Related Experiment Video
Updated: Mar 30, 2026

WheelCon: A Wheel Control-Based Gaming Platform for Studying Human Sensorimotor Control
Published on: August 15, 2020
Experience Replay for Optimal Control of Nonzero-Sum Game Systems With Unknown Dynamics
This study introduces an approximate online equilibrium solution for N-player nonzero-sum (NZS) games with unknown dynamics. The method uses neural networks and experience replay for stable control and convergence in complex game systems.
Area of Science:
- Game Theory
- Control Systems
- Artificial Intelligence
Background:
- N-player nonzero-sum (NZS) games with unknown dynamics present significant control challenges.
- Traditional methods often require complete system knowledge or strong persistence of excitation conditions.
Purpose of the Study:
- To develop an approximate online equilibrium solution for N-player NZS game systems with unknown dynamics.
- To ensure stability and convergence of value functions in adaptive dynamic programming (ADP) control schemes.
Main Methods:
- A three-layer neural network (NN) model identifier is used to reconstruct unknown NZS game dynamics.
- Experience replay technique is employed to update identifier weights, relaxing excitation conditions.
- A single-network adaptive dynamic programming (ADP) algorithm with experience replay solves coupled nonlinear Hamilton-Jacobi (HJ) equations.
Main Results:
- The proposed method provides a feedback Nash equilibrium solution.
- A novel critic NN weights tuning law guarantees closed-loop system stability and value function convergence.
- Lyapunov-based stability analysis confirms uniform ultimate boundedness of the closed-loop system.
Conclusions:
- The developed approximate online equilibrium solution effectively addresses N-player NZS games with unknown dynamics.
- The integration of neural networks and experience replay offers a robust approach to adaptive control.
- Simulation examples validate the proposed control scheme's effectiveness.
Related Concept Videos
Parameters Affecting Nonlinear Elimination: Zero-Order Input, First-Order Absorption and Two-Compartment Model
When a drug is administered through a constant intravenous infusion and eliminated via nonlinear pharmacokinetics, it follows zero-order input. For example, oral drugs undergo first-order absorption upon administration and are eliminated through nonlinear pharmacokinetics.
In the case of subcutaneously administered drugs,...
Stability of Equilibrium Configuration: Problem Solving
Problem-solving in the context of the stability of equilibrium configuration...
Dynamic Equilibrium
Alternative Sets of Equilibrium Equations
One example of such a situation can be observed in a...
Differential Equations: Problem Solving
Multi-input and Multi-variable systems
In the absence of...
