道路交通噪音預測模式之研究

Abstract

本研究主要目的為瞭解各型道路、建築物型式、及車輛種類、數量與交通噪音位 準之關係，探討並建立都市道路交通噪音預測模式，作為道路交通噪音防制的參考。研究對 象為臺北市主要道路沿線的測定站。利用 Rion, NA-20, Rion, SV-73 測得噪音之 Leq、Lx ，所得資料以單因子變異數分析，複迴歸分析。主要結果如下： 1. 本研究之測量時段所測得臺北市道路交通工具數量， 以小型車（包括轎車）最多（ 50.71% ），其次機車（ 45.03% ）、中型車（ 7.02% ）、大型車（ 2.95% ）。 2. 道路噪音平均值 L �煄BL �t， Leq 各為 78.87、 73.01、 75.82dBA。 以 Leq 平均值 75.82dBA 與環保署所草擬的「環境噪音品質標準」路邊地區第三類標準值 75dBA 比較，發 現超越限定值，如果以最大值則嚴重噪音污染。 3. 一般而言，道路愈寬交通流量愈大，尤其 6 線快車道 2 線慢車道之交通流量。 機車數 量於快 6 慢 2 道路數量最多，而其他類型道路之機車數差距不大。 4. 以施氏模式（ Shih, 1980 ）使用於本研究，雖然達顯著水準，但解釋力只有 14% （ L �煄^、16% （ L �t）、11% （ Leq ），可能未符合施氏模式的使用條件。 5. 用狄拉尼（ Delany, 1972 ）模式及蓋諾威（ Galloway, 1969 ）模式， 則解釋力很低 ，尤其是使用蓋諾威模式，只有 10% （ L �煄^、12% （ L �t）、8% （ Leq ）。 6. 使用曾國雄（ Tzeng, 1981 ）模式之第 6 個預測模式， 則解釋力提高為 17% （ L �� ）、19% （ L �t）、13% （ Leq ），但尚嫌不高，可能是八年來車輛增加太多，變因增加 的緣故。 7. 參考施氏、曾氏、狄拉尼、蓋諾威等模式，並利用經驗法則，選出三個變項，作為預測 L�煄BL�t、Leq 值，解釋力各為 18%、20%、15%。模式如下： (1)□L��=55.46-0.001Q+1.892 log (Q□(1+P□))+5.965 log(Q□(1+P□)) R��=0.18 (2)□=50.12-0.001Q+1.614 log(Q□(1+P□))+5.899 log(Q□(1+P□)) R��=0.20 (3)□□=55.98-0.001Q+1.765 log(Q□(1+P□))+5.039 log(Q□(1+P□) R��=0.15 本模式只適用於本研究測定之道路與其時段（非上下班時間） 本研究建議禁止車輛隨意變換車道、淘汰舊車以使流通順暢，在建立類似模式時，應考慮內 車道與外車道車輛數分別計算、交通號誌之間隔、交通順暢與否等因素。

The main purposes of this study were: (1) to understand the relationship between the type of roads, the type of building structure, the number of various vehicles and traffic noise value, and (2) to establish Predictive Model of road traffic noise in urban area in order to help control road traffic noise. The study samples were the test positions along the main roads in Taipei city. Rion, Sound level Meter (Rion, NA-20) and Rion, Noise level analyzer (Rion, SV-73) were used to measure Leq and Lx. One-way ANOVA, multiple correlation & reqression analysis were used to analyze the data of Leq, Lx. The main results as followings: 1. Among all the percentages of the numbers of various vehicles, small-size cars (including sedans) were the highest (50.70%), and the percentages of motorcycles, mid-size cars and big-size cars were 45.03%, 7.02% and 2.95%, respectively. 2. The average values of the road traffic noise in L��, L�t, Leq were 78.87, 73.01 and 75.82 dBA, respectively. The average value for Leq in this study were higher than that requlated by the Environmental Noise Quality Standard Act proposed by Environmental Protection Administration Government of the R.O.C., especially for the maximum value of Leq. 3. Genrally, the width of road is positively related to the number of various vehicles, especially for the roads with six high-speed lines and two slow-speed lines. The number of motorcycles on the roads with six high-speed lines was the highest among all the roads, but the differences between the numbers of the motorcycles on other roads were small. 4. By using the Shih model (1980), the explaratory power to the variation of traffic noise value in this study was only 14% for L��, 16% for L�t, and 11% for Leq. The possible reason for this result was that the condition of this study was not completely suited to that of the Shih model. 5. By using the Delany model (1972) and Galloway model (1969), the explanatory powers of this study were very low, especially for Galloway model. 6. By using the sixth model proposed by Tzeng (1981), the explanatory powers were up to 17% for L��, 19% for L�t and 13% for Leq; however the powers were not satisfactory. The possible reason for the low explanatory powers was that the number of vehicles in Taipei city increased rapidly during last eight years. 7. By using the Shih model, Tzeng model, Delany model, Galloway model and experience law, three variables were selected to predict the values of L��, L、, Leq, and the explanatory powers were 18%, 20% and 15%, respectively. The predictive models established by the study were as follows: (1)□L��=55.46-0.001Q+1.892 log (Q□(1+P□))+5.965 log(Q□(1+P□)) R��=0.18 (2)□=50.12-0.001Q+1.614 log(Q□(1+P□))+5.899 log(Q□(1+P□)) R��=0.20 (3)□□=55.98-0.001Q+1.765 log(Q□(1+P□))+5.039 log(Q□(1+P□) R��=0.15 The study recommends that we should: (1) ban drivers shift lines arbitarily, (2) get rid of old vehicles to promote traffic flow. Other suggestions for future studies done for establishing the model include:(1) counting the numbers of inner and outer tracks of road separately (2) considering the distance between traffic lights, and (3) considering the situation of traffic flow.