using simple (top) and saturated (bottom) integration
Figure 4.23. Dimensioning of the BT accumulator according to the assignment to be completed and as a function of the dimensions of the DG (at given SC dimensions, for F_SC = 6.1 mHz). For a color version of this figure, please see www.iste.co.uk/robyns/energy.zip
Figure 4.24. Reduced energy impact of the SC (in this case represented by F_SC) on the BT dimensioning
Figure 4.25. Example of a behavioral analysis of the DG and BT during m₁₁
Figure 4.26. Illustration of the temporal degradation constraint
Figure 4.27. Influential factor on the dimensioning of the engine generator and battery with respect to the cycle use of the battery
Figure 4.28. Trade-off between diesel fuel consumption and investment cost
Figure 4.29. Selected configuration for the PLATHEE locomotive
Figure 4.30. Electrical configuration of the PLATHEE demonstrator
Figure 4.31. Pareto front (multi-criteria optimization) for the yard assignment section
Figure 4.32. Block diagram of the power HIL simulation of a pantograph–catenary system [BRU 11]
Figure 4.33. PLATHEE hybrid locomotive during the assembly phase
Figure 4.34. Configuration of the signal HIL simulator of the hybrid locomotive
Figure 4.35. Equivalent electrical circuit of the DC bus
Figure 4.36. Equivalent electrical circuit of the supercapacitor module
Figure 4.37. Equivalent wiring diagram of the electrochemical battery
Figure 4.38. Diesel combustion engine mechanically coupled to an excitation three-phase synchronous motor and a diode rectifier supplying the DC bus [MAY 14]
Figure 4.39. Experimental mapping of diesel engine consumption [MAY 14]. For a color version of this figure, please see www.iste.co.uk/robyns/energy.zip
Figure 4.40. Examples of measured values during an assignment: temporal profiles of train speed and charging power
Figure 4.41. HIL test bench of the hybrid locomotive
Figure 4.42. PLATHEE testing
Figure 4.43. Experimental results of an assignment completed by the hybrid locomotive: a) charging power, b) engine generator power, c) electrochemical battery power, d) supercapacitor power
Figure 4.44. Speed curve for the tramway application
Figure 4.45. Speed curve for the suburban train application
Figure 4.46. Speed curve for the regional train application
Figure 4.47. Speed curve for the high-speed train application
Figure 4.48. Tramway power curve over the course of a cycle
Figure 4.49. Suburban train power curve over the course of a cycle
Figure 4.50. Regional train power curve over the course of a cycle
Figure 4.51. High-speed train power curve over the course of a cycle
Figure 4.52. Iso consumption and iso power curves according to the speed of the DG
Figure 4.53. Volume (m³) and purchase cost (in k€ not including full installation) of the DG as a function of its nominal power (kW)

