Browsing by Author "Liu, Kun-Hsing"

Now showing 1 - 3 of 3

Parallelised branch-and-bound algorithm for raster-based landfill siting
(2013-03-01)
Liu, Kun-Hsing
;
Kao, Jehng-Jung
;
環境工程研究所
;
Institute of Environmental Engineering
Landfill siting analysis is complex when compactness and other factors are simultaneously evaluated. In our previous work, a raster-based mixed-integer model was proposed to overcome this difficulty, and a C program was further developed to improve the computational time for solving the raster-based model. In this study, an enhanced parallelised branch-and-bound algorithm was proposed to shorten the solving time further. A parallelised computing environment with five computers was established for implementing the proposed algorithm. For comparison purpose, the un-parallelised algorithm was also tested on a single computer. The results show that the parallelised algorithm and computing environment can increase the speed by about three to seven times, while compared to the original algorithm implemented on a single computer.
Planning for hazardous campus waste collection
(2011-05-15)
Liu, Kun-Hsing
;
Shih, Shao-Yang
;
Kao, Jehng-Jung
;
環境工程研究所
;
Institute of Environmental Engineering
This study examines a procedure developed for planning a nation-wide hazardous campus waste (HCW) collection system. Alternative HCW plans were designed for different collection frequencies, truckloads, storage limits, and also for establishing an additional transfer station. Two clustering methods were applied to group adjacent campuses into clusters based on their locations, HCW quantities, the type of vehicles used and collection frequencies. Transportation risk, storage risk, and collection cost are the major criteria used to evaluate the feasibility of each alternative. Transportation risk is determined based on the accident rates for each road type and collection distance, while storage risk is calculated by estimating the annual average HCW quantity stored on campus. Alternatives with large trucks can reduce both transportation risk and collection cost, but their storage risks would be significantly increased. Alternatives that collect neighboring campuses simultaneously can effectively reduce storage risks as well as collection cost if the minimum quantity to collect for each group of neighboring campuses can be properly set. The three transfer station alternatives evaluated for northern Taiwan are cost effective and involve significantly lower transportation risk. The procedure proposed is expected to facilitate decision making and to support analyses for formulating a proper nation-wide HCW collection plan. (C) 2011 Elsevier B.V. All rights reserved.
Service Distance and Ratio-Based Location-Allocation Models for Siting Recycling Depots
(2010-04-01)
Kao, Jehng-Jung
;
Wen, Long-Mao
;
Liu, Kun-Hsing
;
環境工程研究所
;
Institute of Environmental Engineering
Siting appropriate locations for placing recycling depots is essential for promoting an efficient recycling program. This study develops three optimization models to facilitate siting analysis for district-based, district open, and nondistrict situations. An enhanced model to improve drawbacks of locating recycling depots that mainly serve residents in adjacent districts using the district open model is also proposed. Three factors of service distance, local service ratio, and service ratios for different distance ranges are used to compare the effectiveness of alternatives obtained from different models. A case study involving 16 city districts is implemented to demonstrate the applicability of the proposed models. Findings show that the district-based alternatives have best overall service distance and service ratio, but with a poor local service ratio. The enhanced model obtains alternatives that achieve good local service ratio with acceptable service distance and service ratios for different distance ranges.