Trajectory Inference of Unknown Linear Systems Based on Partial States Measurements

Proliferation of cheaper autonomous system prototypes has magnified the threat space for attacks across the manufacturing, transport, and smart living sectors. An accurate trajectory inference algorithm is required for monitoring and early detection of autonomous misbehavior and to take relevant countermeasures. This article presents a trajectory inference algorithm based on a acrlong CLOE approach using partial states measurements. The approach is based on a physics informed state parameteterization that combines the main advantages of state estimation and identification algorithms. Noise attenuation and parameter estimates convergence are obtained if the output trajectories fulfill a persistent excitation condition. Known and unknown desired reference/destination cases are considered. The stability and convergence of the proposed approach are assessed via Lyapunov stability theory under the fulfillment of a persistent excitation condition. Simulation studies are carried out to verify the effectiveness of the proposed approach.