垃圾公車清運系統之收集點位置規劃

Abstract

隨著經濟的快速發展，人民生活水準的提升，垃圾清運工作除了成本效益的考量外，服務品質之提升已成為地方政府努力的方向。新竹市因而推行垃圾公車制度以提升服務品質，然而垃圾收集定點之選定是影響清運成本及服務效率的重要因子，選擇收集點的數量與位置不恰當，會增加清運成本，且導致民眾的滿意度降低。然而所選定收集點往往因主觀直覺的選定，令收集點之位置與數目未能有效決定。本研究以民眾行走距離長短為服務品質之主要評估因子，建立優選模式，除了應用過去研究者以Location Set-Covering (LSC)及Maximal Covering Location (MCL)原則所建立的兩個模式，本研究並依據總行走距離最短為目標建立改良的Shortest Service Location (SSC)模式，令所得結果能更具效率。 本研究針對新竹市垃圾公車清運系統案例，依上述三個模式選取收集點，並依據民眾平均行走距離、總行走距離、收集點服務範圍等因子，並配合GIS圖形協助分析展示成果，以比較LSC, MCL及SSL三個簡易模式所得結果之差異性與優劣點。依案例結果及GIS圖形展示顯示， SSL模式雖然求解的時間較LSC與MCL模式長，但求解時間仍在可接受範圍內，能縮短平均行走距離約10%，且降低收集服務區覆蓋的情況；而在不同的行走距離區間下，需求點行走距離低於平均行走距離的數量，SSL模式也較LSC及MCL模式多；另外，考慮模式所選擇之收集點所服務的需求點數量來看，大於10個以上的收集點，SSL模式也較LSC與MCL模式有效率。上述結果顯示本研究所建議之SSL模式能選定位置較具效率之收集點，提高民眾傾倒垃圾之便利性，提升服務品質。

With rapid economical growth and increasing living standard, in addition to the cost reduction, service quality enhancement has become an important issue for municipal solid waste collection (MSWC). The Hsinchu City therefore initiated a waste bus system to promote MSWC service quality. For a successful waste bus system, selection of proper collection points is a critical factor. Inappropriate locations or number of collection points will increase the cost and may not be accepted by the general public. Currently, collection points were subjectively selected which result in the number and locations of the collection points being chosen inefficiently. In this study, other than applying the Location Set Covering (LSC) and Maximal Location Covering (MCL) methods proposed in previous literature, an enhanced model, Shortest Service Location (SSL), was proposed based on minimizing total walking distance. The results obtained from these models are intended to improve the decision of the number and locations of collection points. The waste bus system of Hsinchu City was the case analyzed in this study. Comparison of LSC, MCL and SSL models were made based on average and total walking distances from demand points to collection points and the service area of a collection point. A geographical information system (GIS) was utilized to facilitate data analysis and presentation. According to the results obtained by the models and illustrated by the GIS, the SSL solution, while compared with LSC and MCL solutions, can reduce average distance about 10% and decrease the overlapping of service areas, although its computational solving time is longer than those for other two models. For demand points whose walking distances to collection points being shorter than the average, the number of such demand points of the SSL solution is significantly larger than other two solutions. Furthermore, for collection points that served more than ten demand points, the number of such collection points of the SSL solution is also larger than the LSC and MCL solutions. The SSL model can choose efficient collection points, thereby increasing the service quality of MSWC.