Machine Learning Approaches for Predicting Ignition Delay in Combustion Processes: A Comprehensive Review

Abstract

This review explores machine learning approaches for predicting ignition delay in combustion processes. Ignition delay is a vital parameter in optimizing the engine design, fuel formulations, and combustion efficiency. The review examines the applications of artificial neural networks (ANNs) and convolutional neural networks (CNNs) in various combustion processes and equipment, such as engines, boilers, and rapid compression machines. The differences between ANNs and CNNs are discussed, highlighting their capabilities and limitations. Numerous studies are presented, demonstrating the successful application of neural networks in predicting ignition delay for different fuels and engines. Overall, machine learning approaches show great promise in accurately predicting the ignition delay and advancing energy utilization.

Development of a Robust CNN Model for Mango Leaf Disease Detection and Classification: A Precision Agriculture Approach

“…The pooling layer takes a very small region of the Output given by the convolutional layer and further down samples it to give a single output. The last two layers in a typical CNN model are the FC layer and the output layer. , The FC layer receives the output from the final pooling layer and generates a final output to be processed further in the output layer. − The working of the convolutional layer (eq ), activation layer (ReLU) (eq ), pooling layer (eq ), fully connected layer (eq ), and the output layer (Softmax) (eq ) can be mathematically represented as below.…”