Data-Driven Risk-Aware MPC for Constrained Linear Stochastic Systems

A data-driven risk-aware model predictive control framework for discrete-time linear systems under process noise. The dataset, published under CC-BY-4.0 by Pouria Tooranjipour, includes files in PNG and EPS formats totaling 223.8 KB. It was last updated on June 2, 2026.

Use Cases

• Benchmarking risk-aware control algorithms based on the described Conditional Value-at-Risk (CVaR) framework.
• Studying the performance of disturbance feedback policy parameterization for recursive feasibility.
• Implementing the semidefinite programming (SDP) optimization solution for multistage CVaR problems.
• Analyzing the trade-off between safety and performance predictability in stochastic linear systems.

Strengths

• Theoretical properties like recursive feasibility and risk-aware exponential stability are demonstrated through rigorous analysis.
• The method is described as computationally efficient, cast as a tractable semidefinite programming problem.
• The dataset is openly licensed under CC-BY-4.0, facilitating reuse.

Limitations

• The dataset is very small at 223.8 KB, suggesting limited scope or illustrative examples rather than large-scale application data.
• Column-level documentation is absent; field semantics must be inferred after download.
• Row count is unknown, which may limit suitability assessment for empirical modeling.

Provenance

Source

figshare

Collection Method

Likely contains numerical simulation results or theoretical analysis outputs supporting the described control framework.

Freshness

Last updated 2026-06-02 07:22:53; freshness should be verified.