Visualization and Decision Support
From Grid Optics
As the scale and complexity of the power grid increase, so does the need for human analysts to make effective tactical and strategic grid management decisions. As renewables are integrated into the power supply and demand-side management becomes possible, operators will be challenged to make decisions that best balance generation options, load management, storage type and location, and environmental impact, among other criteria.
The Initiative's Visualization and Decision Support research concentrates on the development of capabilities that convert large-amounts of data into knowledge and actionable information to support operation, planning, and policy decisions. This research area is combining multiple, coordinated visualization interfaces and online aids to decision making in a modular, extensible software environment that can be used for both real-time grid operations as well as long-term planning. To this end, it will incorporate visual interfaces to the output of modeling and simulation capabilities developed in our modeling, simulation, and analysis research. We are pursuing research in four areas:
- human factors
- visualization development
- automated reasoning and
- interoperability specifications.
Developing new operations and planning interfaces for the power grid demands that we first understand and document the decision making environment of our anticipated users. Our studies are enumerating the classes of decisions that must be made (what types of decisions and over what time scales) as well as the input to those decisions (e.g., data or models). We are also performing post mortem analyses of both typical and emergency decisions to understand where automated support is needed.
To achieve meaningful visual representations across temporal and spatial scales, new visual metaphors and aggregates must be developed. Metaphors like the traditional line diagram representation of the power system often used in control rooms do not account for geography (the location and proximity of particular buses, lines, transformers, and so on) or for external influences on grid conditions. Because of the modular and multi-scale nature, users will be able to choose from alternate visual representations for a given data source, perhaps selecting an abstract representation for contextual information and more detailed representations for local data.
Future power grid operation will require coordinated decision making among humans in the context of the power systems. Based on the assessment of automated aids to decision making carried out in the formative human factors activities, cognitive assistants that incorporate new mathematical and statistical models of organizations’ decision-making processes and grid operators’ behaviors will be developed. Such systems can assist human operators and planners in reasoning with uncertainty in both raw sensor measurement and derived data.
The developed capabilities will be based on a software integration framework that allows multiple visualization interfaces to interoperate over the same data streams in a coordinated fashion. Engaging representatives from other national laboratories, universities, and commercial partners, we will hold a series of grid visualization workshops to articulate the basic requirements for interoperable grid analysis software. We will work toward a community agreement on protocols for linking grid visualization and analysis components.
List of Projects: