Browsing by Author "Zhao, Junjie"

Now showing 1 - 12 of 12
  • Thumbnail Image
    ItemOpen Access
    Abrupt fault detection and isolation for gas turbine components based on a 1D convolutional neural network using time series data
    (AIAA, 2020-08-17) Zhao, Junjie; Li, Yiguang
    The FDI step identifies the presence of a fault, its level, type, and possible location. Gas turbine gas-path fault detection and isolation can improve the availability and economy of gas turbine components. Data-driven FDI methods are studied in this paper. Some notable gas turbine FDI challenges include: insensitivity to operating conditions, robust separation of faults, noisy sensor readings and missing data, reliable fault detection in time-varying conditions, and the influence of performance gradual deterioration. For conventional ML methods, the problem with handling time series data is its volume and the associated computational complexity; therefore, the available information must be appropriately compressed via the transformation of high-dimensional data into a low-dimensional feature space with minimal loss of class separability. In order to improve the detection and isolation sensitivity, this paper develops a method for FDI based on CNNs. Work in this paper includes: (1) Defining the problem and assembling a dataset. (2) Preparing data for training, validation and test: data generation, feature engineering, data pre-processing, data formatting. (3) Building up the model. (4) Training and validating the model (evaluation protocol). (5) Optimizing: a. deciding the model size. b. regularizing the model by getting more training data, reducing the capacity of the network, adding weight regularization or adding dropout. c. tuning hyperparameters. (6) Evaluation.
  • Thumbnail Image
    ItemOpen Access
    Co-simulation digital twin framework for testing future advanced air mobility concepts: a study with BlueSky and AirSim
    (IEEE, 2023-11-10) Zhao, Junjie; Conrad, Christopher; Fremond, Rodolphe; Mukherjee, Anurag; Delezenne, Quentin; Su, Yu; Xu, Yan; Tsourdos, Antonios
    The UK Future Flight Vision and Roadmap outlines the anticipated development of aviation in the UK by 2030. As part of the Future Flight demonstration segment, project HADO (High-intensity Autonomous Drone Operations) will develop, test, and deploy fully automated unmanned aircraft system (UAS) operations at London Heathrow Airport. Cranfield University is leading the synthetic test environment development within the HADO project, and a digital twin (DT) prototype was developed to enable mixed-reality tests for autonomous UAS operations. This paper enhances the existing DT by introducing new co-simulation capacities. Specifically, a co-simulation DT framework for autonomous UAS operations is proposed and tested through a demonstrative use case based on BlueSky and AirSim. This prototype integrates the traffic simulation capabilities of BlueSky with the 3D simulation capabilities of Airsim, to efficiently enhance the simulation capacities of the DT. Notably, the co-simulation framework can leverage the 3D visualization modules, UAS dynamics, and sensor models within external simulation tools to support a more realistic and high-fidelity simulation environment. Overall, the proposed co-simulation method can interface several simulation tools within a DT, thereby incorporating different communication protocols and realistic visualization capabilities. This creates unprecedented opportunities to combine different software applications and leverage the benefits of each tool.
  • Thumbnail Image
    ItemOpen Access
    A co-simulation digital twin with SUMO and AirSim for testing lane-based UTM system concept
    (IEEE, 2024-05-13) Wen, Zhang; Zhao, Junjie; Xu, Yan; Tsourdos, Antonios
    The UAS (Unmanned Aircraft System) Traffic Management (UTM) System Concept of Operations (ConOps) is the first formal design reference document of the UTM system, ConOps aims to bring Class G Airspace into government regulation. However, it should be noted that there are still some shortcomings in ConOps that require further discussion. For example, there are concerns about operational rights, privacy rights, and the potential interference caused by high-rise buildings in urban core areas. The Lane-based UTM systems could potentially help in solving the above issues. The flight paths of Unmanned Aerial Vehicles (UAVs) in urban areas or other areas will interact with the road network, which can facilitate airspace traffic development. Ground traffic flow simulation is generally conducted on three levels: macroscopic, mesoscopic, and microscopic. Some of the commonly used car traffic flow simulation tools include Vissim, SUMO, and MATSim. However, UAV traffic simulation is mostly at a single level, and all of the current mainstream simulation software for UAV, such as Gazebo, AirSim, and Flight Gear, are microscopic-level analyses of UAV operations, lacking uniform management of drone traffic flow and operations. In addition, these UAV traffic simulation studies do not consider the city traffic and road network. In this context, a lane-based cosimulation UAV traffic simulation method is proposed in this study. The co-simulation architecture will be based on the highfidelity three-dimensional (3D) environment developed in the Unreal Engine, UAV simulation with AirSim, and twodimensional (2D) road network simulation with SUMO. A standardized and universal co-simulation architecture and communication interface to ensure interoperability, compatibility, and synchronization will be developed in this study. The lane-based co-simulation method will effectively leverage the road network simulation capacities to turn complex 3D space planning into simple 2D planning, it could reduce computational load and improve system efficiency. The 3D environment will also enhance the simulation capacities with its unique and high-fidelity simulation capacities. Overall, the proposed co-simulation method will support the Digital Twin development by interfacing several simulation tools, incorporating different communications, and adding realistic visualization, which could create unprecedented opportunities for software tool combinations.
