Book Read Free

Digital Marketplaces Unleashed

Page 105

by Claudia Linnhoff-Popien


  (68.4)

  The constraints for the optimization are defined by: Every vehicle gets a recommended charging point with the matrix X.

  Only one single vehicle can charge at a charging point at the same time.

  Usually there are less charging points than vehicles in the market. Because of the last two constraints, every driver/vehicle will get at least one optimized recommended charging option. So they could get the recommendation to charge at the same charging point, because there are not enough charging points for all vehicles in the market. Out of those vehicles, only the vehicle with the shortest distance (matrix D) will realize the charging option. Others will have to wait.

  The optimization aims to reduce overall waiting times for all vehicles by recommending better charging points with an improved configuration of the matrix X, considering the maximal driving ranges of the individual vehicles, their states of charge and current positions.

  The general coordination problem provokes an exponential complexity of combinations. To solve the problem fast during real‐time traffic conditions, a heuristic approach was chosen. Within the mentioned constraints, the genetic algorithm will search for a configuration X which should lead to an improved situation of charging for all drivers with less waiting times and blocking between the vehicles on arrival at the charging points. The results are shown in the next section.

  68.5 Results of the Optimization

  To evaluate the optimization, vehicles and charging points were initialized with random positions. The capacity of the batteries of the vehicles was set to 20 kWh and the average consumption was 15.3 kWh / 100 km. This default configuration represents common vehicles with a typical driving range of about 130 km under real driving conditions. Fully charged vehicles are not looking for charging points and totally depleted vehicle are technically not able to reach any charging point.

  So the initial state of charge was initialized randomly between 10 and 50% of the maximal capacity of their batteries to generate directly a representative situation of competition for the simulation with no totally charged or depleted vehicles. Vehicles are (currently) driving by air‐line distance directly to the charging points. The implementation of real street data will follow in 2017 by using OpenStreetMap [15] and the traffic simulation framework MATsim [16].

  This article focusses on the general optimization of the coordination problem and how a genetic algorithm could be used for this optimization. An optimized configuration X will lead to less blocking between vehicles and will allow to deliver a higher amount of energy from the charging points to the requesting vehicles. So the indicator for measuring the performance of the optimized “vehicle‐to‐charge‐point assignment” is based on the delivered amount of energy per time to the vehicles.

  The results of the genetic algorithm [17] are compared with the default scenario. The delivered amount of energy in the default scenario defines the basic benchmark with 100%. In the default scenario, vehicles are only driving to the next charging point in range without using additional information and without any further optimization. Applying the genetic algorithm leads to the following results (see Table 68.1). Table 68.1Results of the optimization with the genetic algorithm

  Amount vehicles

  Amount charging points

  Performance “default situation” (in %)

  Performance “genetic algorithm” (in %)

  1

  100

   10

  100

  143.6

  2

  100

   25

  100

  152.5

  3

  100

   50

  100

  165.2

  4

  200

   20

  100

  141.8

  5

  200

   50

  100

  140.7

  6

  200

  100

  100

  162.9

  7

  300

   30

  100

  131.3

  8

  300

   75

  100

  151.2

  9

  300

  150

  100

  149.2

  The configuration of the genetic algorithm is based on standard values with a population size of 50 individuals and a selection of 5% of the best individuals per generation. The genetic crossover function is defined with 80%. The stopping criteria is 1% over 50 generations. So if 50 generations will not generate an individual which is at least 1% better than the last known best one, the genetic algorithm will stop. The general situation of competition amongst the vehicles is influenced by the availability of charging points in relation to the amount of requesting vehicles. A situation of low competition is defined with a sufficient amount of charging points with 50% of the number of vehicles, shown with line 3, 6 and 9 in Table 68.1. The genetic algorithm is able to gain on average a + 59% improvement in such situations with a low competition compared to the default situation without any optimization. A situation with a high competition is generated with a low amount of charging points with only 10% of the number of vehicles. This is shown with line 1, 4 and 7 in Table 68.1. The genetic algorithm is still able to gain on average a + 39% improvement in such situations with a high competition. The average performance of the genetic algorithm over all nine simulated scenarios results in an improvement of + 48.7% compared to the default scenario where drivers will only drive to the nearest charging point in range without using additional information or optimized coordination. The implemented genetic algorithm took about 294 generations per simulation to find an optimum within 13 s with common hardware of the year 2016 (4 × 3.5 GHz CPU, 8 GB RAM).

  68.6 Conclusion and Future Work

  Electric vehicles have to find unoccupied charging points and require long charging times. To reduce the blocking between vehicles at charging points, an optimization model is needed. The presented matrix model shows a solution for aggregating data in traffic scenarios with electric vehicles. Actual positions and states of charge of vehicles are combined with the locations of charging points in a uniform way. A matrix model was designed to be expandable by additional matrices. For example, individual preferences of drivers like preferred energy prices of charging points or preferences of a desired arrival time at specific charging points could be integrated easily with additional matrices. The integration of those additional matrices should lead to a more accurate and personal coordination of electric vehicles and is part of future work and could be used in related traffic management systems.

