Real-time hazard detection and mitigation on construction sites based on artificial neural networks

Abstract

As construction site safety problems appear more and more often, China, as a global infrastructure power, gives high importance and attention to them, so the research of construction site safety early warning technology is very necessary. In this paper, we introduce the technology related to the topic and analyze the early warning process. The anchor nodes are classified by K-nearest neighbor algorithm, and the optimal number of anchor nodes are selected to participate in the positioning. By optimizing the weights of the weighted prime positioning algorithm to make it more rationalized, it is experimentally proven to have better accuracy and higher efficiency than the existing personnel positioning algorithm; it is expected to be able to provide smooth and stable, timely and accurate early warning, predict ten common construction risks, and effectively avoid the construction risks.