The optimization task in many virtual power plant (VPP) scenarios comprises the search for appropriate schedules spaces from distributed energy resources. In with a decoupling of modeling and control, these are as well. If merely controller unit knows about subset operable that allowed to be considered by central scheduling unit, then sets have effectively communicated. We discuss an approach learning envelope separates non-operable inside space all means support vector data description....

The increasing pervasion of information and communication technology (ICT) in energy systems allows for the development new control concepts on all voltage levels.In distribution grid, this is accompanied by a still penetration with distributed resources like photovoltaic (PV) plants, wind turbines or small scale combined heat power (CHP) plants.Combined shiftable loads electrical storage, these units set up flexibility potential grid that can be tapped ICT-based following long-term goal...

A steadily growing pervasion of the energy distribution grid with communication technology is widely seen as an enabler for new computational coordination techniques renewable, distributed generation well bundling controllable consumers. Smart markets will foster a decentralized management. One important task prerequisite to management ability group together in order jointly gain enough suitable flexibility and capacity assume responsibility specific control grid. In self-organized smart...

In many virtual power plant (VPP) scenarios, numerous individually configured units within a VPP have to be scheduled regarding both global constraints (i.e. external market demands) and local technical, economical or ecological aspects for each unit). Approaches constraint handling been discussed in the relevant literature independently. A hybrid approach is proposed that combines decentralized combinatorial optimization heuristic with encoding of constrained search spaces into...

A new application for support vector machines is their use meta-modeling feasible regions in constrained optimization problems. We here describe a solution the still unsolved problem of standardized integration such models into (evolutionary) algorithms with help decoder based approach. This goal achieved by constructing mapping function that maps whole unconstrained domain given to region solutions model. The applicability real world problems demonstrated using load balancing from smart grid domain.

The sets of feasible load schedules that distributed energy resources are able to operate, jointly define the search space within many virtual power plant optimization tasks. If a centralized approach is considered, central, single scheduling unit needs know for each resource what comply with all given constraints, because only these operable and might be taken into account optimization. As constraints depend on state or time, currently alternatives have repeatedly communicated scheduler in...

The current upheaval in the electricity sector demands distributed generation schemes that take into account individually configured energy units and new grid structures. At same time, this change is heading for a paradigm shift controlling these resources within grid. Pro-active scheduling of active power (from perspective) loosely coupled group planning optimization methods individual feasible region local search spaces modeled by surrogate models. We propose method uses support vector...

Following the long-term goal of substituting conventional power generation, market oriented approaches will lead to interaction, competition but also collaboration between different units. Together with expected huge number actors, this in turn a need for self-organized and distributed control structures. Virtual plants are an established idea organizing generation. A frequently arising task is solving scheduling problem that assigns operation schedule each energy resource taking into...

In this contribution we present an approach on how to include local soft constraints in the fully distributed algorithm COHDA for task of energy units scheduling virtual power plants (VPP).We show a flexibility representation based surrogate models is extended and trained using like avoiding frequent cold starts combined heat plants.During scheduling, agents representing these machines indicators their choice new operation schedule.Using example VPP that our enables reflect without...

Classification of high-dimensional data with imbalanced classes poses problems. Especially such time series classification tasks are problematic, because the ordering each step (feature) is important and therefore dimensionality reduction feature selection cannot be applied. The cascade model was developed for tasks. classifier splits into a low-dimensional But can only handle sets structure that easily learned in space. In this paper, we propose generalized version also deal more complex...

Following the long-term goal of substituting conventional power generation with cleaner energy will lead to an integration a large share small units imposing problem sizes for coordination.The expected huge number entities leads need new techniques reducing computational effort coordination.Predictive scheduling is frequent task in grid control.For resources, schedules have be found that fulfill several objectives at same time.Considering day-ahead scenarios 96-dimensional imposes additional...