Privacy-Enabled Secure Iterative Learning Control of Networked Systems Subject to Disclosure Attacks

In this brief, a secure iterative learning control architecture is proposed for networked systems, aiming to track the desired output and protect the privacy of signals. The main technique challenges lie in that the homomorphic encryption is integrated into the control scheme without destructing the performance of systems. Therefore, the iterative learning control algorithm based on the quantized feedback and saturated input is presented to support the design of encrypted controller, and the conditions on the parameters of cryptosystem are established to ensure that the inputs obtained by actuator are reliable. As the results, the calculation of secure control law only relies on the encrypted data, thus immunizing the disclosure attacks of adversaries. To the best of our knowledge, there is currently little work to investigate the secure iterative learning control scheme, which can enhance the practicability and security of systems due to the data-driven strategy and encryption mechanism.