Distributed Differentially Private Model Predictive Control for Energy Storage

A review of preserving privacy in data collected from buildings with differential privacy

“…The analysis of local energy storage (e.g., rechargeable battery) attached to a building has gained increased attention in past decades, and there have also been studies [78][79][80][81] that concentrated on implementing differential privacy by hiding the raw energy consumption characteristics with a battery. The use of the battery (shown in Figure 2) in a building has the ability to shift the peak of end users' power consumption, which can result in reduced electricity prices when dynamic electricity pricing is adopted.…”

“…For this reason, a technique for adjusting and scaling the noise generation is applied to resolve these constraints. Zellner et al [80] developed a model predictive controller leveraging the local energy storage (e.g., battery) for differentially privatizing the load profile of smart meters while minimizing the cost of the electricity they use. An optimization problem for distributed (local) smart meters was considered, and two main objectives were targeted: 1) minimization of the energy costs for smart meter users under a dynamic electricity pricing scheme and 2) smoothing (and obfuscating) the load profile to lower the operation cost of a utility company.…”

Cloud-Based MPC with Encrypted Data

“…In control systems, ensuring the privacy of the measurements and control inputs from eavesdroppers and from the controller has been so far tackled with differential privacy, homomorphic encryption and transformation methods. Kalman filtering with DP was addressed in [15], current trajectory hiding in [16], linear distributed control [17], and distributed MPC in [18]. The idea of encrypted controllers was introduced in [19] and [20], using PHE, and in [21] using FHE.…”