建筑施工安全事故预警研究

发布时间：2018-03-19 08:54

本文选题：建筑施工　切入点：安全预警　出处：《哈尔滨工业大学》2017年硕士论文　论文类型：学位论文

【摘要】：建筑业的快速发展在促进就业岗位增加的同时,也带动了许多其他行业的进步,在国民经济的发展历程中占据着重要地位。但是,建筑工程的建设周期长、施工过程复杂,影响因素较多,这使得建筑施工安全事故的发生频率也较高,由此造成的人员伤亡与财产损失数量惊人。虽然国家对此十分重视,颁布了许多与建筑安全管理相关的法律法规,并严格要求施工企业做好安全管理工作,但是仍然不能将建筑安全从被动管理的状态中解脱出来。因此,实现建筑业由事后应急处理向事前预警防范的转变至关重要。目前,国内外学者对于预警防范方面的研究较多,并取得了一定成果。但是对于建筑施工安全预警,现阶段的研究缺乏精确度高、适用性强的预警模型,因此,建立一种精确的高效的建筑施工安全预警模型具有重要意义。本文通过对建筑施工安全管理的现状进行分析,指出较大事故及较大以上事故的死亡率较高,阐明进行建筑施工安全事故预警研究的必要性。通过对我国2011年以来发生的建筑施工安全事故的类型及规律分析,综合运用事故致因理论,找出影响建筑施工安全事故的主要因素,并借助SPSS软件进行影响因素的筛选与综合,最终建立起具有五个一级指标、十九个二级指标的建筑安全预警指标体系。通过分析与对比各种预警方法的特点及适用范围,指出BP神经网络用于建筑施工安全事故预警的优越性,并针对基于BP神经网络的建筑施工安全事故预警传统模型的局限性,在收敛速度与预测精度方面进行了改进;最后,结合项目实际情况借助Matlab软件对改进模型进行实证研究,为提高建筑安全管理的效率提供理论依据。
[Abstract]:The rapid development of the construction industry in promoting employment increases at the same time, also led to many other industries progress, plays an important role in the development of the national economy. However, construction of the long construction period, complex construction process, the influence of many factors, the frequency of accidents of construction personnel is also higher. Casualties and property losses caused by a staggering amount. Although the country attaches great importance to the promulgation of a number of related laws and regulations of construction safety management, and strict requirements for construction enterprises to do safety management work, but still can not be free from the passive management of construction safety state. Therefore, the construction industry to achieve by the emergency treatment is crucial to prevent to advance warning changes. At present, domestic and foreign scholars for early warning researches, and has made some achievements. But for the construction Construction safety warning, the present study is lack of high accuracy, strong applicability and early warning model, therefore, it is important to establish an accurate and efficient construction safety early warning model. By analyzing the current situation of construction safety management, points out that more accidents and accidents than larger high mortality, it is necessary for construction safety accident early warning research. Through the analysis of the types and laws of building construction safety accidents happened in China since 2011, the comprehensive use of accident causation theory, find out the main factors influencing the construction safety accidents, and the factors influencing the selection and integrated with SPSS software, set up with five an index, construction safety early warning index system of nineteen level two indexes. The characteristics and applicable scope of analysis and comparison of various methods of early-warning, pointed out that the BP of God The superiority of early warning network for construction safety accidents, and aiming at the limitations of BP neural network construction safety accident early warning based on the traditional model, the convergence speed and prediction accuracy is improved; finally, combined with the actual situation of the project with the help of Matlab software, the empirical research on the improved model, in order to improve the efficiency of construction safety management to provide a theoretical basis.

【学位授予单位】：哈尔滨工业大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TU714

【参考文献】