The embodiment of the invention discloses a training method and a detection method of a model for elevator guide rail state detection. At least two dimensions of elevator data at multiple moments in guide rail vibration information, elevator running speed and elevator position information are acquired, and a preset neural network is trained at least based on the elevator data at the multiple moments. Due to the fact that guide rails play a role in guiding when an elevator runs and bear impact generated when an elevator car and the elevator are braked as well as impact force generated when a pair of safety tongs is braked emergently, the running state of the guide rails can indirectly reflect the state of the whole elevator system, by detecting the guide rail state of the elevator through the embodiment of the invention, safety vulnerabilities existing in an existing elevator safety monitoring method are solved, and therefore the safety of the elevator is guaranteed.
本发明实施例公开了一种用于电梯导轨状态检测的模型的训练方法、检测方法。通过获取包括导轨振动信息、电梯运行速度和电梯位置信息中至少两个维度的多个时刻的电梯数据,并至少基于多个时刻的电梯数据,对预设的神经网络进行训练。因为电梯运行时导轨起导向作用,同时承受轿厢,电梯制动时的冲击,安全钳紧急制动时的冲击力,因此其运行状态能间接反应整个电梯系统的状态,所以本发明实施例通过对电梯的导轨状态进行检测,解决了目前电梯的安全监测方法中存在的安全漏洞,进而使电梯的安全得到保障。
Training method and detection method of model for elevator guide rail state detection
用于电梯导轨状态检测的模型的训练方法、检测方法
2021-04-06
Patent
Electronic Resource
Chinese
IPC: | B66B Aufzüge , ELEVATORS |