| 摘要: | 營造業相較於其他行業，其職業災害更為嚴重。有鑑於營造業職業災害類型變動的不確定性，本研究以行政院勞工委員會1999年迄2009年營造業主要職業災害類型之數據，運用灰色理論GM(1,1)模型建構營造業主要職業災害類型之人數與百分比模型，並以2010年數據驗證預測模型的精準性。
研究結果顯示，各職業災害類型人數的平均絕對值百分比誤差(MAPE)為0.372%~3.269%，職業災害類型百分比的MAPE則為1.047%~4.276%，皆低於10%，以2010年之殘差值驗證，各職業災害類型人數及百分比也皆在10%合格範圍以內，證實運用GM(1,1)模型預測我國營造業主要職業災害類型均具有良好之預測精準度。以本研究建構之GM(1,1)模型預測未來趨勢顯示，被切、割、擦傷類型以及交通事故職災類型為所有類型上升幅度最大者；被撞與被夾、被捲呈現微幅上升；物體倒塌、崩塌有明顯下降之趨勢；墜落、滾落、跌倒、物體飛落則為小幅度下降。
關鍵字：營造業、職業災害類型、灰色理論、GM(1,1)模型
Abstract
The construction industry has serious occupation accidents among all industries in Taiwan. In view of the variation of occupational accidents types in construction industry is uncertain, this study applied GM(1,1) model of grey system theory to construct the number of people and the percentage forecasting model of the major occupational accidents types based on the occupational accidents data of construction industry from 1999 to 2009 from Council of Labor Affairs.
The results represent that the mean absolute percentage errors (MAPE) of the number of people for the various occupational accidents types are between 0.372% and 3.269%, and the MAPE of the percentage for the various occupational accidents types is 1.047-4.276%. Moreover, checking the validity from the residuals of 2010, the number of people and the percentage for the various occupational accidents types are all within 10% of eligible. Predicting the trend of the occupational accidents of construction industry from GM(1,1) model, cut, scrathed and traffic accidents are significant increase, hit, caught and sucked slightly increase. On the contrary, collapse shows a significant decline in the trend, and fall, tumble, and falling object decrease slightly.
Keyword: Construction Industry, Occupational Accident Types, Grey System Theory, GM(1,1) model |
