Implementación de un sistema para la evaluación de amortiguadores de vehículos usando sensores IOT y modelos de redes LSTM

Abstract

This research proposes the design and development of an intelligent system for assessing the condition of shock absorbers in light vehicles, integrating recent technological advances such as the Internet of Things (IoT) and Long Short-Term Memory (LSTM) neural network models. This structure is based on the capture of dynamic data from inertial sensors, capable of recording critical variables such as vertical acceleration, vibration frequencies, and dynamic chassis displacements during driving. The data is transmitted in real time to a platform, a server, via the MQTT communication protocol and stored in NoSQL databases optimized for time series. A preprocessing process was then performed, including data cleaning, normalization, and segmentation, to ensure data quality before being entered into the artificial intelligence model. An LSTM network was subsequently developed and trained to identify patterns in the sequential data and predict anomalous behaviors that could indicate shock absorber wear or failure. Finally, the trained model is integrated into a real-time monitoring platform, achieving a 99% efficiency rate, considering that multiple data subsets are evaluated to determine the shock absorber's condition. This approach not only enhances vehicle safety and performance but also optimizes operating costs, contributing to the digitalization of automotive maintenance.