Smart Meter Privacy Control Strategy Including Energy Storage Degradation

Abstract

In this paper, we present a degradation-aware privacy control strategy for smart meters by taking into account the capacity fade and energy loss of the battery, which has not been included previously. The energy management strategy is designed by minimizing the weighted sum of both privacy loss and total energy storage losses, where the weightage is set using a trade-off parameter. The privacy loss is measured in terms of Bayesian risk of an unauthorized hypothesis test. By making firstorder Markov assumptions, the stochastic parameters of energy loss and capacity fade of the energy storage system are modelled using degradation maps. Using household power consumption data from the ECO dataset, the proposed control strategy is numerically evaluated for different trade-off parameters. Results show that, by including the degradation losses in the design of the privacy-enhancing control strategy, significant improvement in battery life can be achieved, in general, at the expense of some privacy loss.

Optimal privacy-by-design strategy for user demand shaping in smart grids

“…To simplify the analysis, we assume that the adversary is unaware of the presence of the ESS and makes a guessĤ k using only the causal information available at time k. Further, we model the ESS using a memoryless stochastic model given in [11], which is characterized by a conditional probability distribution P Z k+1 ,E k ,B k |D k ,Z k where E k , B k denote the energy losses and capacity degradation of the ESS due to the EMU control actions. Let L k denote the expected ESS usage cost which is given by L k = E ρB k +λ k E k , where ρ is the price (in e) of unit capacity loss and λ k is the price (in e) of unit energy.…”

“…In this section, we compute an optimal EMU strategy for the finite-horizon K using the MDP framework. Here, we provide only the key steps in the EMU strategy design and the detailed derivation can be done following the framework in [11]. Let…”

“…denote the set of all information available to the EMU at time-slot k. Further, let µ k denote a randomized strategy of the EMU at time-slot k that specifies a distribution of control action D k given I k , i.e., P D k |I k . Let D k denotes the alphabet of D k satisfying the ESS operating constraints given in [11], I k denote the alphabet of I k , and U k denotes the set of all mappings from I k to ∆ |D k | . For a given strategy µ k ∈ U k and for a given information i k ∈ I k , the weighted cost C k defined in (5) becomes where y k = x k + d k , andζ * denotes the optimal adversarial strategy to make a guessĤ k given the SM measurements y k 1 and the HMMθ.…”

“…For a given strategy µ k ∈ U k and for a given information i k ∈ I k , the weighted cost C k defined in (5) becomes where y k = x k + d k , andζ * denotes the optimal adversarial strategy to make a guessĤ k given the SM measurements y k 1 and the HMMθ. According to [11], a sufficient statistic for the computation of the optimal adversarial strategyζ * iŝ…”

“…In [8], the authors presented the optimal privacy-cost trade-off with the time-varying electricity price by measuring privacy in terms of variations in the load compared to a fixed-target value using a lossless ESS model. In [9]- [11] the Bayesian risk is minimized while taking into account the ESS aspects such as the energy losses and capacity degradation using detailed ESS models. In contrast to the previous work, here we leverage the energy storage capacity of the smart-metering system.…”