5 Railway System: Hybrid Railway Power Substation

Figure 5.1. Block diagram of the current railway electrification system
Figure 5.2. Configurations of the main types of power trains: a) electric, b) diesel-electric and c) bi-mode
Figure 5.3. Economic balance sheet for the electrification of a line
Figure 5.4. Hybrid railway power substation
Figure 5.5. Equivalent diagram of electric traction for railway applications
Figure 5.6. DC electrification diagram
Figure 5.7. AC electrification diagram at industrial frequency
Figure 5.8. Load curve measured at the level of a DC substation
Figure 5.9. Interconnection of the HRPS by means of the direct current bus (DC) at 1,500 V DC
Figure 5.10. Interconnection of the HRPS by means of the direct current bus (DC) at 25 kV/50 Hz AC
Figure 5.11. Illustration of the smoothing procedure of peak transit services at the level of a DC substation
Figure 5.12. Stationary flywheel application of the Rennes underground train [BOI 10]
Figure 5.13. Energy recovery techniques in the DC railway power grid [PEC 14]
Figure 5.14. a) PV systems in a railway station; b) generic configuration for the installation of PV systems in a station [FAR 07]
Figure 5.15. 200 kW PV system installed in the Takasaki station [HAY 11]
Figure 5.16. Energy flow at the level of the studied HRPS
Figure 5.17. HRPS energy supervision structure [BUZ 15]
Figure 5.18. Main functional graph [BUZ 15]
Figure 5.19. Detailed functional graph for mode N1 [BUZ 15]
Figure 5.20. Detailed functional graph for mode N2 [BUZ 15]
Figure 5.21. Detailed functional graph for mode N3 [BUZ 15]
Figure 5.22. Membership functions of the supervisor inputs and outputs [BUZ 15]
Figure 5.23. Main operational graph [BUZ 15]
Figure 5.24. Detailed operational graph for mode N1 [BUZ 15]
Figure 5.25. Detailed operational graph for mode N2 [BUZ 15]
Figure 5.26. Detailed operational graph for mode N3 [BUZ 15]
Figure 5.27. Consumption and production profiles of the considered HRPS [BUZ 15]
Figure 5.28. Forecast storage profile and average hourly tariff of the electrical energy purchased [BUZ 15]
Figure 5.29. Storage model
Figure 5.30. Simulation results obtained with the supervisor combining the forecast and real-time modes (LMT and ST) [BUZ 15]. For a color version of this figure, please see www.iste.co.uk/robyns/energy.zip
Figure 5.31. Optimization variables of the supervisor’s input parameters [BUZ 15]. For a color version of this figure, please see www.iste.co.uk/robyns/energy.zip
Figure 5.32. Optimization variables of the supervisor’s output [BUZ 15]
Figure 5.33. Methodology for the optimization of the parameter supervisor [BUZ 15]
Figure 5.34. Experimental design results [BUZ 15] (see the definition of the optimization variables in Table 5.7)
Figure 5.35. Empirical (solid lines) and optimum (dashed lines) membership functions – OF min [BUZ 15]. For a color version of this figure, please see www.iste.co.uk/robyns/energy.zip
Figure 5.36. Simulation results generated by the optimum solution [BUZ 15]
Figure 5.37. General configuration of the HRPS test platform [BUZ 15]
Figure 5.38. General configuration of the HRPS test platform [BUZ 15]
Figure 5.39. Base renewable energy profiles and energetically equivalent profiles (1–3) [BUZ 15]
Figure 5.40. Base charge profiles and energetically equivalent profiles (1–3) [BUZ 15]
Figure 5.41. Influence of renewable energy production on the CMDPS [BUZ 15]
Figure 5.42. Influence of the variation in renewable energy production on energy costs, consistently for the base profile and profiles 1–3 [BUZ 15]. For a color version of this figure, please see www.iste.co.uk/robyns/energy.zip
Figure 5.43. Influence of the variation in renewable energy production on the final SOC, consistently for the base profile and profiles 1–3 [BUZ 15]. For a color version of this figure, please see www.iste.co.uk/robyns/energy.zip
Figure 5.44. Influence of the variation in the charge profile on the energy cost, consistently for the base profile and the charge profiles 1–3 [BUZ 15]. For a color version of this figure, please see www.iste.co.uk/robyns/energy.zip
Figure 5.45. Influence of the variation in the charge profile on the renewable energy private consumption rate, consistently for the base profile and the charge profiles 1–3 [BUZ 15].
Figure 5.46. Influence of the variation in the charge profile on the final SOC, consistently for the base profile and the charge profiles 1–3 [BUZ 15].
Figure 5.47. Possible energy exchanges at the level of “a railway smart grid” [BUZ 15]

Foreword

In this second book, called Electrical Energy Storage in Transportation Systems, Professor Robyns and his coauthors accomplish the aim they set upon from the beginning of their project, that is to present the reader with rigorous methodological approaches based on the concept of energy management supervisors for complex systems, including different sources and stock elements.

As indicated in the title, this new book is dedicated to electric transport, which is of particular relevance because of its broad applications – such as onboard networks in aeronautics, the integration of electric vehicles in the electric network, hybrid vehicles, or even hybrid railway traction and its installations. We have already highlighted, in our previous preface, the importance for our society of offering alternative solutions to currently available energy systems based on fossil or nuclear energy, via transitory solutions such as the hybrid vehicle. However, numerous scientific challenges still remain, the two principal ones being the limited performance of batteries and the difficulty to manage very complex systems in real time. This book particularly aims to provide an answer to this challenge. We can admit that the bet has been successful.

Indeed, pursuing the themes of the previous book, dealing with the management of energy production systems based on renewable sources and stock units, the authors deepen their methodology of fuzzy logic supervision. The presentation that they provide is highly pedagogic, and we follow with interest all the stages that lead to the elaboration of supervision (i.e. drafting the specifications, defining entry and exit variables, conceiving functional graphs corresponding to all necessary hierarchical levels, fuzzification of the problem and optimizing supervisor structures with the aid of genetic algorithms and experimental design).

In this book, apart from implementing the structured methodology based on the design of an energy management supervisor developed at the Laboratoire d'Electrotechnique et d'Electronique de Puissance of Lille, other complementary methods developed at Belfort and Toulouse Laboratories are equally explored. They concern fuzzy logics of type 2, filtering methods and explicit optimization.

Finally, they systematically propose an experimental validation of the studied structures, and this is not a minor contribution in the exposed works.

Thus, happily matching industrial theory and practice, this book will become an indispensable reference for all engineers and researchers working in the field of electric energy management of onboard systems.

Eric MONMASSON
June 2016

Introduction

At the end of the 19th Century, electrical energy was used in railway and road transport; the experimental electric vehicle "La Jamais Contente" (The Never Satisfied) exceeded 100 km/h in 1899. However, the difficulty of storing electrical energy in sufficient quantities, within reasonable volume and weight limits, represented one of the major obstacles in the development of autonomous electric vehicles that are able to travel medium- to long distances. At present, the development of renewable energy sources and the demand for low-carbon modes of transport are generating renewed interest in the storage of electrical energy, which becomes a key element for sustainable development. From this moment on, modern storage technologies make it possible to envisage the development of electric vehicles with acceptable performance levels, more efficient electrification of aircraft, the development of hybrid autonomous vehicles and locomotives, but also using storage to improve energy efficiency and to secure the supply of electrical transport systems. The aim of this book is to contribute to a better knowledge and understanding of these developing technologies within the framework of transport systems and more particularly with regard to their management and operation.

