Error proportional derivative mamdani fuzzy-based negative pressure regulator for negative pressure wound therapy machine
Negative Pressure Wound Therapy (NPWT) is a technique that is able to improve the wound healing process by applying negative pressure on the wound in an enclosed environment. However, there are some NPWT wound-injury cases due to the unstable negative pressure that are generated by the NPWT device i...
Saved in:
Main Author: | |
---|---|
Format: | Thesis |
Language: | English |
Published: |
2020
|
Subjects: | |
Online Access: | http://eprints.utm.my/id/eprint/102409/1/TanJiaHouPSBME2020.pdf |
Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
Summary: | Negative Pressure Wound Therapy (NPWT) is a technique that is able to improve the wound healing process by applying negative pressure on the wound in an enclosed environment. However, there are some NPWT wound-injury cases due to the unstable negative pressure that are generated by the NPWT device in which this problem exists due to the application of Classical Controller (CC) in NPWT device. The negative pressure controlling performance is potentially affected by the CC due to its linearity properties. The Input-Output Mapping Relationship (IOMR) of non-linear system is difficult to be described by the linear CC. Therefore, the aim of this research is to improve the negative pressure controlling performance of NPWT. A Fuzzy-based Negative Pressure Regulator (NPR) is developed based on NPWT tolerant band modulated mode which constructed based on the level and characteristic of negative pressure that suitable for NPWT. Its Membership Function (MF) is developed based on the three valued logic which is suitable for resolve boundary problem of CC and its sensitivity is evaluated by MF Sensitivity Test (MFST). It is next improved with the MF Enhancement (MFE) whereby its performance is able to be enhanced with the incorporation of different MF shapes. The next enhancement is the incorporation of Error-Derivative (ED) technique whereby the control mechanism of NPR is able to be improved with the additional derivative input. In addition, mimic-NPWT experiment, which is developed based on the real NPWT, is enrolled for the performance evaluation of the developed NPR. Based on the result from MFST, the 4 Relationship Equation (4RE) MF is the most suitable sensitivity setting for stabilizing the negative pressure since it achieves the highest ±1 tolerant band percentage (63.38%). Based on the result from NPR with MFE, IOMR with incorporation of Zmf+Smf and Dsig MF is the best MF shape setting for improving the NPR whereby increment of 20.3% in ±1 tolerant band percentage is achieved. Next, it is observed that the undershoot and overshoot percentage achieved by NPR with incorporation of ED are 1.27% and 0.43% reduced and it contribute to the 1.68% of increment in ±1 tolerant band percentage. Apart from that, the result from the mimic-NPWT experiment also shows that performance of the NPR is enhanced with 16.26% of increment in ±1 tolerant band percentage and 23.56% higher than that of Conventional Based NPR (C-EPD). Based on the result obtained, the adaptation of the Fuzzy-based NPR to the non-linear NPWT had been successfully accomplished due to its MF adjustment flexibility which could relatively resolve the controlling problem of linear CC. Next, due to the MF adjustment flexibility of Fuzzy, the MFE is applicable for the NPR enhancement whereby the limitation of linear CC had been further enhanced by Fuzzy with MFE and NPR performance on negative pressure stabilization is further improved. Furthermore, the control mechanism of NPR is further improved with the incorporation of ED since the rate of negative pressure control is supported with this incorporation and it favours for reducing the undershoot and overshoot. In conclusion, the adaptation and enhancement of the developed NPR to the non-linear NPWT had been successfully implemented with the Fuzzy with MFE and incorporation of ED which can contribute to the improvement of wound therapy performance and safety. |
---|