CAJ | 학술논문

建筑施工安全问题日益凸显，迫切需要进行建筑施工安全预警模型方面的研究。本文结合以往的建筑施工安全事故案例，从人、物、环境、管理四个方面，对建筑施工安全影响因素进行分析，构建建筑施工安全预警指标体系。本文将建筑施工安全预警状态分为5个等级，并采用专家打分法对各预警区间定量划分，并借鉴模糊神经网络预警模型在其他学科领域的应用，建立了基于T-S网络的建筑施工安全预警模型。然后经过大量样本的训练和检测，使网络具有良好的泛化性。本文选取沈阳市某3个在建项目进行预警，发现与传统预警模型得出的结果一致，进一步验证了T-S网络预警模型的准确性和可行性。最终，得出T-S网络预警模型在建筑施工安全预警中具有适用性的结论。
건축시공안전문제일익철현，박절수요진행건축시공안전예경모형방면적연구。본문결합이왕적건축시공안전사고안례，종인、물、배경、관리사개방면，대건축시공안전영향인소진행분석，구건건축시공안전예경지표체계。본문장건축시공안전예경상태분위5개등급，병채용전가타분법대각예경구간정량화분，병차감모호신경망락예경모형재기타학과영역적응용，건립료기우T-S망락적건축시공안전예경모형。연후경과대량양본적훈련화검측，사망락구유량호적범화성。본문선취침양시모3개재건항목진행예경，발현여전통예경모형득출적결과일치，진일보험증료T-S망락예경모형적준학성화가행성。최종，득출T-S망락예경모형재건축시공안전예경중구유괄용성적결론。
Due to the construction safety issues become increasingly prominent, it is very necessary to study about construction safety warning model. The Paper analyzes the factors which affect the construction safety from the aspects of people, objects, environment and management combined with the previous construction accident cases, and builds security early-warning indicator system construction. It divides the construction safety warning state into five levels, and applies expert scoring method to quantify division of each warning interval. Learning the application of fuzzy neural network in other disciplines and fields, the article builds the construction safety warning model based on T-S network . After training and testing large number of samples, the network has a good generalization. Then it selects three construction projects in Shenyang City for early-warning, it finds that the results are consistent with traditional early-warning model, further validate the accuracy and feasibility of T-S network early-warning model. Finally, it gets the conclusion that T-S network has the applicability in construction safety warning.