.png){kind=icon}
.png){kind=icon}
Return
A Privacy-Preserving Framework for Cloud-Based HVAC Control
.png){kind=icon}
PRE
AI
DOI：10.1109/TCST.2024.3487019.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
The objective of this work is: 1) to develop an encrypted cloud-based heating, ventilation, and air-conditioning (HVAC) control framework to ensure the privacy of occupancy information; 2) to reduce the communication and computation costs of encrypted HVAC control; and 3) to reduce the leakage of private information via the triggering time instances in event-based encrypted HVAC control systems. Occupancy of a building is sensitive and private information that can be accurately inferred by cloud-based HVAC controllers using HVAC sensor measurements. To ensure the privacy of the occupancy information, in our framework, the sensor measurements of an HVAC system are encrypted by a fully homomorphic encryption (FHE) technique prior to communication with the cloud controller. We first develop an encrypted fast gradient algorithm that allows the cloud controller to regulate the indoor temperature and CO2 of a building by solving two model predictive control (MPC) problems using encrypted HVAC sensor measurements. We next develop an event-triggered control policy to reduce the communication and computation costs of the encrypted HVAC control. We cast the optimal design of the event-triggering policy as an optimal control problem wherein the objective is to minimize a linear combination of the control and communication costs. Using Bellman's optimality principle, we study the structural properties of the optimal event-triggering policy and show that the optimal triggering policy is a function of the current state, the last communicated state with the cloud, and the time since the last communication with the cloud. We also show that the optimal design of the event-triggering policy can be transformed into a Markov decision process (MDP) by introducing two new states. As the triggering time instances are not encrypted, there is a risk that the cloud may use them to deduce sensitive information. To mitigate this risk, we introduce two randomized triggering strategies that reduce the leakage of private information via the triggering time instances. We finally study the performance of the developed encrypted HVAC control framework using the TRNSYS simulator. Our numerical results show that the proposed framework not only ensures efficient control of the indoor temperature and CO2 but also reduces the computation and communication costs of encrypted HVAC control by at least 60%.
Keywords:
Building automation
encrypted model predictive control (MPC)
event-triggered control
optimal control
reinforcement learning
Building automation
encrypted model predictive control (MPC)
event-triggered control
optimal control
reinforcement learning
Journal
IF:
3.9
Papers:
4.9K
Citations:
1.7W