Physics

Pathology

Event (particle physics)

Mathematics

Multi-agent system

Economic growth

Evolutionary biology

Biology

Algorithm

Mathematical optimization

Regular polygon

Subgradient method

Function (biology)

Convex function

Artificial intelligence

Economics

Set (abstract data type)

Information exchange

Programming language

Alternative medicine

Geometry

Convergence (economics)

Protocol (science)

Computer science

Quantum mechanics

Projection (relational algebra)

Medicine

Telecommunications

International Journal of Systems Science

This paper discusses a convex optimisation problem with a common set of constraints in the framework of multi-agent systems. Each agent only exchanges information with its neighbours and collaboratively searches for the optimal solution of the global function. To this addressed problem, a distributed multi-step subgradient projection algorithm is developed, where an adaptive event-triggering protocol is designed to govern the information exchange. It is disclosed that the state of each agent representing the estimate of the optimal solution asymptotically converges to one of the optimal solutions under suitably chosen stepsizes and momentum parameters. Simulation results verify that the proposed algorithm has better convergence performance than the standard event-triggered subgradient projection algorithm. In addition, the communication frequency between agents can be effectively reduced to save communication resource consumption.

Distributed multi-step subgradient projection algorithm with adaptive event-triggering protocols: a framework of multiagent systems