Single parameter adaptive neural network control for multi-agent deployment with prescribed tracking performance

This study addresses the problem of deploying multi-agent systems using single-parameter adaptive neural network control in time and space where the system is modeled by a parabolic partial differential equation (PDE). We investigate the agent model, simplify the pointwise dynamics using a PDE model, and consider the deployment problem when the number of agents is relatively large. In order for the deployed agents to follow the desired trajectory, we augment the agent dynamics with individual control inputs, accounting for the unknown interference faced by each agent during the deployment process. In the proposed approach, a radial basis function neural network structure is introduced to enhance the systems’ adaptivity under unknown interference. The unknown parameter is estimated via the single-parameter idea for reducing the computation of the entire process and increasing the calculation speed. Asymmetric performance constraints are imposed on the tracking error of the system to ensure that each agent is deployed in the required position. The results of numerical simulation prove the effectiveness of the method.