Validation of Cross-Correlation Detonation Wave Mode Identification Through High-Speed Image Analysis

Abstract Utilizing a neural network, individual down-axis images of combustion waves in a Rotating Detonation Engine (RDE) can be classified according to the number of detonation waves present and their directional behavior. While the ability to identify the number of waves present within individual images might be intuitive, the further classification of wave rotational direction is a result of the detonation wave’s profile, which suggests its angular direction of movement. The application of deep learning is highly adaptive and therefore can be trained for a variety of image collection methods across RDE study platforms. In this study, a supervised approach is employed where a series of manually classified images is provided to a neural network for the purpose of optimizing the classification performance of the network. These images, referred to as the training set, are individually labeled as one of ten modes present in an experimental RDE. Possible classifications include deflagration, clockwise and counterclockwise variants of co-rotational detonation waves with quantities ranging from one to three waves, as well as single, double and triple counter-rotating detonation waves. After training the network, a second set of manually classified images, referred to as the validation set, is used to evaluate the performance of the model. The ability to predict the detonation wave mode in a single image using a trained neural network substantially reduces computational complexity by circumnavigating the need to evaluate the temporal behavior of individual pixels throughout time. Results suggest that while image quality is critical, it is possible to accurately identify the modal behavior of the detonation wave based on only a single image rather than a sequence of images or signal processing. Successful identification of wave behavior using image classification serves as a stepping stone for further machine learning integration in RDE research and comprehensive real-time diagnostics.

Download Full-text

Application of a Convolutional Neural Network for Wave Mode Identification in a Rotating Detonation Combustor Using High-Speed Imaging

Journal of Thermal Science and Engineering Applications ◽

10.1115/1.4049868 ◽

2021 ◽

pp. 1-22

Author(s):

Kristyn B. Johnson ◽

Donald H. Ferguson ◽

Robert S. Tempke ◽

Andrew C. Nix

Keyword(s):

Neural Network ◽

Detonation Wave ◽

High Speed ◽

Wave Mode ◽

Classification Performance ◽

Detonation Waves ◽

Single Image ◽

High Speed Imaging ◽

Mode Identification ◽

Rotating Detonation

Abstract Utilizing a neural network, individual down-axis images of combustion waves in a Rotating Detonation Engine (RDE) can be classified according to the number of detonation waves present and their directional behavior. While the ability to identify the number of waves present within individual images might be intuitive, the further classification of wave rotational direction is a result of the detonation wave's profile, which suggests its angular direction of movement. The application of deep learning is highly adaptive and therefore can be trained for a variety of image collection methods across RDE study platforms. In this study, a supervised approach is employed where a series of manually classified images is provided to a neural network for the purpose of optimizing the classification performance of the network. These images, referred to as the training set, are individually labeled as one of ten modes present in an experimental RDE. Possible classifications include deflagration, clockwise and counterclockwise variants of corotational detonation waves with quantities ranging from one to three waves, as well as single, double and triple counter-rotating detonation waves. The ability to predict the detonation wave mode in a single image using a trained neural network substantially reduces computational complexity by circumnavigating the need to evaluate the temporal behavior of individual pixels throughout time. Results suggest that while image quality is critical, it is possible to accurately identify the modal behavior of the detonation wave based on only a single image rather than a sequence of images or signal processing.

Download Full-text

Application of a Convolutional Neural Network for Wave Mode Identification in a Rotating Detonation Combustor Using High-Speed Imaging

10.1115/1.0003218v ◽

2021 ◽

Author(s):

Kristyn Johnson ◽

Donald Ferguson ◽

Andrew Nix ◽

Robert Tempke

Keyword(s):

Neural Network ◽

Convolutional Neural Network ◽

High Speed ◽

Wave Mode ◽

High Speed Imaging ◽

Mode Identification ◽

Rotating Detonation

Download Full-text

High-speed brightfield 3-D imaging and 3-D image analysis of thick and overlapped cell clusters in cytological preparations

Proceedings, annual meeting, Electron Microscopy Society of America ◽

10.1017/s0424820100168827 ◽

1994 ◽

Vol 52 ◽

pp. 218-219

Author(s):

Robert W. Mackin

Keyword(s):

