Full Text Available
Note: Clicking the button above will open the full text document at the original institutional repository in a new window.
Performability of Integrated Networked Control Systems
A direct sensor to actuator communication model (S2A) for unmodified Ethernet-based Networked Control Systems (NCSs) is presented in this research. A comparison is made between the S2A model and a previously introduced model including an in-loop controller node. OMNET simulations showed the success...
Saved in:
| Main Author: | |
|---|---|
| Format: | Thesis |
| Published: |
AUC Knowledge Fountain
2014
|
| Subjects: | |
| Tags: |
No Tags, Be the first to tag this record!
|
| Summary: | A direct sensor to actuator communication model (S2A) for unmodified Ethernet-based Networked Control Systems (NCSs) is presented in this research. A comparison is made between the S2A model and a previously introduced model including an in-loop controller node. OMNET simulations showed the success of the S2A model in meeting system delay with strict zero packet loss (with no over-delayed packets) requirements. The S2A model also showed a reduction in the end-to-end delay of control packets from sensor nodes to actuator nodes in both Fast and Gigabit switched Ethernet-Based. Another major improvement for the S2A model is accommodating the increase in the amount of additional load compared to the in-loop model. Two different controller-level fault-tolerant models for Ethernet-based Networked Control Systems (NCSs) are also presented in this research. These models are studied using unmodified Fast and Gigabit Ethernet. The first is an in-loop fault-tolerant controller model while the second is a fault-tolerant direct Sensor to Actuator (S2A) model. Both models were shown via OMNeT++ simulations to succeed in meeting system end-to-end delay with strict zero packet loss (with no over-delayed packets) requirements. Although, it was shown that the S2A model has a lower end-to-end delay than the in-loop controller model, the fault-tolerant in-loop model performs better than the fault-tolerant S2A model in terms of less total end-to-end delay in the fault-free situation. While, on the other hand, in the scenario with the failed controller(s), the S2A model was shown to have less total end-to-end delay. Performability analysis between the two fault-tolerant models is studied and compared using fast Ethernet links relating controller failure with reward, depending on the system state. Meeting control system's deadline is essential in Networked Control Systems and failing to meet this deadline represents a failure of the system. Therefore, the reward is considered to be how far is the total end-to-end delay in each state in each model from the system deadline. A case study is presented that simultaneously investigates the failure on the controller level with reward. |
|---|