首页|Studies Conducted at National University of Defense Technology on Machine Learni ng Recently Reported (Analysis of the Combustion Modes In a Rocket-based Combine d Cycle Combustor Using Unsupervised Machine Learning Methodology)
Studies Conducted at National University of Defense Technology on Machine Learni ng Recently Reported (Analysis of the Combustion Modes In a Rocket-based Combine d Cycle Combustor Using Unsupervised Machine Learning Methodology)
扫码查看
点击上方二维码区域，可以放大扫码查看
原文链接
NETL
NSTL
By a News Reporter-Staff News Editor at Robotics & Machine Learning Daily News Daily News-Investigators publish new report on Ma chine Learning. According to news reporting originating from Changsha, People's Republic of China, by NewsRx correspondents, research stated, "Combustion mode a nalysis is essential to a rocket-based combined cycle (RBCC) combustor because i t may experience multiple combustion modes during the operation. In this study, a method based on an autoencoder and a K-means algorithm was proposed for combus tion mode analysis." Funders for this research include National Natural Science Foundation of China ( NSFC), National Natural Science Foundation of China (NSFC), Science and Technolo gy on Scramjet Laboratory. Our news editors obtained a quote from the research from the National University of Defense Technology, "Flame chemiluminescence images and schlieren images of three combustion modes observed in an RBCC combustor were used to evaluate this method. Two autoencoders that followed the same encoderdecoder architecture wer e developed separately to generate the latent space representations of flame che miluminescence images and schlieren images. In the latent space, the centroids a nd boundaries of different combustion modes were determined using the K-means al gorithm. Each autoencoder was trained using 750 images and tested using another 3000 images. The method achieved an accuracy up to 99% on both fla me chemiluminescence images and schlieren images. The images generated by the de coder suggested that the autoencoder captured the important features (e.g., prim ary reaction zone and shock wave) of the reacting flow field. The autoencoder de veloped for flame chemiluminescence images also successfully detected the combus tion mode transition during an ignition process, which suggested that it had the potential to monitor the combustion mode in a real time manner. However, the au toencoder failed on monitoring combustion mode transition when it came to the sc hlieren images because the optical access of the training data was not exactly t he same."
ChangshaPeople's Republic of ChinaAs iaChemiluminescenceCyborgsEmerging TechnologiesHealth and MedicineMach ine LearningNational University of Defense Technology
NETL
NSTL
