Data-based system representations from irregularly measured data

Mohammad Alsalti,Ivan Markovsky,Victor G. Lopez,Matthias A. Müller

DOI: https://doi.org/10.1109/TAC.2024.3423053

2024-07-08

Abstract:Non-parametric representations of dynamical systems based on the image of a Hankel matrix of data are extensively used for data-driven control. However, if samples of data are missing, obtaining such representations becomes a difficult task. By exploiting the kernel structure of Hankel matrices of irregularly measured data generated by a linear time-invariant system, we provide computational methods for which any complete finite-length behavior of the system can be obtained. For the special case of periodically missing outputs, we provide conditions on the input such that the former result is guaranteed. In the presence of noise in the data, our method returns an approximate finite-length behavior of the system. We illustrate our result with several examples, including its use for approximate data completion in real-world applications and compare it to alternative methods.

Systems and Control,Optimization and Control

What problem does this paper attempt to address?

The problem this paper attempts to address is how to obtain a non-parametric representation of a Linear Time-Invariant (LTI) system from irregularly measured data. Specifically, it focuses on how to recover the complete finite-length behavior of the system by leveraging the kernel structure of the Hankel matrix when data samples are missing. This issue is crucial in practical applications because data loss is a common phenomenon, especially in fields such as networked control systems and biomedical applications. Traditional methods based on continuous data often cannot be directly applied in such cases, necessitating the development of new computational methods to handle irregularly measured data.

The main contributions of the paper include:
1. Modifying existing algorithms to handle irregularly measured data in multivariable systems and obtain the kernel representation of the system from such data.
2. Utilizing the kernel structure of the Hankel matrix to recover any complete finite-length behavior of the system, similar to the results of the fundamental lemma, but using only offline data from irregular measurements.
3. Considering the special case of periodically missing outputs and providing input conditions to ensure that the kernel representation can be recovered from the available data.
4. Investigating the presence of noise in the data and providing computational methods to obtain an approximate finite-length behavior of the system.

These contributions make the method more robust and effective in handling irregular and noisy data in practical applications.