A Privacy-Preserving Distributed Control of Optimal Power Flow

Ryu, Minseok; Kim, Ki-Baek

doi:10.1109/tpwrs.2021.3120056

Cited by 13 publications

(7 citation statements)

References 27 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In this setting, agents are not required to share electric power loads consumed at their zones, but it is still possible to reverse-engineer local solutions communicated between agents to estimate the sensitive power load data [38]. To protect data, one can utilize DP algorithms that admit an inevitable trade-off between data privacy and solution quality.…”

Section: Resultsmentioning

confidence: 99%

“…Experimental settings. For the power network instances, we consider case 14 and case 118, each of which is decomposed into three zones as in [38]. For the ADMM parameters, we set ρ t = 100 for all t and η t = 1/ √ t. We consider various ¯ ∈ {0.05, 0.1, 0.5, 1} where smaller ¯ ensures stronger data privacy, as described in Definition 2.1.…”

Section: Resultsmentioning

confidence: 99%

See 1 more Smart Citation

Differentially Private Distributed Convex Optimization

Ryu¹,

Kim²

2023

Preprint

View full text Add to dashboard Cite

This paper considers distributed optimization (DO) where multiple agents cooperate to minimize a global objective function, expressed as a sum of local objective functions, subject to some constraints. In DO, each agent iteratively solves a local optimization model constructed by its own data and communicates some information (e.g., a local solution) with its neighbors in a communication network until a global solution is obtained. Even though locally stored data are not shared with other agents, it is still possible to reconstruct the data from the information communicated among agents, which could limit the practical usage of DO in applications with sensitive data. To address this issue, we propose a privacy-preserving DO algorithm for constrained convex optimization models, which provides a statistical guarantee of data privacy, known as differential privacy, and a sequence of iterates that converges to an optimal solution in expectation. The proposed algorithm generalizes a linearized alternating direction method of multipliers by introducing a multiple local updates technique to reduce communication costs and incorporating an objective perturbation method in the local optimization models to compute and communicate randomized feasible local solutions that cannot be utilized to reconstruct the local data, thus preserving data privacy. Under the existence of convex constraints, we show that, while both algorithms provide the same level of data privacy, the objective perturbation used in the proposed algorithm can provide better solutions than does the widely adopted output perturbation method that randomizes the local solutions by adding some noise. We present the details of privacy and convergence analyses and numerically demonstrate the effectiveness of the proposed algorithm by applying it in two different applications, namely, distributed control of power flow and federated learning, where data privacy is of concern.

show abstract

Section: Resultsmentioning

confidence: 99%

Section: Resultsmentioning

confidence: 99%

Differentially Private Distributed Convex Optimization

Ryu¹,

Kim²

2023

Preprint

View full text Add to dashboard Cite

show abstract

“…In addition, more and more generators/consumers are concerned about the leakage of their private information (such as bus injection power and load). It is difficult for a single control center to ensure that their private information will not be leaked [9][10] . A bus-oriented decentralized PF approach can avoid the above problems.…”

Section: A Research Motivationmentioning

confidence: 99%

“…The simplest and most straightforward approximation is to replace the current value with the previous value of the neighbor bus. Namely, the neighbor bus voltage variable whose subscript contains 𝑘 + 1 (𝜃 𝑗,𝑘+1 , 𝑉 𝑗,𝑘+1 ) in formula (10) is directly approximated to the known approximation (𝜃 𝑗,𝑘 , 𝑉 𝑗,𝑘 ) obtained from the previous iteration.…”

Section: B Straightforward Approximationmentioning

confidence: 99%

“…Neglecting the error caused by the estimation and substituting formula (11) to formula (10) [ 𝜃 𝑖,𝑘+1 Formula ( 12) is the iterative formula for the 𝑖-th bus voltage derived by straightforwardly estimating the neighbor bus. At this time, the variables to be solved in the iterative formula are consistent with the number of equations, and when all buses update the voltage through the formula (12) to meet the convergence condition, the PF calculation ends.…”

Section: B Straightforward Approximationmentioning

confidence: 99%

See 1 more Smart Citation

A Bus Privacy Preserving Decentralized Power Flow Algorithm Considering Neighbor Partial Derivative Information

Tan,

Peng,

et al. 2024

IEEE Access

View full text Add to dashboard Cite

To ensure the privacy of power system users while maximizing the efficiency of decentralized computational resources, this paper presents a novel decentralized power flow algorithm. Its core focus lies in preserving bus privacy through the integration of neighboring partial derivative information. Initially, the algorithm utilizes partial derivative data from adjacent buses to formulate a bus-level voltage iterative function, rooted in the bus power balance equation. Each bus in the power network is treated as an individual computational unit managed by an agent. This process involves scanning the network's buses, computing estimated bus voltage values using a designated iteration formula, and iterating until convergence is achieved. Subsequent utilization of the latest neighbor information accelerates this iterative process. To fortify the algorithm's reliability, a strategy assessing local convergence, utilizing the Jacobian matrix at the solution's convergence, is proposed. The simulation results across various IEEE test systems unequivocally validate the proposed algorithm's efficacy in power flow computation, showcasing accelerated speed and robust antiinterference capabilities. Compared to the traditional decentralized power flow method at the bus level, this algorithm significantly optimizes the iterative process, slashing the number of iterations by a minimum of 1.36 times. Notably, this efficiency boost amplifies in larger system scales, boasting an impressive reduction of up to 9.21 times fewer iterations.

show abstract