A digital twin framework for aircraft hydraulic systems failure detection using machine learning techniques

Abstract

Since the last decade, aircraft systems, such as flight control and landing gear, have been requiring increasing power, and consequently, the complexity of hydraulic aircraft systems has escalated. Inevitably, this complexity has resulted in the need for the troubleshooting of hydraulic aircraft systems that are dispersed around an aircraft and supply power to critical flight systems. This study proposes a novel digital twin-based health monitoring system for aircraft hydraulic systems to enable diagnostics of system failures early in the design cycle using machine learning (ML) methods. The scope of the systems is limited to hydraulic systems at the aircraft level using 20 failure scenarios. The support vector machine and several ensemble learning algorithms of ML methods were used to identify these failures. A comparison of the ML methods revealed that the random forest algorithm performed superior to the other ML algorithms. The developed digital twin framework for hydraulic system of aerial vehicle platforms, can help researchers and engineers to evaluate diagnostics systems early in the design phase.

Towards Safe and Efficient Unmanned Aircraft System Operations: Literature Review of Digital Twins’ Applications and European Union Regulatory Compliance

“…An example of the proposed method is also illustrated on an aircraft wing's fatigue crack growth [68]. Kosova et al [69] developed a DT and used ML for a health-monitoring system (limited to aircraft hydraulic systems) to diagnose system failures in the early stages using 20 failure scenarios. Laukotka et al [70] implemented DTs for civil aviation, aircraft, and aircraft cabins, based on modular product family design and model-based systems engineering.…”

“…DTs can be used in any stage of the aircraft life cycle [50][51][52][53][54][55][56][57][58][59][60], such as design, manufacturing, operations, and maintenance. DTs can also be implemented on components as well as systems [61][62][63][64][65][66][67][68][69][70] that provide a comprehensive view of an aircraft and its individual parts. It allows for monitoring and analysis at different levels, enabling engineers to assess the performance and health of specific components as well as understand the overall behavior and interactions within the system.…”

Simulation Method in Automotive, Aviation and Maritime Industries for Digital Twin: A Brief Survey

“…In recent years, data-driven methods have flourished in the aviation industry with the remarkable development of data transmission, data storage, and data fusion. Several algorithms are also in progress, including supervised methods for random forests [15], convolutional neural networks (CNNs) [16], semisupervised or supervised methods for LSTMs [17], and HMMs [18].…”

Health assessment of an electro-hydraulic servo pump control system for servomotor based on LGA deep neural network

“…Kim [ 12 ] applied wireless sensors for status monitoring in system maintenance and management, and developed a method for extracting data features using the Boruta algorithm, which achieved stability assessment of the hydraulic system's various components' current operational state. Kosova [ 13 ] presented a feature extraction method based on linear discriminant analysis (LDA) to reduce the dimension of features. Macaluso [ 14 ] proposed a PHM health assessment system based on critical factor analysis and detection (FMECA), which enabled accurate assessment of system status with fewer sensors.…”