یک سیستم اطلاعات جغرافیایی چند هدفه برای انتخاب مسیر حمل و نقل زباله های هسته ای

Nuclear power is widely used throughout the world today. Functioning nuclear power plants produce large quantities of radioactive wastes needing to be transported to safe sites for proper management. With public emphasis on environmental protection and concern for safe transport of nuclear wastes, the problem of selecting an appropriate route for transporting nuclear wastes is a vitally important issue. The aforementioned route selection problem involves conflicting objectives among interested parties; therefore, we develop a multi-objective geographic information system (GIS) with ESRI ArcView GIS 3.x interface to practically support the involved parties for such a multi-objective route selection problem in engineering practices.

Currently, nuclear power is generally used worldwide; however, radioactive wastes are inevitably combined with nuclear power benefits. If produced nuclear wastes are not properly transported to safe sites for strict management, the natural environment will be destroyed from pollution. Since the public is concerned with safe transport of nuclear wastes [1], [2], [3], [4], [5] and [6], the problem of choosing an appropriate route for nuclear waste transport should be handled with caution. Route selection for nuclear waste transport becomes a decision for transporting hazardous materials [7] and [8]: the model developed by List and Mirchandani [9] is the most similar case to what we proposed in this study. They consider three objectives: total transportation cost, total risk, and equity of the risk imposed. Although many multi-objective models for transporting hazardous materials are proposed, the multi-objective transportation decision is not widely approved among practitioners so far. This is because: (a) these models are validated by a small road-network instead of an actual road-network, and the cost of collecting actual road-network attributes is very high; (b) most practitioners are not familiar with the complicated models; and (c) lack of an effective interface to integrate multi-objective programming and transportation planning in practice. The geographic information system (GIS) is specially designed to assist in finding solutions to geographic problems by computer [10] and [11], e.g., location problems, shortest route problems, distribution patterns of people, etc. In short, the GIS allows both practitioners and theoreticians the opportunity to grab large chunks of earth's surface and explore them around in their hands. Since GIS can effectively storage actual road-network data, we can develop a customized application for an actual road-network with GIS assistance. As suggested by DeMers [12], the GIS encourages us to roll our world in much the same way that geographers, naturalists, and explorers have done, but with a much more precise set of tools. This paper successfully integrates the theoretical aspect (multi-objective shortest route problem), with the practical aspect (actual road-network attributes of GIS), to plan a multi-objective route for transporting nuclear wastes using a practical rather than simplified road-network. Our study is organized as follows: Section 2 illustrates the problem characteristics and the real transportation network for nuclear waste transport in Taiwan. In Section 3, the shortest route model with multiple objectives is established and the concept of multi-objective GIS for selecting such a route for transporting nuclear wastes is introduced. In Section 4, an actual case of Taiwan is used as a numerical example for resolution and discussion. Finally, conclusions and recommendations are presented in Section 5.

The main contribution of this paper is focused on “actual application” for testing the technical feasibility of multi-objective GIS, rather than “theoretical originality” for proposing new multi-objective models for resolution. We successfully developed the multi-objective GIS for route selection of nuclear waste transport, and this system is practically approved by the INER in Taiwan. As the public becomes increasingly concerned about environmental conscience and transportation safety, our efforts are beneficial to the public and the Taiwan government. However, technical success does not necessarily guarantee successful execution of our multi-objective GIS in practices. This means, the planned result of Fig. 5 is not the finally determined route so far. Why? The difficulty is resulting from the political scope rather than the technical scope: the multi-objective GIS is really brilliant to the INER, but the INER will conditionally apply our approach on the nuclear waste transport in the very near future if we can continuously release our model assumptions and resolve the following questions as soon as possible: for example, what information do participants need to engage in the process? Who should be participating currently in waste cleanup dialogues? Perhaps more important, who is missing from the discussion? Are the current interested parties adequate and sufficient to present the public so as to make a good transportation decision? Is the congress really favored of such a MCDM process that makes each route of nuclear waste transport transparent? Although the concept of multi-objective GIS is approved by the INER, the questions just raised relate to qualitative rather than quantitative issues discussed here—any final decision, i.e., any finally determined route of nuclear waste transport in Taiwan is very sensitive in politics to the public. Therefore, the INER continuously pushes us resolving the proposed difficulties above in the coming project before such a system could be actually launched to the public. We believe our proposed model is a simple interface to integrate the public, the scholars (researchers): who are major in multi-objective modeling and the practitioners: who are major in the administrative affairs. We still negotiate with the INER for the coming research project in order to actually implement our idea by conquering the aforementioned difficulties. Since calculating the transportation risk is a tough task, the second and the third objective of our study could be better modified to our current ideas. ArcView 3.x is promoted to ArcView 9.x so far as we know, and ArcView 9.x can directly communicate with Visual Basic and C; therefore, some evolutionary algorithms or fuzzy concepts of MCDM might be included.