  • Thumbnail Image
    ItemOpen Access
    Convolutional neural network denoising autoencoders for intelligent aircraft engine gas path health signal noise filtering
    (American Society of Mechanical Engineers, 2022-10-31) Zhao, Junjie; Li, Yiguang; Sampath, Suresh
    Removing noise from health signals is critical in gas path diagnostics of aircraft engines. An efficient noise filtering/denoising method should remove noise without using future data points, preserve important changes, and promote accurate diagnostics without time delay. Machine Learning (ML)-based methods are promising for high fidelity, accuracy, and computational efficiency under the motivation of Intelligent Engines. However, previous ML-based denoising methods are rarely applied in actual engineering practice because they cannot accommodate time series and cannot effectively capture important changes or are limited by the time delay problem. This paper proposes a Convolutional Neural Network Denoising Autoencoder (CNN-DAE) method to build a denoising autoencoder structure. In this structure, a convolutional operation is used to accommodate time series, and causal convolution is introduced to solve the problem of using future data points. The proposed denoising method is evaluated against NASA's Propulsion Diagnostic Method Evaluation Strategy (ProDiMES) software. It has been proved that the proposed method can accommodate time series, remove noise for improved denoising accuracy and preserve the important changes for enhanced diagnostic information. NASA's blind test case results show that Kappa Coefficient of a common diagnostic method using the processed data is 0.731 and is at least 0.046 higher than the other diagnostic methods in the open literature. Processing health signals using the proposed method would significantly promote accurate diagnostics without time delay. The proposed method could support intelligent condition monitoring systems by exploiting historical information for improved denoising and diagnostic performance.
  • Thumbnail Image
    ItemOpen Access
    Developing a digital twin for testing multi-agent systems in advanced air mobility: a case study of Cranfield University and airport
    (IEEE, 2023-11-10) Conrad, Christopher; Delezenne, Quentin; Mukherjee, Anurag; Mhowwala, Ali Asgher; Ahmed, Mohammad; Zhao, Junjie; Xu, Yan; Tsourdos, Antonios
    Emerging unmanned aircraft system (UAS) and advanced air mobility (AAM) ecosystems rely on the development, certification and deployment of new and potentially intelligent technologies and algorithms. To promote a more efficient development life cycle, this work presents a digital twin architecture and environment to support the rapid prototyping and testing of multi-agent solutions for UAS and AAM applications. It leverages the capabilities of Microsoft AirSim and Cesium as plugins within the Unreal Engine 3D visualisation tool, and consolidates the digital environment with a flexible and scalable Python-based architecture. Moreover, the architecture supports hardware-in-the-loop (HIL) and mixed-reality features for enhanced testing capabilities. The system is comprehensively documented and demonstrated through a series of use cases, deployed within a custom digital environment, comprising both indoor and outdoor areas at Cranfield University and Airport. These include collaborative surveillance, UTM flight authorisation and UTM conformance monitoring experiments, that showcase the modularity, scalability and functionality of the proposed architecture. All 3D models and experimental observations are critically evaluated and shown to exhibit promising results. This thereby represents a critical step forward in the development of a robust digital twin for UAS and AAM applications.
  • Thumbnail Image
    ItemOpen Access
    Developing a stackable programme based on the advanced air mobility systems MSc course
    (Elsevier BV, 2024-09-05) Zhao, Junjie; Gong, Tingyu; Nnamani, Christantus; Conrad, Christopher; Fremond, Rodolphe; Tang, Yiwen; Xu, Yan; Tsourdos, Antonios
    This study proposes the development of content and materials for a stackable programme that aligns with the existing Cranfield University Advanced Air Mobility Systems (AAMS) MSc Course and integrates with ongoing Future Flight Challenge (FFC) projects, emerging research and development (R&D) capacities, and the growing demand for skilled professionals in the sector. The programme is structured into four phases: enhancement of taught modules through technology-enhanced teaching (TET), enrichment of project-based learning, bolstering of student experience and career development, and a stackable approach adaptable to various educational levels. This approach was evaluated using courses from the 2022/23 and 2023/24 academic years.
