Abstract
To expedite the development cycle of the engine control system and address the challenges posed by multi-objective and non-convex optimization problems during the calibration process, this paper introduces a mesh adaptive Bayesian optimization algorithm for online calibration. Specifically, for the calibration of the fuel injection MAP in a torque-based unit-pump diesel engine's fuel injection control strategy, the influence of convergence criterion hyperparameters is analyzed and optimized. Ultimately, the hyperparameters with a mesh size tolerance of 0.5 and a minimum improvement threshold of 0.001 in the objective function are selected for the fuel injection MAP online calibration simulation. The outcomes demonstrate that this algorithm facilitates rapid and automated online calibration of the fuel injection MAP in unit-pump diesel engines, significantly enhancing the efficiency and accuracy of the calibration process.
| Original language | English |
|---|---|
| Pages (from-to) | 70-75 |
| Number of pages | 6 |
| Journal | IFAC-PapersOnLine |
| Volume | 58 |
| Issue number | 29 |
| DOIs | |
| Publication status | Published - 1 Nov 2024 |
| Event | 7th IFAC Conference on Engine and Powertrain Control, Simulation and Modeling, E-COSM 2024 - Dalian, China Duration: 30 Oct 2024 → 1 Nov 2024 |
Keywords
- analysis and optimization
- mesh adaptive Bayesian optimization
- online calibration
- the influence of hyperparameters
- unit-pump diesel engine