應用兩階層規劃模型於差別補貼費率訂定之研究－以台灣廢印表機資源回收系統為例

An Application of Bi-level Programming to the Settings of Differential Subsidy – A Case of the Subsidy Dicision to Recycled Printers in Taiwan

張雅翔(Ya-Shiang Chang)

指導教授 ： 時序時

共同指導教授 ： 鄭啟斌(Chi-Bin Cheng)

淡江大學/管理學院/管理科學學系碩士班/碩士(2012年)

https://doi.org/10.6846/TKU.2012.00855

摘要

為使環境汙染成本內部化，我國行政院環境保護署(簡稱環保署)於1997年推動資源回收計畫，並建立資源基金制度，委由資源回收基金管理委員會(以下簡稱基管會)進行基金管理。基管會對於公告應回收廢棄物之製造及輸入業者依核定費率收取回收清除處理費用後，再扣除行政、稽核等相關成本，其餘基金則補貼給回收處理業者，以獎勵回收工作並期望降低環境污染。
目前基管會每年的可用基金額度約有20億元新台幣。然而自制度實施以來，各項費率之多寡始終為各方利益關係人爭議的焦點，尤其又需考慮回收效率與公平性原則，因此本研究嘗試以兩階層分散式規劃(Bi-level decentralized programming, BLDP)用於差別補貼費率訂定之研究以期解決爭議，並以廢資訊物品中的廢印表機為對象進行探討。其中，上階為基管會，經分析其目標有四：補貼費用極小化、回收率極大化、徵收費用極小化以及行政成本極小化；下階則為回收處理廠，其目標為利潤最大化。由於兩者目標不一，透過上下兩階層衝突互動後期望能得到最適分配。此外，由於模型無法直接求解，本研究使用模糊趨近法(Fuzzy approach)以及直覺模糊趨近法(Intuitionistic fuzzy sets)求取妥協之最適解。除了費率問題外，由於印表機廢棄物總量不變，目前單一費率制度方式易造成處理技術及能力停滯不前，因此本研究導入回收處理廠分級概念，以差別補貼方式激勵回收處理廠升級，期增加資源化價值與降低環境污染。目前補貼費率由基管會統一訂定，當市場存在單一費率時，隱含供給與回收處理廠獲得之效益並無差異。故現行單一缺乏變化之費率制度無法凸顯廢資訊物品市場回收處理能力對資源再利用的影響，因此面對不同處理量之回收處理廠本研究嘗試在可用資金範圍內訂定差別補貼費率，將改善目前單一費率所造成的不合理現象，且符合經濟效益。
　　本研究結果顯示如欲提升回收率，目前回收清除處理補貼費率以及回收獎勵金皆有調整空間。而本研究所使用之模糊趨近法以及直覺模糊趨近法解值較能符合基管會以及回收處理廠之期望，也較符合目前現況。回收率雖然與目前回收率相差不大，但卻能達到鼓勵回收處理廠升級之目的，不僅基管會資金流動效率更好，回收處理廠亦可提升其處理能力。因此，本研究結果可提供基管會做為未來修改費率時之參考。

關鍵字

兩階層分散式規劃 ； 印表機回收 ； 回收清除處理補貼費率 ； 差別補貼資源再利用率 ； 模糊趨近法 ； 直覺模糊趨近法 ； 回收率

並列摘要

To internalize environmental pollution cost, the Environmental Protection Administration (EPA), Executive Yuan, R.O.C, promoted the resource recycling program in 1997, and established the resource funding system, of which funding was to be managed by the Recycling Fund Management Board (RFMB). After collecting recycling and treatment fees according to regulated rates from manufacturers and importers of wastes announced to be recycled, RFMB deducts administration and auditing costs and then subsidizes the remaining fund to recycling firms as incentive for their work and in hope of reducing environmental pollution. The annual fund quota available to RFMB is currently around two billion NTdollars. However, since the enforcement of this system, the amount of various rates has been an issue disputed by stakeholders, especially as recycling efficiency and social fairness are considered. Thus, this study uses Bi-level decentralized programming (BLDP) to investigate differential subsidy rates in an attempt to solve the dispute; waste printers are used as subject for discussion.
To increase the waste printer recycling rate and the capabilities of recycling and treatment plants in Taiwan, this study uses BLDP to determine the optimal subsidy rate. The high level is RFMB which has three goals: minimize subsidy fees, minimize and maximize recycling rate, and minimize the administration cost. The low level is the recycling and treatment firms, of which the goal is to maximize profits. As the two levels have different goals, it is anticipated that an optimal allocation may be achieved after conflicting interactions. This study uses fuzzy approach and intuitionistic fuzzy sets to determine an optimal solution. In addition to the rates problem, the single rate system currently applied tends to cause stagnancy in treatment technique and ability because the total amount of printer wastes does not change. Thus, this study induced the concept of recycling and treatment firms grading and used differential subsidy to stimulate these firms to upgrade in hope to increase resource recovery rate and reduce recycling treatment wastes. The subsidy rates in this study was established by RFMB, so when single rate exists in the market, there is no difference between the benefits acquired by recycling and treatment firms and the implicit supply. Therefore, the single rate system currently adopted, which lacks change, is unable to highlight the impact of information waste market's recycling and treatment abilities on reusable resource. For recycling and treatment firms with different treatment quantities, this study attempted to stipulate a differential subsidy rate within available fund. This will improve the unreasonable phenomenon caused by the single rate system and conform to economic benefits.
The results of this study indicate that the subsidy rate for recycling, disposal and treatment and the recycling incentive may be adjusted to increase recycling rate. The results obtained by intuitive fuzzy approach conform better to not only the expectations of RFMB and recycling and treatment firms but also current situation. Although the recycling rate will be similar to the current rate, the goal of encouraging upgrading of recycling firms may be fully achieved. This will not only allow better efficiency of RFMB's fund flow, but also the profits of recycling firms will increase much more than current rates. Further, though this study allows all recycling firms to be capable of upgrading, it is suggested that the subsidy rate of recycling, disposal and treatment be slightly decreased if differential subsidy is desired to achieve the competing objective of recycling firms. The results of this study can provide a direction for RFMB resetting subsidy in the future.

並列關鍵字

Bi-level decentralized programming ； Fuzzy approach ； Intuitionistic fuzzy sets ； Printers recycling ； Treatment fee ； Differential subsidy ； Resource recovery rate ； recycling rate

參考文獻

14. 陳永仁，黃揮原（2003），廢棄物回收政策－台灣經驗，台灣衛誌，第二十二卷第四期，318-326。

12. 郭育孟（2009），廢棄電子資訊物品逆物流回收之需求分析與預測研究－以台灣地區為例，淡江大學運輸管理學系運輸科學碩士班碩士論文，未出版，新北市。

21. 彭孟吟（2008），「兩階層數學規劃費率制訂之研究－以台灣玻璃容器資源回收機制為例」，淡江大學管理科學研究所，碩士論文，新北市。

。行政院環境保護署96年度委託研究計畫，未出版，台北市。

26. 溫麗琪（2005b），廢棄物回收處理費之回收處理市場及誘因效果。人文及社會科學集刊，第十七卷第三期(94/9)，491-520，台北市。