  • Thumbnail Image
    ItemOpen Access
    A digital twin mixed-reality system for testing future advanced air mobility concepts: a prototype
    (IEEE, 2023-05-15) Zhao, Junjie; Conrad, Christopher; Delezenne, Quentin; Xu, Yan; Tsourdos, Antonios
    The UK Future Flight Vision and Roadmap defines how aviation in the UK is envisioned to develop by 2030. As part of the Future Flight demonstration segment, project HADO (High-intensity Autonomous Drone Operations) will develop, test, and deploy fully automated Unmanned Aircraft System (UAS) operations at London Heathrow airport. The resource-demanding nature of real-world tests, however, suggests that developing and improving the reliability and efficiency of virtual environment-based testing methods is indispensable for the evolution of such operations. Nonetheless, developing a high-fidelity and real-time virtual environment that enables the safe, scalable, and sustainable development, verification, and validation of UAS operations remains a daunting task. Notably, the need to integrate physical and virtual elements with a high degree of correlation presents a significant challenge. Consequently, as part of the synthetic test environment work package within the HADO project, this paper proposes a Digital Twin (DT) system to enable mixed-reality tests in the context of autonomous UAS operations. This connects a physical world to its digital counterpart made up of five distinct layers and several digital elements to support enhanced mixed-reality functionality. The paper highlights how the static layers of the synthetic test environment are built, and presents a DT prototype that supports mixed-reality test capabilities. In particular, the ability to inject virtual obstacles into physical test environments is demonstrated, highlighting how the sharp boundaries between virtual environments and reality can be blurred for safe, flexible, efficient, and effective testing of UAS operations.
  • Thumbnail Image
    ItemOpen Access
    A hierarchical structure built on physical and data-based information for intelligent aero-engine gas path diagnostics
    (Elsevier, 2022-12-23) Zhao, Junjie; Li, Yi-Guang; Sampath, Suresh
    Future trends in engine health management (EHM) systems are information fusion, advanced analytical methods, and the concept of the Intelligent Engines. Machine Learning (ML)-based aero-engine gas path diagnostic methods are promising under the motivation of these trends. However, previous ML-based diagnostic structures are rarely applied in actual engineering practice because they are purely mathematical and lack physical insight or are limited by the error accumulation problem. Developing an accurate, flexible and interpretable intelligent diagnostic method has always posed a challenge, especially when physical knowledge is also available for more diagnostic information. Instead of modifying and applying existing ML methods for classification or regression, this study proposes a novel hierarchical diagnostic method to get insight into the physical systems, build hierarchies automatically, and recommend the classification structures. The proposed hierarchical diagnostic method is evaluated against a NASA model high-bypass two-spool turbofan engine. NASA's blind test case results show that Kappa Coefficient of the proposed hierarchical diagnostic method is 0.693 and is at least 0.008 higher than the other diagnostic methods in the open literature. It has been proved that the proposed method can quantify the dependence relationships between the fault classes for enhanced diagnostic information, recommend the best diagnostic structure for reduced complexity, and solve the error accumulation problem for improved diagnostic accuracy. The proposed method could support intelligent condition monitoring systems by effectively exploiting physical and data-based information for improved model interpretability, model flexibility, diagnostic visibility, diagnostic accuracy, and diagnostic reliability.
  • Thumbnail Image
    ItemOpen Access
    Modelling and performance analysis of vaneless counter rotating turbine in gas turbine engines
    (ISABE, 2017-09-08) Jia, Linyuan; Chen, Yuchun; Gao, Yuan; Zhao, Junjie
    The objective of this article is to develop a method for 1+1/2 type vaneless counter rotating turbine (VCRT) modelling in gas turbine engines and to analyse the influence of the VCRT on a turbofan engine. By studying the off-design features of the VCRT, a new method to describe the VCRT characteristic map was established. The new turbine maps were integrated into the gas turbine performance simulation model based on component maps. The throttle performance of a low bypass counter rotating turbofan engine was simulated and the results were validated with engine test data. The validation results show that the new method improves the accuracy of the VCRT engine performance simulation. The influence of the VCRT on the engine steady state performance was illuminated by comparing the performance of a VCRT engine with that of a conventional engine. Under low engine load conditions, the insufficient work capacity of the VCRT low pressure turbine was detected due to the decrease of both its efficiency and pressure ratio. Consequently, the VCRT engine’s turbine inlet temperature, fan rotating speed, mass flow rate and engine thrust decreased while its specific fuel consumption increased compared to that of a conventional engine.
