Multi-Horizon Refinement of EDMDc for Aeroengine Model Predictive Control across Multiple Flight Conditions

Abstract Aeroengine model predictive control (MPC) requires accurate predictions over multiple steps across changing flight conditions. This study develops multi-horizon refinement of extended dynamic mode decomposition with control (EDMDc) using measurement histories, protected fuel commands, and measured altitude and Mach number. The refinement minimizes normalized free-running errors over several horizons while preserving prescribed history shifts, with validation-based model selection and spectral screening. AGTF30 identification compares five dictionaries using 1500 s trajectories across sampled conditions spanning 0--12.192 km and Mach 0--0.8. The selected linear-delay model reduces the normalized 15 s test prefix score from 0.21807 to 0.20978 relative to its ridge initializer, with improvements varying by horizon and output. The refined predictor is integrated into constrained MPC with native protection, validated online updates, and proportional--integral (PI) fallback. Compared with PI under common command limits, the complete controller reduces corrected-fan-speed tracking root mean square error from 97.11 to 64.93 rpm in a 240 s repeated-command experiment and from 298.81 to 256.50 rpm in a 75 s staircase and large acceleration/deceleration experiment. Both improvements are accompanied by greater fuel-command variation and characterize the complete controller on the evaluated profiles.

Authors

Institutions

Publication Details

Journal
Research Square
Published
2026-11-17
DOI
https://doi.org/10.21203/rs.3.rs-11464272/v1
Primary Topic
Advanced Control Systems Optimization
Type
preprint
Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
NOV
preprint

Multi-Horizon Refinement of EDMDc for Aeroengine Model Predictive Control across Multiple Flight Conditions

Qiang Shen, Biao Chen, Junfeng Shen, Huihui Li et al.
2 links
Research Square
Advanced Control Systems Optimization
preprint

Multi-Horizon Refinement of EDMDc for Aeroengine Model Predictive Control across Multiple Flight Conditions

Qiang Shen, Biao Chen, Junfeng Shen, Huihui Li, Jing Wang
preprint en

Abstract

Abstract Aeroengine model predictive control (MPC) requires accurate predictions over multiple steps across changing flight conditions. This study develops multi-horizon refinement of extended dynamic mode decomposition with control (EDMDc) using measurement histories, protected fuel commands, and measured altitude and Mach number. The refinement minimizes normalized free-running errors over several horizons while preserving prescribed history shifts, with validation-based model selection and spectral screening. AGTF30 identification compares five dictionaries using 1500 s trajectories across sampled conditions spanning 0--12.192 km and Mach 0--0.8. The selected linear-delay model reduces the normalized 15 s test prefix score from 0.21807 to 0.20978 relative to its ridge initializer, with improvements varying by horizon and output. The refined predictor is integrated into constrained MPC with native protection, validated online updates, and proportional--integral (PI) fallback. Compared with PI under common command limits, the complete controller reduces corrected-fan-speed tracking root mean square error from 97.11 to 64.93 rpm in a 240 s repeated-command experiment and from 298.81 to 256.50 rpm in a 75 s staircase and large acceleration/deceleration experiment. Both improvements are accompanied by greater fuel-command variation and characterize the complete controller on the evaluated profiles.

Research Square
Shanghai Jiao Tong University (CN), Innovation Academy for Microsatellites of Chinese Academy of Sciences (CN)
Advanced Control Systems Optimization
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.

Laika access coming soon