Evolutionäre Algorithmen

Weicker, Karsten

doi:10.1007/978-3-658-09958-9

Cited by 35 publications

(26 citation statements)

References 85 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…For the parameters used in the algorithm, a standard suggestion is used [14], advising that the population size should lie between 1 and 30 solutions, and that the best selection for the quotient between parent solutions and number of solutions in the evaluation of the fitness has a value between 1 5 and 1 7 . In our case this leads to a start population of 20 solutions, and in each evolution process 40 child individuals are generated via recombination and another 40 child individuals are generated via mutation.…”

Section: Discussionmentioning

confidence: 99%

Charging optimization of enterprise electric vehicles for participation in demand response

Detzler

Eichhorn

Karnouskos

2015

2015 International Symposium on Smart Electric Distribution Systems and Technologies (EDST)

View full text Add to dashboard Cite

Abstract-In the emerging smart grid, using flexible infrastructures to realize better energy management via demand response scenarios is at its core. The potential of electric vehicles used to realize such flexibility is widely experimented with. In this work, we take a closer look at a specific case of enterpriseowned electric vehicles, parked at enterprise premises, and how their charging can be optimized in order to both adhere to the enterprise operational constraints, as well as consider dynamic changes stemming from other grid stakeholders. A simplified optimization using an evolutionary algorithm is realized, and the approach is evaluated under two scenarios of interest. Of specific interest to the many smart grid stakeholders are the DR scenarios [4], as these can yield additional (usually monetary) benefits to the involved stakeholders, while in parallel tackling key problems in the grid due to highly dynamic energy production stemming from Renewable Energy Sources (RES). The role of the Electric Vehicles (EVs) in smart grid is increasingly investigated [6], including their utilization as dynamic storage [7], since, if a critical mass of them is reached, they can have a significant energy impact on existing infrastructure, future planning and naturally in any energy optimization scenario. Index Terms-EnergyEVs can be an active participant in DR, since they provide flexibility during longer standing times. This is especially of interest when larger numbers of them are available, which is the case e.g., for EV fleets. Several uncertainties are coupled with individual EVs including, their presence, the authorization to centrally control charging, the acceptance by the consumer (EV owner), the impact on the EV battery, etc. However, many of these considerations, can be set aside in specific cases, such as those involving enterprise-owned cars.In this work we focus on this area and introduce two DR scenarios (a price-based and an incentive-based one) that are attractive for enterprise fleets of electric vehicles. The latter can react to DR events, with the flexibility given by long and predictable parking times without interfering with their operational plan. We introduce an optimization approach that allows operators of such EV fleets to react to two different types of DR events. Finally we evaluate these two scenarios and the optimization realized, with real world data both for available RES and enterprise EV fleet. II. DEMAND RESPONSE FOR DSO AND SUPPLIERSIn the context of enterprise-owned EVs, different motivations exist for shifting electrical loads over time. From a local point of view, such load shifts can help reducing consumption (due to EV charging) at times where electricity is expensive in order to reduce overall enterprise costs. In addition, EVs can also be used to prevent the overall power draw from exceeding technical or contractual limitations which could lead to physical damages or penalty fees. DR allows extending this concept from local boundary conditions to more global aspects, as due t...

show abstract

Section: Discussionmentioning

confidence: 99%

Charging optimization of enterprise electric vehicles for participation in demand response

Detzler

Eichhorn

Karnouskos

2015

2015 International Symposium on Smart Electric Distribution Systems and Technologies (EDST)

View full text Add to dashboard Cite

show abstract

“…Domain experts are able to define a set of planning rules and to embed this information into an Evolutionary Algorithm used here. Literature refers to memetic algorithms or hybrid evolutionary computation if standard evolutionary algorithms are combined with heuristic rules for guiding the search process [5,6]. Due to different options in combining standard evolutionary algorithms with heuristics there is a large variety of possible memetic approaches:…”

Section: Optimization Approachmentioning

confidence: 99%

Resilient and comprehensive networks planning is key to a successful integration of renewable energy resources

Jendernalik¹,

Dominique²,

Engels³

et al. 2016

CIRED Workshop 2016

View full text Add to dashboard Cite

The German "Energiewende" leads to significant uncertainty in terms of future demand and generation development. Distribution networks are mainly affected by the resulting different divergent requirements: Distribution networks are transformed into local transportation grids. Additionally network investments are restricted by regulation. The solution presented here combines strategic Asset Management (age & condition) and Network Development (capacity) in a synergetic fashion to raise significant technical & economical optimization potential. An adaption to uncertainty in the future is based on the analysis of a variety of load and feed-in scenarios. This approach enables optimal risk and cost aware decisions.

show abstract

“…For the intelligent charging plan algorithm we use an evolutionary algorithm whose general process [6] can shortly be described by these steps: 1) Select a start population of individuals randomly or partly randomly (first generation). 2) Analyze the fitness of each individual in that population.…”

Section: Algorithmmentioning

confidence: 99%

“…For the parameters used in the algorithm, a standard suggestion is used [6], advising that the population size lies between 1 and 30 solutions and the best selection for the quotient between parents solutions and number of solutions in the evaluation of the fitness has a value between . In this case this leads to a start population of 20 solutions and in each evolution process 40 child individuals are generated with recombination and 40 with mutation.…”

Section: B Assessmentmentioning

confidence: 99%

A model and an evolutionary algorithmic approach towards optimization of Electric Vehicle fleet charging

Detzler

Karnouskos

2015

2015 International Symposium on Smart Electric Distribution Systems and Technologies (EDST)

View full text Add to dashboard Cite

Abstract-The prevalence of the Smart Grid and its capabilities, has enabled sophisticated energy management that can be realized as a multi-constraint optimization problem and tailored to the specific scenario needs. In conjunction with the increasing introduction of Electric Vehicles (EVs), energy management tools can now consider expanded conditions including grid balance, cost optimization, EV characteristics, asset utilization, operational goals etc. In this work we analyze such a scenario and demonstrate how an EV fleet charging can be optimized in a timely manner while taking into consideration local conditions e.g., individual EV needs as well as global ones e.g., grid limits and energy price. We formalize a model that reflects the EV restrictions, and use it to assess an algorithmic approach that solves this non-linear optimization problem.

show abstract

Evolutionäre Algorithmen

Cited by 35 publications

References 85 publications

Charging optimization of enterprise electric vehicles for participation in demand response

Charging optimization of enterprise electric vehicles for participation in demand response

Resilient and comprehensive networks planning is key to a successful integration of renewable energy resources

A model and an evolutionary algorithmic approach towards optimization of Electric Vehicle fleet charging

Contact Info

Product

Resources

About