塔吊施工重物坠落情景智能感知方法研究

目前施工现场的塔吊使用过程依赖于安全管理人员在现场进行安全监管，但仅依靠人进行施工全过程的安全管理难免会有疏漏。为进一步实现塔吊使用过程安全管理智能化，提出了一套基于计算机视觉和故障树分析的塔吊施工重物坠落危险情景智能感知方法。为验证该方法的有效性，考虑到上述情景在施工现场复现的危险性，在实验室环境下建立了塔吊使用过程中重物脱钩坠落伤人危险情景，利用提出的感知系统对情景进行了智能感知实验。结果表明，提出的智能感知方法能够对塔吊重物坠落情景进行准确的感知，自动感知频率为2次/秒，满足实时性要求。

Research on Intelligent Perception Method of Heavy Load Falling in Tower Crane Operation

Currently, the operation process of tower cranes at the construction site relies on safety management personnel for safety supervision. However, only relying on people for the safety management of the whole construction process is inevitably negligent. To further realize the intelligent safety management of tower crane operation，a set of intelligent perception methods of the heavy load falling hazard scenarios based on computer vision and fault tree analysis is proposed. To verify the effectiveness of the method，considering the danger of the above scenario reproduced at the construction site，a dangerous scenario of heavy objects falling off the hook and injuring people during the use of tower cranes is established in the laboratory environment. The proposed perception system is used to conduct intelligent perception experiments on the scenario. The results show that the proposed intelligent perception method can accurately perceive the tower crane heavy load falling scenario，and the automatic perception frequency is two times/second，which meets the real-time requirement.