Scalability and Predictability Model of PLC Systems
(2024) In CODEN:LUTEDX/TEIE EIEM01 20241Industrial Electrical Engineering and Automation
- Abstract
- This master’s thesis is commissioned by SAAB Group with the purpose of developing a model which can be used to estimate the margin for expansion of devices in PLC systems. Although commercial PLC systems are hard real-time systems, they have inherited unpredictable characteristics due to off-the-shelves components. This makes it difficult to measure the impact on performance when a new device is connected to the system.
By pressuring the CPU using a computationally heavy control program, the scheduler is forced to prioritize tasks associated with the addition of a device. This may increase the accuracy of measuring the impact of an increasing number of devices on CPU performance. The commercial PLC system and all devices used are... (More) - This master’s thesis is commissioned by SAAB Group with the purpose of developing a model which can be used to estimate the margin for expansion of devices in PLC systems. Although commercial PLC systems are hard real-time systems, they have inherited unpredictable characteristics due to off-the-shelves components. This makes it difficult to measure the impact on performance when a new device is connected to the system.
By pressuring the CPU using a computationally heavy control program, the scheduler is forced to prioritize tasks associated with the addition of a device. This may increase the accuracy of measuring the impact of an increasing number of devices on CPU performance. The commercial PLC system and all devices used are manufactured by B&R, a member of ABB group providing solutions within industrial automation.
The method consists of measuring idle and cyclic CPU usage (%) before and after a device is connected to the system. This is done using Profiler, a profiling tool provided by B&R. The resulting models, using linear approximation and ridge regression, are capable of predicting the impact on idle and cyclic CPU usage (%) when adding a device to a test setup of similar size to previously measured systems. Although the gathered quantity of data samples can be considered low, the predictions regarding idle load (%) was accurate.
Furthermore, the industrial Ethernet protocol (Powerlink) used in the B&R system is evaluated. Specifically to detect circumstances which might impact the CPU load and the number of devices existing within a single network. Program optimization is briefly investigated and two programming conventions are identified as important for maintaining a large number of devices. (Less) - Popular Abstract (Swedish)
- PLC-system (Programmable Logic Controller) är robusta realtidssystem som används för att styra fysiska processer i krävande miljöer. Moderna PLC-system är väldigt kapabla, men skulle systemet mot förmodan gå på knä, kan man alltid installera ännu ett PLC-system i mån av plats. En lyx som inte finns ombord på ubåtar, där man istället behöver hushålla med befintliga system. Med hjälp av två matematiska modeller lyckades det att prediktera hur systemet belastas under olika förhållanden. Även begränsande faktorer i mjukvara och kommunikation undersöktes.
Please use this url to cite or link to this publication:
http://lup.lub.lu.se/student-papers/record/9178861
- author
- Fornander, Max LU
- supervisor
-
- Ulf Jeppsson LU
- organization
- course
- EIEM01 20241
- year
- 2024
- type
- H3 - Professional qualifications (4 Years - )
- subject
- keywords
- automation, plc, evaluate, cpu, usage, real-time, devices, predicting, performance, cyclic, idle, load, network, program
- publication/series
- CODEN:LUTEDX/TEIE
- report number
- 5523
- language
- English
- id
- 9178861
- date added to LUP
- 2025-01-13 21:27:49
- date last changed
- 2025-01-13 21:27:57
@misc{9178861,
abstract = {{This master’s thesis is commissioned by SAAB Group with the purpose of developing a model which can be used to estimate the margin for expansion of devices in PLC systems. Although commercial PLC systems are hard real-time systems, they have inherited unpredictable characteristics due to off-the-shelves components. This makes it difficult to measure the impact on performance when a new device is connected to the system.
By pressuring the CPU using a computationally heavy control program, the scheduler is forced to prioritize tasks associated with the addition of a device. This may increase the accuracy of measuring the impact of an increasing number of devices on CPU performance. The commercial PLC system and all devices used are manufactured by B&R, a member of ABB group providing solutions within industrial automation.
The method consists of measuring idle and cyclic CPU usage (%) before and after a device is connected to the system. This is done using Profiler, a profiling tool provided by B&R. The resulting models, using linear approximation and ridge regression, are capable of predicting the impact on idle and cyclic CPU usage (%) when adding a device to a test setup of similar size to previously measured systems. Although the gathered quantity of data samples can be considered low, the predictions regarding idle load (%) was accurate.
Furthermore, the industrial Ethernet protocol (Powerlink) used in the B&R system is evaluated. Specifically to detect circumstances which might impact the CPU load and the number of devices existing within a single network. Program optimization is briefly investigated and two programming conventions are identified as important for maintaining a large number of devices.}},
author = {{Fornander, Max}},
language = {{eng}},
note = {{Student Paper}},
series = {{CODEN:LUTEDX/TEIE}},
title = {{Scalability and Predictability Model of PLC Systems}},
year = {{2024}},
}