Neural network-based (NN) Model Predictive Control (MPC) algorithm to control a multi-agent system (MAS) with stochastic communication topology.
Paper available: https://www.sciencedirect.com/science/article/pii/S095219762200361X Floriano, B. R., Vargas, A. N., Ishihara, J. Y., & Ferreira, H. C. (2022). Neural-network-based model predictive control for consensus of nonlinear systems. Engineering Applications of Artificial Intelligence, 116, 105327.
Folder includes the NN-based MPC algorithm to achieve consensus in MAS for quadrotor fleet system, system with disturbances and nonlinear robot-car system.
The main file is "main.m". In order to run it, uncomment line:
- Line 17 - "start_WW20_Fig1" to run linear system (based on Wang et al. (2018));
- Line 18 - "start_GL20_Fig3" to run system with disturbances (based on Gao et al. (2020));
- Line 21 - "start_Car" to run nonlinear robot-car system.
The output is the evolution of the system's states over time.
Includes the data (.mat) and the code to generate the graphs and tables.
The subfolder "results" contain all the data (.mat) shown in the Results section. It is organized according to the folder's names as follows:
-
WW20: refers to the results of the linear system
-
GL20: refers to the results of the system with disturbances
-
Car: refers to the results of the nonlinear robot-car system
-
Crossval: refers to the cross-validation results
-
Horizon: refers to the results for different horizons
-
Pi_x: refers to the results with the transition matrix \Pi = \Pi_x
To generate the graphics of state's progression over time, run "analysis.m". To generate the cross-validation and horizon graphics, run "analysis_crossval.m".