The aims of this book are to:

– highlight the importance of storing electrical energy in accordance with the principles of sustainable development in transport, in the context of deploying smart electric power grids or “smart grids”, grids with which a certain number of transport systems will interact to an increasing extent, such as electric vehicles, plug-in hybrids, trains, underground trains, trams and electric buses;
– present the variety of services provided with respect to the storage of electrical energy;
– present methodological tools that make it possible to build an energy storage management system following a generic and pedagogical approach. These tools rely on artificial intelligence and explicit optimization methods. They are presented throughout the book with respect to practical case studies;
– illustrate these methodological approaches using several practical and pedagogical examples regarding the electrification of transport units and their integration into electric power grids, in specific cases, with respect to the production of electrical energy from variable renewable energy sources.

The first chapter formulates the issues of storing electrical energy in transport systems. The storage requirements of these applications are highlighted, along with their numerous contributions. A design methodology for storage system management, relying on artificial intelligence, is introduced; it is particularly well adapted for the management of complex systems involving uncertainties related to the forecast of the production of variable renewable energy, the consumption induced by the trajectory or the power profile of the vehicle or aircraft, but also of the electrical grid when the system in question is connected to it. This methodology serves several objectives requiring real-time processing.

The second chapter presents the integration of electrical energy storage into aeronautic onboard grids. The increase in the number of electric charges as well as the gradual substitution of the actuators, originally hydrostatic or mechanical, by electro-hydrostatic or electro-mechanical actuators, are the main causes of the increased electrification of aircraft. The onboard electric power grids developed alternative solutions with fixed and variable frequency, and configurations of direct current grids (local or distributed) for the exchange of energy including storage are developed. Direct current grids, including storage, facilitate bidirectional electrical power flows, making it possible to recover the braking energy of the actuators, reduce the number of electronic power converters and the cable diameter between the main electrical grid and the actuators, thus allowing for gains in volume and mass. They also provide the possibility of increasing the reliability of these grids owing to the use of storage as a local emergency power supply. A structured methodology for the development of an energetic supervisor in real time, based on fuzzy logic, has been applied to the management of energy in a local energy exchange direct current grid, from the creation of a list of functional specifications (objectives and constraints) to the optimization stage of the supervisor parameters. A comparison is made between supervision strategies that use fuzzy logic only and solutions that do not resort to it (using, for example, a PI controller) together with combined solutions. Implementation on an experimental basis in real time is also addressed.

The third chapter refers to autonomous road vehicles. The first part refers to the charge management of electric vehicles, so as to incorporate them into the electric power grids harmoniously and to give priority to a charge using renewable energy as a guarantee of low environmental impacts. The design methodology of a supervisor based on fuzzy logic is implemented until the optimization of the parameter supervisor. The prospect of a more active contribution of electric vehicles to electric power grids (Vehicle to Grid and Vehicle to Home) is also addressed. The final part of this chapter provides an overview of various configurations of hybrid power trains which are implemented practically. The management of a hybrid vehicle, comprising electrochemical batteries, supercapacitors and a fuel cell, is then developed using fuzzy logic. A variant of fuzzy logic, referred to as type-2 fuzzy logic, which involves the uncertainty related to the determination of membership functions, is implemented.

The fourth chapter addresses hybrid electric traction for railway applications. The first part of the chapter, dedicated to rolling stocks, provides a detailed description of storage systems for the diesel hybrid traction unit, which includes two onboard technologies for storage systems: electrochemical batteries and supercapacitors. The energy management of these systems is carried out by combining digital filtering and explicit optimization methods. Experimental implementation on a real locomotive is also described. The second part introduces, in a pedagogical manner, practices dedicated to using onboard storage systems for electrical traction. The analysis is based on the kinematic profile to relate back to the basic energy requirements of the railway system. Starting from this description, the set of relevant applications of the energy storage systems is presented in more detail. In this part, numerical applications derived from real cases are presented to illustrate the scope and the energy issues of onboard energy storage systems in the railway sector.

The integration of systems for producing variable renewable energy (photovoltaic and/or wind) and for storing the energy directly in the feed system of the railway system introduces the notion of the railway “smart grid”. The fifth chapter describes this evolution, along with an analysis of the services that can be provided by the new hybrid railway power substations (HRPS) which supply the railway network. Reference is made to the services provided to the railway system itself, but also to electric power grids conducting and distributing the electrical energy and to local producers of renewable energy. A two-stage energy management process for an HRPS is then developed, with a first forecast stage (long-term management) and a different one in real time (short-term management), making it possible to adapt to fluctuations, as well as uncertainties in predicting the production of renewable energy, but also to deviations in the charge profile represented by the movement of trains. The supervision stage in real time is constituted following the structured methodology based on an artificial intelligence tool, namely fuzzy logic. The parameters of the supervisor are optimized and a sensitivity study, within an experimental platform in laboratories, makes it possible to evaluate the robustness of the supervisor. Finally, the development prospects for railway smart grids conclude this chapter.