基于T-S网络的建筑施工安全预警体系研究

发布时间：2018-05-05 02:05

本文选题：建筑施工 + 安全预警　；参考：《沈阳建筑大学》2015年硕士论文

【摘要】：我国将新型城镇化发展纳入十二五规划中,为建筑业进一步发展提供新的契机。建筑业为我国经济做出巨大贡献的同时,也面临着严峻的挑战。为了预防建筑安全事故的发生,必须做到对建筑施工安全状态的提前预知。与国外很多发达国家相比,我国在建筑施工安全预警水平上还很落后。因此,采取科学、有效的预警方法对建筑施工安全状态进行预警,进而制定相应的预控对策就显得尤为重要从系统工程学角度出发,将“人-机材-环境-管理”作为一个系统,采用事故致因理论对建筑施工安全影响因素进行分析。按照科学性、实践性等原则选取主要影响因素,建立了建筑施工安全预警指标体系。确定建筑施工指标时,尽量对指标能够定量化评价,减少预警指标打分时人为主观因素对预警结果准确性的影响。通过对模糊神经网络的研究,发现模糊神经网络汲取了模糊系统和神经网络的优点,适合解决数学分析、非线性系统建模问题。借鉴在其他学科领域的应用,创新性的构建了基于T-S模糊神经网络的建筑施工安全预警模型,对预警模型的网络结构和算法进行了设计。经过大量样本的训练和检测后,结果表明,所构建的预警模型具有较好的准确性和泛化性。通过信号灯形式表示预警结果,不仅直观地反映出某一周期的建筑施工安全状态,还能对一定周期内建筑施工安全总体态势进行宏观的把控。根据建筑施工安全预警处理结果,从技术、培训、管理等方面提出了宏观预控对策。选取“郑港新城”项目作为实例,运用所构建的建筑施工安全预警模型,得出其施工安全状态为“一般安全”。针对该项目预警结果,结合项目实际施工安全特点,从人员、机材、环境、管理四个方面提出了具体的预控对策和建议。证明构建的基于T-S模糊神经网络的建筑施工安全预警模型具有很好的适应性。
[Abstract]:Our country brings the new urbanization development into the 12th five-year plan, which provides a new opportunity for the further development of the construction industry. Construction industry makes great contribution to our economy, but also faces severe challenges. In order to prevent the occurrence of building safety accidents, it is necessary to predict the construction safety state in advance. Compared with many developed countries, our country still lags behind in construction safety warning level. Therefore, it is very important to take the scientific and effective early warning method to warn the construction safety state, and to formulate the corresponding pre-control countermeasures from the angle of system engineering. In this paper, the theory of accident cause is used to analyze the influence factors of construction safety in the system of "man-machine, environment-management". According to the principles of science and practicality, the main influencing factors are selected, and the early warning index system of construction safety is established. When determining the construction index, the index can be quantitatively evaluated as far as possible, and the influence of subjective factors on the accuracy of early warning result can be reduced when the early warning index is graded. Through the research of fuzzy neural network, it is found that fuzzy neural network has the advantages of fuzzy system and neural network, which is suitable for mathematical analysis and nonlinear system modeling. Based on the application of other disciplines, the paper creatively constructs the building construction safety early-warning model based on T-S fuzzy neural network, and designs the network structure and algorithm of the early-warning model. After a large number of samples are trained and tested, the results show that the proposed early warning model has good accuracy and generalization. The early warning result expressed by signal lamp can not only directly reflect the construction safety state of a certain period, but also control the overall situation of construction safety in a certain period. According to the result of early warning and treatment of construction safety, this paper puts forward macro-control countermeasures from the aspects of technology, training, management and so on. Taking the project of "Zhenggang New Town" as an example and using the building construction safety warning model, it is concluded that the construction safety state is "general safety". In view of the early warning result of the project, combined with the actual construction safety characteristics of the project, this paper puts forward some concrete pre-control countermeasures and suggestions from four aspects: personnel, mechanical material, environment and management. It is proved that the building safety warning model based on T-S fuzzy neural network has good adaptability.
【学位授予单位】：沈阳建筑大学
【学位级别】：硕士
【学位授予年份】：2015
【分类号】：TU714

【相似文献】