小客車成長預測之研究

Abstract

鑒於小客車快速成長帶來之交通擁擠、停車位不足、環境品質惡化、 廢棄車輛棄置、道路闢建及能源諸問題；且過去研究並未依施政期之長短 ，分別以適當方法預測小客車之成長，本研究爰依短（一至二年）、中（ 二至五年）、長期（五年以上）構建預測模式，以作為政府施政之參據。 短期蒐集民國67至83年小客車持有數、平均國民所得、人口數之月資料或 季資料，來構建ＡＲＩＭＡ及轉換函數之時間數列模式；中期則以民國66 至82年之小客車持有數及相關社經變數，構建多元迴歸模式；長期則利用 群落分析及判別分析，篩選與我國地理社經背景類似之國家，從國外實證 探討影響飽和率之因素，繼之以平均國民所得為影響變數構建羅吉斯模式 。 經研究發現：短期小客車持有數有明顯的季節性變化；中期小客車 持有數與影響變數間並非呈線性關係，且以平均國民所得、公共運輸延人 公里、小客車進口關稅及服務業就業人口數較為顯著；長期則以公共運輸 比例、小客車取得成本及小客車使用成本，構建飽和率預測式，得知我國 未來小客車持有率為每人0.36輛。 此外，本研究建立之預測模式經與 世界先進國家之小客車持有率比較，短期預測模式有低估之現象，中期預 測模式則明顯的高估，長期預測模式似較為合理；且未來我國小客車發展 可參考日本之模式。 For the rapid growth of passenger cars，it brings us many problems，such as traffic congestion，shortage of pparking spaces，environmentalpollution，deserted vehicles，road building ，energy and so on. Previousstudies never use proper methods to forecast the growth of passenger c-ars according to the period of time. The thesis is to construct the sh-ort-term(one to two years)、middle-term(between two and five years)、l-ong-term(over five years) forecasting models separately for the policyplanning of government sectors. In short-term it constructs ＡＲＩＭＡ and transfer function modelsof time series. Time series data from 1978 through 1994 including the number of passenger cars、 population and average national income are u-sed to calibrate these models. In middle-term it constructs two multiple regression models to for-ecast the growth of passenger cars. Time series data from 1977 through1993 including the number of passenger cars and social-economic factorsaffecting the ownership of passenger cars are used to calibrate these models. In long-term it considers the factors affecting the saturation rateby the positive study of foreign countries，then to set it for Taiwan area，and select average national income to construct logistic model. Three models are finally obtained separately. The thesis discovers:In short-term forecasting model，the number of passenger cars have sea-sonal change obviously；in middle- term forecasting model，it indicatesfour explanatory variables- average national income，passenger-mile of public transportation ，car tariff and population of service employmentare significant ，and also found that the number of passenger cars and each of the explanatory variable are nonlinearly related. In long-term， a forecasting model of saturation rate is obtained in which three e-xplanatory variables-ratio of public transportation， acquiring cost ofpassenger cars and use cost of passenger cars are included.We get thefuture rate of car-ownership in Taiwan area is 0.36. Furthermore，we compare the forecasting models with advanced coun-tries in rate of car-ownership，it is found the short-term forecastingmodel underestimates the rate，middle- term forecasting model overesti-mates the rate，long-term forecasting model can be accepted reasonably.The result suggests we can follow the trace of Japan to develop the therelative policy of passenger cars.