Distributed Load Management Using Additive Increase Multiplicative Decrease Based Techniques

Due to the expected increase in penetration levels of Plug-in Electric Vehicles (PEVs), the demand on the distribution power grid is expected to rise significantly during PEV charging. However, as PEV charging in many cases may not be time critical, they are suitable for load management tasks where the power consumption of PEVs is controlled to support the grid. Additionally, PEVs may also be enabled to inject power into the grid to lower peak demand or counteract the influence of intermittent renewable energy generation, such as that produced by solar photovoltaic panels. Further, PEV active rectifiers can be used to balance reactive power in a local area if required, to reduce the necessity for long distance transport of reactive power. To achieve these objectives, we adapt a known distributed algorithm, Additive Increase Multiplicative Decrease, to control both the active and reactive power consumption and injection. Here, we present this algorithm in a unified framework and illustrate the flexibility of the algorithm to accommodate different user objectives. We illustrate this with three scenarios, including a domestic scenario and a workplace scenario. In these scenarios the various objectives allow us to define a type of “fairness” for how the PEVs should adapt their power consumption, i.e. equal charging rates, or charging rates based on energy requirements. We then validate the algorithms by simulations of a simple radial test network. The simulations presented use the power simulation tool OpenDSS interlinked with MATLAB.