Architecture and sustainability assessment of cable multi-state monitoring system driven by green computing

Abstract

The rapid development of urban infrastructure requires the development of advanced monitoring systems for the continuous assessment of cable health to guarantee operational reliability, safety, and long-term sustainability. The research proposed a novel Cable Multi-State Monitoring System (CMSMS) leveraging the computational capabilities of Edge compute allied with Green Computing principles to recover energy consumption and monitor. The suggested system employs heterogeneous sensors for real-time data acquisition, capturing dangerous cable parameters such as temperature, vibration, and strain. Pre-processing includes handling missing values and feature extraction using Discrete Wavelet Transform (DWT) to enhance the quality and relevance of the sensor data. Edge devices, clearly low-power platform such as Raspberry Pi and NVIDIA Jetson, serve as dispersed nodes for local data processing. These strategies permit the categorization of cable conditions into three discrete states: normal, degradation, and fault prediction, thereby support early detection of potential cable failure. For fault detection, the system includes an Extreme Gradient Boosting (XGBoost) model to adeptly handle complex, non-linear interdependencies with sensor data. Its parallel processing capabilities significantly improve computational competence, making it well-suitable for edge-based application. To further reduce energy consumption, the Shuffled Frog Leaping Algorithm (SFLA) is employed for the optimization of system parameters; ensure a balance between computational performance and energy efficacy. A comprehensive sustainability valuation is conduct to evaluate system performance, converging on energy consumption, processing speed, and fault detection accuracy. The simulation result implement using python, SFLA-XGBoost method outperformed the existing method in CMSMS fault identification, as established by it’s almost "higher accuracy" of 99.50 %. The outcomes establish a considerable decrease in effective costs and energy usage while preserve high precision in fault classification and detection. The recommended CMSMS design suggest a scalable, reliable, and energy-efficient key that is suitable across several industry, including telecommunications, power distribution, and smart cities.