On the Comparative Analysis of Various PSO Variants and Hybrids for Waste Collection Route Optimization

Efficient waste collection is essential to reduce fuel consumption, carbon emissions and unnecessary waste collection trips. Diesel-powered collection trucks emit large amounts of pollutants into the air daily. In addition, improper routing and scheduling of collection trucks may lead to an overflow of waste, which could affect the cleanliness of the environment and may pose health risks to citizens. This paper reviews various swarm intelligence-based route optimization algorithms, including ant colony optimization (ACO) and particle swarm optimization (PSO), specifically for waste collection. Various PSO variants including standard PSO, extrapolation PSO (e-PSO), global-local best PSO (gl-PSO), and faster PSO (f-PSO) were considered for evaluation. A novel hybridization strategy was introduced in this paper, which was used to form hybrids of these variants. The efficacy of these PSO variants and created hybrids was initially validated through four standard benchmark problems. These algorithms were then simulated for waste collection route optimization application, considering shortest route, weight of the bins and obstacles to obtain the most fuel-efficient path. From the simulated results, a comparative analysis was conducted using statistical analysis, Analysis of Variance (ANOV A) tests and parametric two-way t-test. From the analytical and statistical results, the supremacy of hybridized e-PSO algorithms, specifically a hybrid of e-PSO,f-PSO and gl-PSO, in determining the most fuel-efficient route for optimizing waste collection in intelligent waste management systems is demonstrated in this paper.