A Recursive Non-Uniform Sampling Estimator for Asynchronous Nonlinear Systems

Area of Science:

• Control Systems Engineering
• Nonlinear System Analysis
• Signal Processing

Background:

• Asynchronous sampling and packet losses complicate state estimation in nonlinear systems.
• Accurate dynamic models are often difficult to obtain due to inherent uncertainties and unmodeled dynamics.
• Existing estimation methods may struggle with the combined challenges of asynchronous data and data loss.

Purpose of the Study:

• To develop a robust state estimation method for randomly sampling nonlinear systems experiencing packet losses.
• To address the synchronization issue in asynchronous systems by weighting state updates.
• To propose an observation inference technique for improved estimation accuracy despite modeling errors.

Main Methods:

• Synchronization of asynchronous sampling via state weighting of adjacent update points.
• Application of the projection theorem for state estimation at sampling instances.
• Development of observation inference using interpolation techniques to handle modeling uncertainties.
• Extension of the algorithm for distributed fusion estimation in multi-sensor systems.

Main Results:

• Successful synchronization of asynchronous sampling systems.
• Effective state estimation even with packet losses at control and measurement points.
• Improved estimation accuracy through observation inference, mitigating challenges from modeling errors.
• Validation of a distributed fusion estimator for multi-sensor applications.

Conclusions:

• The proposed observation inference method effectively handles asynchronous estimation in nonlinear systems with packet losses.
• The developed algorithm provides a robust solution for state estimation in challenging system dynamics.
• The extension to multi-sensor systems demonstrates the algorithm's versatility and potential for practical applications.