  Within this work, the matrix model has been applied to describe the general situation of competition between electric vehicles with the mentioned matrix Z. To optimize the coordination problem for electric vehicles in search for charging points, a genetic algorithm was utilized for several situations of competition.

  It has been shown that a genetic algorithm could be a suitable approach to optimize the general coordination problem. If every driver will cooperate and adapt to those recommended alternative charging options, there is a huge potential. The optimization could lead to an average performance of about + 48.7% of served energy to the electric vehicles, compared to the default driving scenario, where drivers are only driving to the nearest charging point in range without optimization.

  Under real driving conditions, this potential of optimization would be less, because not every driver will cooperate and adapt to those recommended and optimized alternative charging points. The effect of those uncooperative drivers will be examined in future work.

  References

  1.

  Bundesregierung, “Regierungsprogramm Elektromobilität,” 01 05 2011. [Online]. Available: https://​www.​bmbf.​de/​fi
les/​programm_​elektromobilitae​t.​pdf. [Accessed 11 01 2016].

  2.

  Clean Energy Ministerial, “Global EV Outlook,” OECD/IEA, OECD/IEA, Paris, France, 2013.

  3.

  Daimler AG, “Premiere: Apple “CarPlay” in der neuen C-Klasse: Mercedes-Benz bleibt Trendsetter bei Smartphone-Integration,” Stuttgart, Germany, 2014.

  4.

  Volkswagen AG, “VOLKSWAGEN CAR-NET - APPS UND DIENSTE IM ÜBERBLICK,” Wolfsburg, Germany, 2015.

  5.

  Audi AG, “Audi connect & Mobilität,” Ingolstadt, Germany, 2015.

  6.

  BMW AG, “Rede Harald Krüger, Vorsitzender des Vorstands der BMW AG,” Munich, Germany, 2015.

  7.

  M. Zuch, “Effizienz im elektromobilen Massenmarkt,” Lecture Notes in Informatics, Gesellschaft für Informatik, Cottbus, 2015.

  8.

  Muhammad Ismail, Ehab F. El-Saadany, Weihua Zhuang, Mostafa F. Shaaban, Real-Time PEV Charging-Discharging Coordination in Smart Distribution Systems, Waterloo, Canada: IEEE Transactions on Smart Grid (Volume: 5 , Issue: 4), 2014.

  9.

  Rutger Claes, Stijn Vandael, Niels Leemput, Tom Holvoet, Geert Deconinck, Kristof Coninx, Anticipatory Coordination of Electric Vehicle Allocation to Fast Charging Infrastructure, Leuven, Belgium: EnergyVille, Department of Electrical Engineering, 2014.

  10.

  R. Waraich, Agent-based simulation of electric vehicles: design and implementation of a framework, Zürich, Switzerland: ETH-Zürich, 2013.

  11.

  Y. Han, Y. Chen, F. Han and K. J. Ray Liu, “An optimal dynamic pricing and schedule approach in V2G,” IEEE, Maryland, USA, 2012.

  12.

  V. Larsson, Route Optimized Energy Management of Plug-in Hybrid Electric Vehicles, Göteborg, Sweden: Department of Signals and Systems - Chalmers University of Technology, 2014.

  13.

  M. N. Mariyasagayam and Y. Kobayashi, “Electric Vehicle Route Assistance Using Forecast on Charging Station,” in ENERGY 2013, The Third International Conference on Smart Grids, Green Communications and IT Energy-aware Technologies, Lisbon, Portugal, 2013.

  14.

  M. Hirzel, “Partition and Compose: Parallel Complex Event Processing,” DEBS ’12 Proceedings of the 6th ACM International Conference on Distributed Event-Based Systems, New York, USA, 2012.

  15.

  openstreetmap.de, “OpenStreetMap - Deutschland,” 11 10 2016. [Online]. Available: www.​openstreetmap.​de. [Accessed 11 10 2016].

  16.

  Technische Universität Berlin, Swiss Federal Institute of Technology Zurich, Senozon AG, “The Multi-Agent Transport Simulation,” 11 10 2016. [Online]. Available: http://​matsim.​org. [Accessed 11 10 2016].

  17.

  The MathWorks GmbH, “How the Genetic Algorithm Works,” 20 08 2016. [Online]. Available: https://​de.​mathworks.​com/​help/​gads/​how-the-genetic-algorithm-works.​html. [Accessed 20 08 2016].

