鎮暴決策支援系統建立之研究

Abstract

當暴動事件一旦發生，決策者必需在時間壓力下搜集各方面的資訊以作出 適當的鎮暴決策，其中包含如何決定防守據點、人力和武器配備、調派兵 力及研擬部隊行進的路線等。而此決策過程的完成有賴於一系列智慧決策 程序的達成方可成功。本研究中作佈兵決策是一種專家的經驗知識，應用 專家系統應可解決此問題；調派兵力是資源配對問題，必須發展適合鎮暴 問題之特殊資源配對模式，以支援決策者在調派兵力上的運作；而研擬部 隊行進的路線是典型的車輛繞行問題，必須評估現有的演算式加以修改使 成為適合鎮暴環境的車輛繞行模式；除此之外，在佈兵、調派兵力、研擬 部隊行進路線等決策過程中，決策者必須隨時查閱地圖以獲得相關的地理 資訊，在此我們必須根據鎮暴的需要，發展適合鎮暴問題的地理資訊系統 介面，以方便決策者作查詢之用。是故本研究擬發展一鎮暴決策支援系統 ，將以地理資訊系統為基礎，整合專家系統、乏晰推理、人機交談模式、 探試模型，並建立車輛繞行、最短路徑及資源配對等人機交談模式以支援 整個鎮暴事務的運作。 If a riot breaks out in city,it should be handled carefully by decision maker（ＤＭ）. In the repression process,a DM has to consider many factors in limited time, in which various kinds of resources has to be dispatched effectively and properly to repress the riot. The Problem is usually solved by DM's personal experiences and which it is a very time-consuming task. Therefore, this research `develops a prototype of a decision support system to support the DM over entire riot- repression process. The proposed DSS based on a GIS user interface integrates expert's knowledge in fields of expert system, heuristics,man-machine interfacing and mathematic model in its architecture. Various algorithms are also improved to meet practical situations. The DSS can collect and manage the information of the riot-repression to support the DM dynamically, in which evaluated alternatives of resource mamagement polices and the best routing suggestions are included.