Data-driven Control with Real-time Uncertainty Compensation for Multi-Fuel Engines

2026-06-15 · Source: Machine Learning · Field: Science & Research — Engineering & Applied Sciences, Artificial Intelligence & Machine Learning · Depth: Expert, quick

Summary

A novel data-driven real-time uncertainty compensation framework is introduced for combustion control in multi-fuel compression ignition (CI) engines. This framework addresses the challenge of achieving consistent, optimal combustion phasing across diverse operating conditions despite modeling uncertainties. It employs a pseudo-engine speed for dynamic control input adaptation and models the combustion process using a Gaussian Process Regression (GPR) model trained on input-output data. Control inputs are synthesized via GPR model inversion and augmented with an uncertainty compensator to mitigate deviations from dynamic variations. This integrated strategy enables real-time input corrections within a finite number of combustion cycles, with theoretical guarantees for finite-time convergence. Simulation results confirm its ability to steer combustion phasing to desired values, offering a scalable and adaptive solution.

Key takeaway

For Control Systems Engineers optimizing multi-fuel CI engine performance, this framework offers a robust approach to manage combustion phasing uncertainties. You should consider integrating data-driven Gaussian Process Regression models with real-time uncertainty compensation. This ensures consistent performance and dynamic adaptation across varying operating conditions, providing finite-time convergence guarantees for your control systems.

Key insights

A data-driven framework uses GPR and uncertainty compensation for real-time combustion control in multi-fuel engines.

Principles

• Pseudo-engine speed enables dynamic control adaptation.
• GPR models nonlinear, fuel-dependent combustion behavior.
• Uncertainty compensation mitigates dynamic deviations.

Method

Train a GPR model on input-output data, synthesize control inputs via model inversion, and augment with an uncertainty compensator for real-time corrections.

In practice

• Apply GPR for complex engine combustion modeling.
• Integrate uncertainty compensators for dynamic system control.

Topics

• Multi-fuel Engines
• Combustion Control
• Gaussian Process Regression
• Uncertainty Compensation
• Data-driven Control
• Real-time Systems

Best for: AI Scientist, Machine Learning Engineer, Research Scientist

Related on AIssential

• Learning-Based Decision Making for Combustion Phasing Control in Multi-Fuel CI Engines with Latent Fuel Reactivity Estimation — Aligning fuel-reactivity inference with control policy learning avoids train-deploy inconsistency in dynamic systems.
• A Multi-Plant Machine Learning Framework for Emission Prediction, Forecasting, and Control in Cement Manufacturing — Data-driven ML models can significantly reduce cement plant NOx emissions and costs without hardware changes.
• Local Guidance, Global Impact: Gaussian-Reshaped Trust Region Unlocks Behavior Transitions — GTR improves PPO in non-stationary environments by using a Gaussian-reshaped, non-monotonic trust region and adaptive anchors.
• Timescale Separation Enables Deep Reinforcement Learning Control of Rotating Detonation Engine Mode Transitions — Reformulating DRL in a moving reference frame enables effective control of multi-timescale systems like RDEs.
• Does "Do Differentiable Simulators Give Better Policy Gradients?'' Give Better Policy Gradients? — Discontinuities in differentiable simulators introduce bias, but variance control often dominates practical policy gradient performance.
• Families of Control-Cost-Parametrized Inverse-Optimal Universal Stabilizers — A new control-cost-parametrized stabilizer family offers design freedom via a nonlinear "expander" formula and neural operator approximation.

Open in AIssential →

Editorial summary, takeaway, and curation by AIssential. Original article published by Machine Learning.