Preset-Time Robust Multi-Objective Optimization Over Directed Networks Based On Multi-Agent Framework

The present study addresses the preset-time multi-objective optimization problem subject to external disturbances on directed topologies. Firstly, a novel preset-time robust algorithm is designed by employing two types of time-regulator functions, the linear weighted sum method, global information of cost functions, and integral sliding mode control with robustness. This algorithm is tailored for solving multi-objective optimization problems on strongly connected networks. Additionally, two distributed preset-time estimators are proposed and incorporated into the design of the optimization algorithm, effectively eliminating the dependence on global information. Distinct from existing optimization outcomes, the algorithms developed in this study exhibit superior performance in terms of network structure (digraph), convergence time (preset-time), and robustness. Finally, the correctness and effectiveness of the designed optimization algorithms are corroborated by a bilateral negotiation model.