  • Thumbnail Image
    ItemOpen Access
    Research on quantitative evaluation method of scramjet and integration
    (ISABE, 2017-09-08) Zhang, Huanrong; Chen, Yuchun; Cai, Yuanhu; Zhao, Junjie
    The objective of this paper is to develop a method for the advancement quantitative evaluation of scramjet and hypersonic vehicle/scramjet integration. By studying the theory of analytic hierarchy process (AHP), a new method to build the quantitative evaluation system was established. The theory of Nondimensional parameter was also proposed and the results showed that this calculation method is scientific and effective. Both the numerical simulations of the advancement quantitative evaluation of scramjet and hypersonic vehicle/scramjet integration were carried out with a number of calculation examples and the advancements of different projects were obtained, which proved the feasibility and reliability of the evaluation method. This quantitative evaluation method simples the complex problems of scramjet and integration evaluation and lays the foundation for the further research of hypersonic technology.
  • Thumbnail Image
    ItemOpen Access
    Self-supervised learning-based two-phase flow regime identification using ultrasonic sensors in an S-shape riser
    (Elsevier, 2023-09-07) Kuang, Boyu; Nnabuife, Somtochukwu G.; Whidborne, James F.; Sun, Shuang; Zhao, Junjie; Jenkins, Karl W.
    Two-phase flow regime identification is an essential transdisciplinary topic that spans digital signal processing, artificial intelligence, chemical engineering, and energy. Multiphase flow systems significantly impact pipeline safety, heat transfer, and pressure drop; therefore, precisely identifying the governing flow regime is crucial for effective modeling and design. However, it is challenging due to the geometrical complexity of flow regimes in multiphase flow. With the advances in sensor measurement and machine learning, applying non-destructive tests and self-supervised learning to practical industrial problems has become technically feasible and cost-effective. This study applies a weak-supervised learning-based two-phase flow regime identification solution using a non-destructive tests ultrasonic sensor in an S-shape riser experimental bed by proposing a self-supervised feature extraction algorithm. The proposed self-supervised feature extraction algorithm reduces time/labor consumption and human error in data annotation using SSL, which provides full supervision without manual annotation. The self-supervised feature extraction algorithm uses a bottlenecked neural network and encoder-decoder structure to extract compact features. The self-supervised feature extraction algorithm performance is evaluated using an established convolutional neural network-based classifier. The source data was collected from a 10 × 50 m riser experimental rig. The dataset is made available to the community as part of this study. The performance of the approach is comparable with state-of-the-art methods and is also the first successful attempt to apply self-supervised learning to multiphase flow regime ultrasonic signal identification. This study achieved 98.84%, 0.000663, 0.00312, and 7.71 × 10^5 in accuracy, root mean square error, categorical cross-entropy, and model complexity, respectively. The practical experiment justifies the robustness, fairness, and practicability in the practical application environment. The proposed self-supervised feature extraction brings new approaches and inspirations for the feature extraction step in identifying a two-phase flow regime, and it will be beneficial to generalize this study in different riser shapes in the future.
  • Thumbnail Image
    ItemOpen Access
    A study on co-simulation digital twin with MATLAB and AirSim for future advanced air mobility
    (IEEE, 2024-05-13) Turco, Lorenzo; Zhao, Junjie; Xu, Yan; Tsourdos, Antonios
    The exponential growth in Unmanned Aerial Vehicle (UAV) operations highlights the need for reliable and efficient airspace management. Ensuring the integrity and reliability of UAV Traffic Management (UTM) systems requires extensive testing, verification and validation. Overcoming these challenges is essential to guarantee that UAVs can be successfully integrated into operational airspace while maintaining the highest standards of safety and effectiveness. The development and improvement of virtual environment testing capabilities are critical to the advancement and rapid deployment of UTM services; however, building a high-fidelity digital environment that enables the development, verification and validation of UTM solutions in a compliant, scalable and sustainable manner remains demanding. By integrating MATLAB, Simulink, AirSim, Unreal Engine and Cesium, this study aims to extend the performance and functionality of a Digital Twin (DT) system. This integration leverages the sophisticated modelling capabilities of MATLAB and the advanced 3-dimensional (3D) simulation capabilities of AirSim. It also aligns with emerging trends in the aerospace sector, including autonomous flight and advanced sensing technologies. Firstly, co-simulation is explored as a potential technique to overcome the limitations of individual simulation software available on the market. The general principles are then applied to define and model a communication interface between the software selected to build the DT ecosystem. This interface is then evaluated by proposing practical and achievable test cases. Only after validating the proposed co-simulation framework, further experiments are introduced aiming at improving the vehicle and sensor model performance to obtain more accurate synthetic data. A series of experiments of increasing difficulty is proposed, starting from co-simulating a single sensor (GNSS, INS, LIDAR and camera) up to co-simulating the entire aircraft system for the development of intelligent algorithms such as waypoint following and collision avoidance. Ultimately, this research contributes to the practical application of co-simulation, providing insight into its capabilities and potential influence on the advancement of autonomous aircraft and DT systems.