TY - BOOK
T1 - 20-SIM code generation for PC/104 target
AU - Groothuis, M.A.
PY - 2001
Y1 - 2001
N2 - From version 3.2, 20-Sim will contain a new tool, called C-code generation. With this tool it will be possible to generate C code from a 20-Sim model. This tool works on basis of templates. For each target, a target specific template has to be made.
The goal of this project was to write a new 20-Sim template for a PC/104 target, a embedded Intel 80486 pc on the mobile robot Arty. It should to be possible to control the robot with a 20-Sim model. The template had to be written using the Phar Lap TNT Toolsuite Lite, a toolsuite for real-time embedded development. The written template is working and to demonstrate this, a 20-Sim model that controls the robot is made.
Another goal of this project was testing the code generation tool and doing some recommendations for improvement. From this project it followed that this tool works good, but that the following points have to be improved:
· Controlling the target hardware from a 20-Sim model is possible but only via a time-consuming workaround route.
· The generated code is large, because it contains many not-used functions and redundant variables and equations.
· The Runge Kutta routines fail to work in the existing templates when a model has no states.
AB - From version 3.2, 20-Sim will contain a new tool, called C-code generation. With this tool it will be possible to generate C code from a 20-Sim model. This tool works on basis of templates. For each target, a target specific template has to be made.
The goal of this project was to write a new 20-Sim template for a PC/104 target, a embedded Intel 80486 pc on the mobile robot Arty. It should to be possible to control the robot with a 20-Sim model. The template had to be written using the Phar Lap TNT Toolsuite Lite, a toolsuite for real-time embedded development. The written template is working and to demonstrate this, a 20-Sim model that controls the robot is made.
Another goal of this project was testing the code generation tool and doing some recommendations for improvement. From this project it followed that this tool works good, but that the following points have to be improved:
· Controlling the target hardware from a 20-Sim model is possible but only via a time-consuming workaround route.
· The generated code is large, because it contains many not-used functions and redundant variables and equations.
· The Runge Kutta routines fail to work in the existing templates when a model has no states.
M3 - Report
T3 - Report / Control Laboratory Electrical Engineering Department
BT - 20-SIM code generation for PC/104 target
PB - University of Twente
CY - Enschede, the Netherlands
ER -