Monte carlo logarithmic number system for Model Predictive Control

Panagiotis D. Vouzis; Mark G. Arnold; Sylvain Collange; Mayuresh V. Kothare

doi:10.1109/FPL.2007.4380688

Monte carlo logarithmic number system for Model Predictive Control

Panagiotis D. Vouzis, Mark G. Arnold, Sylvain Collange, Mayuresh V. Kothare

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

2 Scopus citations

Abstract

Simple algorithms can be analytically characterized, but such analysis is questionable or even impossible for more complicated algorithms, such as Model Predictive Control (MPC). Instead, Monte Carlo Arithmetic (MCA) enables statistical experimentation with an algorithm during runtime for detection and mitigation of numerical anomalies. Previous studies of MCA have been limited to software floating point. This paper studies how MCA can be used in an FPGA implementation of the Logarithmic Number System (LNS), forming the Monte Carlo Logarithmic Number System (MCLNS). Simulation studies present how MCLNS affects the accuracy vs. performance of an MPC implementation, and synthesis results give an estimate of the cost of utilizing MCLNS in a Xilinx Virtex-IV FPGA.

Original language	English (US)
Title of host publication	Proceedings - 2007 International Conference on Field Programmable Logic and Applications, FPL
Pages	453-458
Number of pages	6
DOIs	https://doi.org/10.1109/FPL.2007.4380688
State	Published - 2007
Externally published	Yes
Event	2007 International Conference on Field Programmable Logic and Applications, FPL - Amsterdam, Netherlands Duration: Aug 27 2007 → Aug 29 2007

Other

Other	2007 International Conference on Field Programmable Logic and Applications, FPL
Country/Territory	Netherlands
City	Amsterdam
Period	8/27/07 → 8/29/07

ASJC Scopus subject areas

Computer Science Applications
Electrical and Electronic Engineering

Access to Document

10.1109/FPL.2007.4380688

Cite this

Vouzis, PD, Arnold, MG, Collange, S & Kothare, MV 2007, Monte carlo logarithmic number system for Model Predictive Control. in Proceedings - 2007 International Conference on Field Programmable Logic and Applications, FPL., 4380688, pp. 453-458, 2007 International Conference on Field Programmable Logic and Applications, FPL, Amsterdam, Netherlands, 8/27/07. https://doi.org/10.1109/FPL.2007.4380688

@inproceedings{ebfc9e7824144a7fa6a3a6340a2aaaee,

title = "Monte carlo logarithmic number system for Model Predictive Control",

abstract = "Simple algorithms can be analytically characterized, but such analysis is questionable or even impossible for more complicated algorithms, such as Model Predictive Control (MPC). Instead, Monte Carlo Arithmetic (MCA) enables statistical experimentation with an algorithm during runtime for detection and mitigation of numerical anomalies. Previous studies of MCA have been limited to software floating point. This paper studies how MCA can be used in an FPGA implementation of the Logarithmic Number System (LNS), forming the Monte Carlo Logarithmic Number System (MCLNS). Simulation studies present how MCLNS affects the accuracy vs. performance of an MPC implementation, and synthesis results give an estimate of the cost of utilizing MCLNS in a Xilinx Virtex-IV FPGA.",

author = "Vouzis, {Panagiotis D.} and Arnold, {Mark G.} and Sylvain Collange and Kothare, {Mayuresh V.}",

year = "2007",

doi = "10.1109/FPL.2007.4380688",

language = "English (US)",

isbn = "1424410606",

pages = "453--458",

booktitle = "Proceedings - 2007 International Conference on Field Programmable Logic and Applications, FPL",

note = "2007 International Conference on Field Programmable Logic and Applications, FPL ; Conference date: 27-08-2007 Through 29-08-2007",

}

TY - GEN

T1 - Monte carlo logarithmic number system for Model Predictive Control

AU - Vouzis, Panagiotis D.

AU - Arnold, Mark G.

AU - Collange, Sylvain

AU - Kothare, Mayuresh V.

PY - 2007

Y1 - 2007

N2 - Simple algorithms can be analytically characterized, but such analysis is questionable or even impossible for more complicated algorithms, such as Model Predictive Control (MPC). Instead, Monte Carlo Arithmetic (MCA) enables statistical experimentation with an algorithm during runtime for detection and mitigation of numerical anomalies. Previous studies of MCA have been limited to software floating point. This paper studies how MCA can be used in an FPGA implementation of the Logarithmic Number System (LNS), forming the Monte Carlo Logarithmic Number System (MCLNS). Simulation studies present how MCLNS affects the accuracy vs. performance of an MPC implementation, and synthesis results give an estimate of the cost of utilizing MCLNS in a Xilinx Virtex-IV FPGA.

AB - Simple algorithms can be analytically characterized, but such analysis is questionable or even impossible for more complicated algorithms, such as Model Predictive Control (MPC). Instead, Monte Carlo Arithmetic (MCA) enables statistical experimentation with an algorithm during runtime for detection and mitigation of numerical anomalies. Previous studies of MCA have been limited to software floating point. This paper studies how MCA can be used in an FPGA implementation of the Logarithmic Number System (LNS), forming the Monte Carlo Logarithmic Number System (MCLNS). Simulation studies present how MCLNS affects the accuracy vs. performance of an MPC implementation, and synthesis results give an estimate of the cost of utilizing MCLNS in a Xilinx Virtex-IV FPGA.

UR - http://www.scopus.com/inward/record.url?scp=48149096378&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=48149096378&partnerID=8YFLogxK

U2 - 10.1109/FPL.2007.4380688

DO - 10.1109/FPL.2007.4380688

M3 - Conference contribution

AN - SCOPUS:48149096378

SN - 1424410606

SN - 9781424410606

SP - 453

EP - 458

BT - Proceedings - 2007 International Conference on Field Programmable Logic and Applications, FPL

T2 - 2007 International Conference on Field Programmable Logic and Applications, FPL

Y2 - 27 August 2007 through 29 August 2007

ER -

Monte carlo logarithmic number system for Model Predictive Control

Abstract

Other

ASJC Scopus subject areas

Access to Document

Other files and links

Fingerprint

Cite this