Fine-scale volatility-based economic dispatch for smart grid edge networks

The volatility of power delivery from renewable energy sources is currently limiting the scaling up of distributionlevel generation such as rooftop PV solar. Because volatility occurs at short time scales, its management poses challenges in coordination of the dispatch of load events. In this paper, we introduce an algorithm and supporting architecture for the control of volatility at fine time scales. The algorithm dynamically adjusts parameters of load events requested by distributed energy resources, and the architecture and processing can be implemented via integration of presently available technologies. We also introduce a unified cost measure that incorporates economic valuation of both volatility and energy as a function of time-varying pricing policies. This cost measure leads to a constrained non-linear optimization problem, and we propose a genetic algorithm to shape and dispatch load events using the cost measure as a fitness function. Numerical results demonstrate the efficacy of the approach, reveal the trade-off between volatility and energy cost under time-varying pricing policies, and show that increasing the scale of coordinated aggregation can improve performance.