Browsing by Author "Otoo, Dominic"

Now showing 1 - 2 of 2
  • Thumbnail Image
    Item
    Capacitated Arc Routing Problem: Collection of Solid Waste at Kwadaso Estate, Kumasi
    (2012-06-20) Otoo, Dominic
    Solid waste management in most cities in Africa has become more challenging in which Ghana is not an exception. This is as a result of industrialization and urban migration, it has therefore become necessary to formulate a model which can be used by city authorities, Governments and waste management groups alike to use a minimal distance to collect more waste in an area. It is upon this basis why this work was done to come out with a solution that can be used to minimize the tour in a collection area and also give some sort of flow chart for the collection. In this study, we selected an area in Kumasi called Kwadaso estates which have 157 collection points and 588 240 litre bins. We first, found all pair shortest path and partitioned the entire collection points into smaller clusters based on the capacity of the vehicle and then used Ant Colony Optimization (ACO) to find the minimum tour in each cluster, which will also serve as a flow chart to guide collection in a cluster. Our study has improved the total distance by about 40% as compelled to the existing figures given by the waste management group in the area (Zoom Lion Ghana Limited).
  • Thumbnail Image
    Item
    Multi objective node routing problem with time windows: an alternate approach to solid waste collection and disposal in developing countries
    (2015-04-15) Otoo, Dominic
    This thesis is concerned with the collection and disposal of solid waste in the developing countries where logistics for collection and improper road network is a challenge. This problem is modeled as a Capacitated Node Routing Problem with Time Windows (CNRPTW), the effect of smell from uncollected waste on the environment and the cost of transportation to the disposal sites. The first part of the study investigates the generation of waste per person per day in third class communities; the second part provides a proposed location model to optimally assign customers unto a zone base d on two primary parameters. The third part of the study provides a meta-heuristic method, which takes into account the vehicle capacity and time of collection. This is based on the improved Ant Colony Heuristic, enhanced by three parameters. The introduction of soft windows incorporates vehicle breaking time, crew lunch break and drop-off time. The fourth part of the study modeled the effect of smell from uncollected waste by incorporating factors such time and fraction of biodegradable component, the fifth part considered the implementation of a proposed model on fuel consumption in addition to the existing fuel consumption models. Computational testing is carried out on the test problems used in the literature on our improved models gave competitive results. The final part of the study deals with the implementation of our models on the real life solid waste generation, collection and transportation which exists in one of the nine sub-metropolitan areas in Kumasi Metropolitan Assembly in Kumasi. Competing results were obtained compared with the existing practices on the ground.