Smart Waste Management and Classification System for Smart Cities using Hybrid Random Forest and ANN Models

The Web and IoT paradigms are transforming due to the extensive integration of smart devices such as RFIDs, sensors, and actuators. This establishes the foundation for frameworks pertaining to Smart Cities. Intelligent services such as Smart Waste Management are facilitated by sensors embedded in urban surroundings that collect and monitor ambient data. The prompt collection, transportation, and disposal of refuse across several locations, supported by real-time data, is crucial for effective waste management. This research introduces a novel hybrid model that integrates RF with the FFA to enhance the accuracy of hourly global solar radiation forecasts, a crucial element in optimising energy efficiency within waste management systems. PCA and data normalisation enhance the efficacy of preprocessing and feature extraction. Optimising the RF settings with the FFA enhances the model's performance. The parameters encompass the quantity of trees and the number of leaves per tree. The RMSE, MAE, and MSE values for the proposed RF-FFA model were 13.76%, 11.48%, and 12.32%, respectively. It outperformed competing models in both speed and accuracy of predictions. The findings indicate that the model is effective for enhancing decision-making and forecasts in smart city applications, specifically regarding SWM.

Hali tarjima qilinmagan