都會區基地開發常由於現有交通設施不足，對鄰近地區造成嚴重交通衝擊，然而其事前之規劃與評估中，詳細的交通衝擊評估(Traffic Impact Assessment, TIA)需耗費龐大的調查分析成本，故本研究期以較簡單之迴歸分析方式，針對都會區一般性中小型基地開發，有效進行開發之道路交通衝擊預測。 本研究乃以非線性迴歸分析(Nonlinear Regression Analysis)為主軸建立道路交通衝擊預測模式，而參考過去文獻與施行經驗及交通衝擊相關之各項規定，並考量資料蒐集之可行性，選定各項開發案基地主、客體相關變數進行模式初步構建，由於變數間存在非線性關係之考量，故需應用變數變換之方式先行轉換為線性關係，再進行相關參數之校估，最後建立一完整之非線性交通衝擊預測模式。實證對象為台北縣民國96與97兩年度已定稿交通衝擊評估書之開發案件，結果顯示其迴歸預測模式能力良好，但由模式精準度可見實證之樣本資料特性不佳。 未來可應用於其他都會區一般性之基地開發衝擊預測分析，並可搭配鄰近地區之交通現況或其他衝擊評估準則，且結合各地方所建構之地理資訊系統及空間資料庫以方便樣本取得，提供基地開發交通衝擊之初步審查或評估的參考依據。 In the metropolitan areas, the existing shortage of transport facilities of site development often caused serious traffic impact in the neighborhood. However, the detailed traffic impact assessment (TIA) on the prior of its planning and evaluation would cost a huge amount of money on surveying and analysis. Therefore, regarding the common site development of a metropolis, the study hoped using a simpler way of regression analysis to establish a forecasting model and estimate the road traffic impact more efficiently. The mainstream of the study were based on the nonlinear regression analysis which established the forecasting model of road traffic impact, and it referred to the past literatures, the implemented experiences and every related provision about traffic impact. Furthermore, it considered the feasibility of collected data and constructed model initially with related variables between major and objective bases which selected from every development project. Due to the consideration of existing nonlinear relationship between variables, it required to transform the used variables into a linear relationship by variable transformation, and then estimated the relevant parameters, finally, established a complete nonlinear traffic impact forecasting model. The experimental objects were the final drafts of traffic impact reports from Taipei County site cases at 2007 and 2008, the results revealed that this regression forecasting model could get a great outcome. However, the poor characteristics of the sample data could be seen from the experimental objects through the model accuracy. It could be applied to forecast the impact of general site development analysis at other metropolitan areas in the future, and cooperated to the traffic present situation or other impact evaluation criteria, and even combined with each local geographic information system and spatial database, so that it facilitated to acquire samples and provided a basis of preliminary investigation and assessment on traffic impact from site development.
