Diagnostics of Ship Engines Based on Wavelet Neural Network and Image Scanning Using Programmable Logic Circuit

The article is devoted to a diagnostic system for ship engines based on a wavelet neural network and image scanning using a programmable logic circuit and considers a method for analysing multifractal wavelet models. The combination of wavelet neural networks with a programmable PLIC-based (programmable logic integrated circuit) real-time image processing platform has a significant potential for the purposes of non-destructive testing, which makes it possible to accurately diagnose faults and take effective measures for predictive maintenance, which in turn makes it possible to effectively increase safety and reliability of equipment and reduce maintenance costs. The article proposes an improved approach to the diagnosis of ship engines, which is based on a wavelet neural network and image scanning using a programmable logic circuit. Wavelet packet decomposition is a method for local time and frequency analysis. It gradually refines the signal at multiple scales through scaling and conversion operations, and it can automatically adapt to the requirements of time-frequency signal analysis to focus on any detail of the signal. It has the advantage of good diagnostic accuracy for information with different noise levels, as well as high reliability since image data from multiple engine signals is used.