  © Springer-Verlag GmbH Germany 2018

  Claudia Linnhoff-Popien, Ralf Schneider and Michael Zaddach (eds.)Digital Marketplaces Unleashedhttps://doi.org/10.1007/978-3-662-49275-8_69

  69. A Secure and Efficient Communication Tool

  Matteo Cagnazzo1 , Patrick Wegner1 and Norbert Pohlmann1

  (1)Institute for Internet-Security, Gelsenkirchen, Germany

  Matteo Cagnazzo (Corresponding author)

  Email: cagnazzo@internet-sicherheit.de

  Patrick Wegner

  Email: wegner@internet-sicherheit.de

  Norbert Pohlmann

  Email: pohlmann@internet-sicherheit.de

  69.1 Communication Means Sharing Information

  Nowadays information has an unprecedented importance. It has never been easier to share information with other entities. An information is based on characters that are connected via a syntax and grammar to form data. Setting this data into context is information. “Sharing information” creates knowledge and to share information one must communicate. There must be a sender and a receiver who communicate. It surrounds our everyday life more than ever, especially at work. Communication can reach different degrees of efficiency and organized in various ways. Efficiency can be determined by individual skills of participants as well as differences in the established communication structure and culture [1].

  Communication can be ranked in terms of relevance. Should this information be accessible by everyone or just a special group of people? Or is there one specified recipient for one specific information. Depending on this classification and the criticality of the transferred information a secured and trustworthy environment may be needed. Technologically speaking this means to establish a trust and security level through end to end encryption or at least transport layer encryption to secure the transmission. Adding such security measures will result in a complex system which can create a trustworthy connectedness of the participating entities.

  Corporate communication includes every communication process no matter if it is in‐ or outbound. This includes the sharing of information between employees as well as communication of the corporation to employees. Through communication an organization gains knowledge and if this knowledge is managed properly an organization can draw strategic decisions from it which gives them competitive advantages. That is why communication is such an integral part of modern businesses [2].

  69.2 Digital Communication in Organization Nowadays

  Through installed telecommunication systems e. g. mail client, social business network or telephone organizations connect their employees so that they can communicate. Nowadays there is a focus on e‐mails through which most of the communication inside of organizations happens, no matter how big or small. Worldwide the number of in‐ and outbound mails is 116 billion with an upward trend. On average this means that every employee deals with 123 mails per day [3]. The time that is expended daily to answer those mails is not negligible. If every mail is processed for just one minute, an employee is occupied for two hours on average. If it’s an executive this number increases and they are spending most of their time at work with communication.

  Optimization of such communication habits are already on the rise. Especially mailing is being questioned because of the excessive amount of mails being exchanged on a daily basis and therefore not being too efficient. Even though the decade old system works and is being used by a majority of people it is not capable of dealing with today’s technologies.

  Mailing is nowadays characterized by formalities formulated personally but they are not meant personal. Repeating salutation and farewell patterns, empty phrases, and disclaimers are generating a significant overhead. The information of a mail can be one sentence but due to the overhead the mail becomes long and extensive. If this mail becomes part of a collaborative exchange because of a document that a group of people is working on it results in poorly structured, swelling and never ending discussions. This offers a possibility to break with norms and structure communication more efficient, smart and modern.

  Apart from this way of transmitting information mailing is not capable of dealing with today’s security requirements and it is comparatively unsafe to write a mail. It is easy to manipulate whole messages or forge the address of sender or receiver [4, 5].

  There are possibilities to protect against such frauds for example end‐to‐end encryption with “Pretty Good Privacy, PGP” but they are not usable out of the box [6]. Due to the complexity and not usability by non‐professionals such solutions to the trust and security problems are not popular. There are ways to secure the transport layer e. g. STARTLS so that the user can be sure no one eavesdrops on their communication but it does not offer security on the different nodes through which a mail is routed [7]. The harm done by digital attacks against companies is estimated at $400 billion a year [8]. This clarifies the need for a suitable secure communication solution. There has to be an e
xchange of technology which offers new ways to share critical information in businesses.

  69.3 New and Efficient Communication Approaches

  New and modern approaches help to optimize the communication behavior in organizations and establish a culture of efficient communication. A robust foundation is necessary which is depicted by a modern, forward looking and secure architecture. An efficient, smart and secure communication tool is not a trivial task but it is a very complex challenge not just from a technological point of view. One of the main challenges is to offer the user new and efficient ways of sharing information which are intuitively and easily usable and on the other hand offer a reasonably high security level. Based on these more complex scenarios are implementable e. g. depicting business processes to digitize them and transform businesses.

  69.3.1 Transforming Old into New

  A first step towards an efficient and modern communication technique is already available but the idea behind it is not new. Instant messaging is existent for more than 40 years and gained popularity in the mid 1990’s [9]. Programs like ICQ and AOL messenger were the most prominent tools available. They enabled real time communication from sender to receiver no matter how many miles apart [10, 11].

 

‹ Prev