An enhanced aircraft engine gas path diagnostic method based on upper and lower singleton type-2 fuzzy logic system

Abstract

The gas turbine is the most common engine used in the majority of commercial aircraft. Regarding the economic and social importance of aviation, methods that can identify faults in gas turbines with precision are relevant. Aiming to detect and classify the gas turbine faults, this work uses the upper and lower singleton type-2 fuzzy logic system trained by steepest descent method. Succeeding the model presentation, comparisons are performed with other models proposed in the literature to diagnose gas turbine faults. The investigated data set was obtained from the software Propulsion Diagnostic Method Evaluation Strategy, which was developed by the National Aeronautics and Space Administration for benchmarking purposes in aviation gas turbines. The experimental results show that the upper and lower singleton type-2 fuzzy logic system model had a greater detection and classification performance than the models reported in the literature.

Sistemas Fuzzy aplicados ao Handover em Redes LTE com Falhas de Cobertura

“…Os autores de [14] e [15] apontam que estratégias fuzzy são capazes de fornecer desempenho equivalente (ou até melhor), consumindo menos recursos computacionais, quando confrontados com ferramentas clássicas de inteligência artificial. Por essa razão, muitas técnicas baseadas em regras fuzzy são encontradas em diversas aplicações, como em [16], [8] e [7].…”

“…• Type-2 Fuzzy Logic Classifier (T2) [16]; O modelo ALMMoé um algoritmo que pode ser aplicado em diversos tipos de problemas computacionais, como classificação, regressão, identificação, entre outros. O modelo em questãoé capaz de extrair as informações necessárias dos dados para constituir seus conjuntos fuzzy e suas regras de inferência.…”

An Autonomous Model to Classify Lathe's Cutting Tools Based on TSFRESH, Self-Organised Direction Aware Data Partitioning Algorithm and Machine Learning Techniques

“…Lately, to overcome these issues, computational intelligence tools have been used into the development of diagnosis, prognosis and monitoring systems for industrial processes. Regarding the implementation of CBM methods and monitoring systems, the artificial neural networks (ANN's) are the mainly applied tool, as presented in (Calderano et al, 2019;Lee et al, 2010). However, the information provided by monitoring systems applied in machining processes is acquired in the form of extremely dynamic data flows which have large dimensionality.…”

“…Consequently, the development of systems capable of supervising the machining process through real-time monitoring data gained momentum, as well as strategies that minimise human interference in the process. Additionally, The implementation of smarter maintenance procedures intends to replace the schedule-based maintenance by a condition-based maintenance (Calderano et al, 2019;Lee et al, 2010). The CBM allows predictive analysis in sensor monitored equipment based on the historical stored data previously to the fault's occurrence (Ellis, 2008).…”

On Interval Type-2 Fuzzy Logic System Using the Upper and Lower Method for Supervised Classification Problems

“…sensors [70]. Figure 4.9 Pre-processing diagram block of the EFS generated data [36]. (a) IT2-FLS UL-GMFum (b) IT2-FLS UL-GMFus Table 4.11 Confusion matrix of both IT2-FLS models considering at least 1000 flights per false alarm for the test dataset, where F and NF means fault and non-fault respectively.…”