利索-全球专利检索

  1. 登录
  2. 注册
发明公开
CN110599735A 一种基于变电站作业违章行为智能识别的告警方法 无效 - 驳回

一种基于变电站作业违章行为智能识别的告警方法
摘要：
本发明涉及电力领域，尤其涉及一种基于变电站作业违章行为智能识别的告警方法，包括：采集变电站作业现场的图像；通过图像识别算法判断图像中是否有作业人员；对于有作业人员的图像，再次通过图像识别算法判断图像中的作业人员是否存在作业违章行为；若图像中的作业人员存在作业违章行为，则向作业现场发出报警信息。本发明通过将两种定位方式结合，能够实现管线各个目标点的定位信息的采集，最终得到的管线图更加符合实际。通过图像识别算法对变电站作业现场的作业人员进行违章行为的判断，在判断出违章行为后通过向作业现场发出报警信息来提醒作业人员，实现对作业人员在作业现场的监控，在一定程度上降低了事故发生的概率。
摘要（英）：

The present invention relates to the power field, and in particular to an alarming method based on intelligent recognition of a transformer substation operation regulation-breaking behavior. The method comprises: collecting an image of a transformer substation working site; determining whether there are operating personnel in the image by using an image recognition algorithm; for an image with operating personnel, determining, by using the image recognition algorithm again, whether the operating personnel in the image is performing an operation regulation-breaking behavior; and sending alarm information to the working site if the operating personnel in the image is performing an operation regulation-breaking behavior. According to the alarming method, two positioning manners are combined,so that positioning information of each target point in a pipeline can be collected, and a finally obtained pipeline drawing better corresponds to reality. The image recognition algorithm is used to determine the regulation-breaking behavior of the regulation-breaking behavior on the transformer substation working site, and after the regulation-breaking behavior is determined, the alarm information is sent to the working site to warn the operating personnel, so that the operating personnel on the working site is monitored, thereby reducing an accident probability to a certain extent.

0/0