Image Analysis ◽

High Speed ◽

Three Dimensional ◽

Image Data ◽

Ccd Camera ◽

Objective Lens ◽

Array Processor ◽

Vaginal Smears ◽

Low Contrast ◽

Computer Controlled

This paper presents two advances towards the automated three-dimensional (3-D) analysis of thick and heavily-overlapped regions in cytological preparations such as cervical/vaginal smears. First, a high speed 3-D brightfield microscope has been developed, allowing the acquisition of image data at speeds approaching 30 optical slices per second. Second, algorithms have been developed to detect and segment nuclei in spite of the extremely high image variability and low contrast typical of such regions. The analysis of such regions is inherently a 3-D problem that cannot be solved reliably with conventional 2-D imaging and image analysis methods.High-Speed 3-D imaging of the specimen is accomplished by moving the specimen axially relative to the objective lens of a standard microscope (Zeiss) at a speed of 30 steps per second, where the stepsize is adjustable from 0.2 - 5μm. The specimen is mounted on a computer-controlled, piezoelectric microstage (Burleigh PZS-100, 68/μm displacement). At each step, an optical slice is acquired using a CCD camera (SONY XC-11/71 IP, Dalsa CA-D1-0256, and CA-D2-0512 have been used) connected to a 4-node array processor system based on the Intel i860 chip.

Download Full-text

High-Speed Image Analysis of the Rock Fracture Process under the Impact of Blasting

Физико-технические проблемы разработки полезных ископаемых ◽

10.15372/ftprpi20190408 ◽

2019 ◽

Keyword(s):

Image Analysis ◽

High Speed ◽

Fracture Process ◽

Rock Fracture ◽

The Impact

Download Full-text

Auto and Cross Correlation Measurements in a Turbine Cascade Using a Digital Correlator

Volume 1: Turbomachinery ◽

10.1115/82-gt-132 ◽

1982 ◽

Cited By ~ 3

Author(s):

P. J. Bryanston-Cross ◽

J. J. Camus

Keyword(s):

Mach Number ◽

High Speed ◽

Cross Correlation ◽

Turbine Blades ◽

Image Plane ◽

Strongly Correlated ◽

Trailing Edge ◽

Number Range ◽

Cross Correlations ◽

Digital Correlator

A simple technique has been developed which samples the dynamic image plane information of a schlieren system using a digital correlator. Measurements have been made in the passages and in the wakes of transonic turbine blades in a linear cascade. The wind tunnel runs continuously and has independently variable Reynolds and Mach number. As expected, strongly correlated vortices were found in the wake and trailing edge region at 50 KHz. Although these are strongly coherent we show that there is only limited cross-correlation from wake to wake over a Mach no. range M = 0.5 to 1.25 and variation of Reynolds number from 3 × 105 to 106. The trailing edge fluctuation cross correlations were extended both upstream and downstream and preliminary measurements indicate that this technique can be used to obtain information on wake velocity. The vortex frequency has also been measured over the same Mach number range for two different cascades. The results have been compared with high speed schlieren photographs.

Download Full-text

Collision-Free Tool Path Generation for Five-Axis High Speed Machining

Key Engineering Materials ◽

10.4028/www.scientific.net/kem.474-476.961 ◽

2011 ◽

Vol 474-476 ◽

pp. 961-966 ◽

Cited By ~ 1

Author(s):

Li Qiang Zhang ◽

Min Yue

Keyword(s):

Image Analysis ◽

Collision Detection ◽

High Speed ◽

Tool Path ◽

Geometric Model ◽

High Speed Machining ◽

Analysis Method ◽

Geometric Information ◽

Detection Approach ◽

Five Axis

Collision detection is a critical problem in five-axis high speed machining. Using a combination of process simulation and collision detection based on image analysis, a rapid detection approach is developed. The geometric model provides the cut geometry for the collision detection and records a dynamic geometric information for in-process workpiece. For the precise collision detection, a strategy of image analysis method is developed in order to make the approach efficient and maintian a high detection precision. An example of five-axis machining propeller is studied to demonstrate the proposed approach. It has shown that the collision detection task can be achieved with a near real-time performance.

Download Full-text