Quantication of FPGA Requirements for Closed-Loop Combustion Control Implementation
(2021) ICE 2021 15th International Conference on Engines & Vehicles- Abstract
- This paper investigates the quantification of FPGA resources for the implementation of closed-loop combustion control algorithms. Specifically, the study focuses on methods for their in-cycle execution and control of the combustion in real-time. A National Instruments Xilinx Virtex-5 platform was used for the quantification of the resources.
Closed-loop combustion control obtains feedback from fast in cylinder pressure measurements for accurate and reliable information about the combustion progress synchronized with the flywheel encoder. The requirements on the signal processing (evaluation rate and digital resolution) are investigated for an effective in-cycle control of the combustion. Tools for the analysis of the selection and... (More) - This paper investigates the quantification of FPGA resources for the implementation of closed-loop combustion control algorithms. Specifically, the study focuses on methods for their in-cycle execution and control of the combustion in real-time. A National Instruments Xilinx Virtex-5 platform was used for the quantification of the resources.
Closed-loop combustion control obtains feedback from fast in cylinder pressure measurements for accurate and reliable information about the combustion progress synchronized with the flywheel encoder. The requirements on the signal processing (evaluation rate and digital resolution) are investigated for an effective in-cycle control of the combustion. Tools for the analysis of the selection and design of the signal scaling are summarized. The implementation of the in-cylinder pressure signal processing for the computation of the heat release is studied and different alternatives analyzed. By the optimization of the arithmetic and model formulation, the necessary FPGA resources can be reduced by 80%.
In-cycle closed-loop combustion controllers were previously investigated by the authors. In this paper, the resources for the implementation of the different modules and control strategies are studied to identify the hardware requirements. The modules for the heat release calculation, combustion observers, predictive models and in-cycle controller are investigated. The results show that the total number of slices required for the computations are the main limiting factor on the consumed FPGA resources, where a total of 18729 slices out of the 17280 available would be necessary. Suggestions on how to improve these limitations are discussed based on the results.
The quantification of the required hardware provides guidance on how to select an FPGA to implement the different in-cycle combustion control alternatives. This permits to evaluate the total cost of the system as a trade-off between the increased efficiency by the closed-loop combustion control and the cost for its implementation. (Less)
Please use this url to cite or link to this publication:
https://lup.lub.lu.se/record/5d54d23a-2b3e-4954-a9e6-6c0a0dc2da09
- author
- Jorques Moreno, Carlos LU ; Stenlåås, Ola LU and Tunestål, Per LU
- organization
- publishing date
- 2021
- type
- Chapter in Book/Report/Conference proceeding
- publication status
- submitted
- subject
- host publication
- ICE2021, International Conference on Engines and Vehicles
- pages
- 12 pages
- conference name
- ICE 2021 15th International Conference on Engines & Vehicles
- conference location
- Capri, Naples, Italy
- conference dates
- 2021-09-12 - 2021-09-16
- language
- English
- LU publication?
- yes
- id
- 5d54d23a-2b3e-4954-a9e6-6c0a0dc2da09
- date added to LUP
- 2021-04-20 09:52:07
- date last changed
- 2021-04-20 12:25:17
@inproceedings{5d54d23a-2b3e-4954-a9e6-6c0a0dc2da09, abstract = {{This paper investigates the quantification of FPGA resources for the implementation of closed-loop combustion control algorithms. Specifically, the study focuses on methods for their in-cycle execution and control of the combustion in real-time. A National Instruments Xilinx Virtex-5 platform was used for the quantification of the resources.<br/>Closed-loop combustion control obtains feedback from fast in cylinder pressure measurements for accurate and reliable information about the combustion progress synchronized with the flywheel encoder. The requirements on the signal processing (evaluation rate and digital resolution) are investigated for an effective in-cycle control of the combustion. Tools for the analysis of the selection and design of the signal scaling are summarized. The implementation of the in-cylinder pressure signal processing for the computation of the heat release is studied and different alternatives analyzed. By the optimization of the arithmetic and model formulation, the necessary FPGA resources can be reduced by 80%.<br/>In-cycle closed-loop combustion controllers were previously investigated by the authors. In this paper, the resources for the implementation of the different modules and control strategies are studied to identify the hardware requirements. The modules for the heat release calculation, combustion observers, predictive models and in-cycle controller are investigated. The results show that the total number of slices required for the computations are the main limiting factor on the consumed FPGA resources, where a total of 18729 slices out of the 17280 available would be necessary. Suggestions on how to improve these limitations are discussed based on the results.<br/>The quantification of the required hardware provides guidance on how to select an FPGA to implement the different in-cycle combustion control alternatives. This permits to evaluate the total cost of the system as a trade-off between the increased efficiency by the closed-loop combustion control and the cost for its implementation.}}, author = {{Jorques Moreno, Carlos and Stenlåås, Ola and Tunestål, Per}}, booktitle = {{ICE2021, International Conference on Engines and Vehicles}}, language = {{eng}}, title = {{Quantication of FPGA Requirements for Closed-Loop Combustion Control Implementation}}, year = {{2021}}, }