A Distributed Consensus Scenario Approach to Optimization and Control With Uncertainties

This paper addresses the problem of distributed optimization with uncertainties for multiagent systems. We propose a new distributed consensus scenario approach, handling uncertainties through a scenario program method. An approximate solution is developed, followed by a distributed estimation algorithm to manage large data volumes without a centralized approach. Our design inherently offers robustness against adversarial attacks and preserves data privacy. We provide rigorous convergence analysis and demonstrate the effectiveness of our approach through application examples in regression and robust internal model control. This work presents a significant step towards holistic solutions for optimization in safety-critical multiagent systems.