Combining Artificial Neural Network- Genetic Algorithm and Response Surface Method to Predict Waste Generation and Optimize Cost of Solid Waste Collection and Transportation Process in Langkawi Island, Malaysia

Abstract

Solid waste management is an important component in the environmental management system. Due to high fluctuations of the amount of the produced waste in langkawi because of tourism in area, the use of neural networks is appropriate method to predict the amount of the produced waste based on non-linear and complex relationships between inputs and outputs. Collection and transportation of solid waste devote most part of municipality budget about 60% in area. The purposes of this research are to develop a model to predict the generation of solid waste and to reduce the cost of collection and transportation for solid waste management. This research has used the artificial neural network (ANN) and response surface model (RSM) to predict solid waste generation and to optimize the cost of waste collection and transportation. The authors believe that this approach will assist the authorities to determine the amount or quantity of solid waste generated over time. It will also assist the authorities to optimize cost, design appropriate and cost effective measures to collect and transport solid waste. This will improve environmental conditions and the cost saved could be used to provide other important services. We used time-series data with multiple input variables to perform the analyses. The results showed that use of variety of inputs data decreased the number neurons in hidden layer, which reduced the calculations performance and point of dimensionality, and increased accuracy in prediction the amount of produced waste; and whereas there is an increase in solid waste generation from 7825.7 tons (T) in 2009 to 8030.68 T in 2011; cost reduction amount is 10.64%. The methodology or an adapted form of the methodology can be applied to other fields, subject to a study of the requirements in each place.

The use of a recurrent neural network model with separated time-series and lagged daily inputs for waste disposal rates modeling during COVID-19

“… Antanasijevic et al (2013) 26 countries in Europe Back-projection BPANN and general regression GRNN Annual indicators of sustainability (gross domestic product, domestic material consumption and resource productivity) from 2000 to 2002 GRNN model proved to be significantly better than the more traditional BP model. Shamshiry et al (2014) Malaysia ANN, genetic algorithm, response surface method Weekly waste generation data, number of trucks, number of personnel, number of tourists, fuel cost Combined ANN and RSM to predict or forecast solid waste generation and optimize the cost of waste collection and transportation Younes et al (2015) Malaysia Modified Adaptive Neural Inference System (MANFIS) GDP, electricity demand, employment, unemployment, waste generation, population (1981-2013 data) The best input variables were people in age groups 0-14, 15-64, and +65 years, and the best model structure was 3 triangular fuzzy membership functions and 27 fuzzy rules Azadi and Karimi-Jashni (2016) 20 cities in Iran ANN, MLR Used monthly population, waste collection frequency, temperature, altitude, waste generation data between 2009-2010 ANN is better than MLR in predicting the mean seasonal municipal solid waste generation rate Abbasi and Hanandeh (2016) Australia SVM, ANFIS, ANN, kNN Used monthly waste generation data from July 1996 to June 2014 SVM reliably predicted monthly MSW generation, ANFIS predicted the most accurate forecasts of the peaks, kNN was successful in the monthly averages of waste quantities prediction Kumar and Samadder (2017) India MLR Monthly household SW generation rate in 2016 and the socioeconomic parameters R2 was 0.782 for biodegradable waste generation rate and 0.676 for non-biodegradable waste generation rate Kannangara et al (2018) Ontario, Canada Decision trees and neural networks Residential MSW quantities, socio-economic (earnings and income, education, employment, industries and occupations, dwelling characteristics, household characteristics, workplace and demograph...…”

Performance Improvement of Machine Learning Model Using Autoencoder to Predict Demolition Waste Generation Rate

“…In this regard, several studies have employed AI technology for predicting the volume of C&DW or MSW generated. For instance, an artificial neural network (ANN) algorithm has been applied to predict WG [11][12][13][14][15][16][17][18][19]. In addition, a support vector machine (SVM) algorithm has been utilized to develop a WG prediction model [12,13,18,[20][21][22][23][24][25][26][27][28], and several studies have developed a WG prediction model using linear regression (LR) [11,13,[29][30][31][32][33][34][35][36].…”

Artificial Neural Network Applied in Forecasting the Composition of Municipal Solid Waste in Iasi, Romania

“…They used nonlinear autoregressive with external (exogenous) input model (NARX) for prediction (Younes et al, 2015). In another study the authors used data from 2004 to 2009 and applied ANN and response surface model for solid waste generation forecasting (Shamshiry et al, 2014). The input variables considered by Shamshiry et al (2014) were: fuel consumption, 4-ton truck, 10-ton truck, number of trips made by the trucks to the landfill, number of times the personnel entered into the landfill, number of tourists, and salary per worker per day.…”