Safety Risk Prediction Model of High-Rise Building Construction Based on Key Physiological Index

The tasks conducted on a high-rise building are complex and dangerous, and the construction safety of the construction personnel needs to have a higher guarantee. In this study, the key physiological indicators of high-rise construction workers were monitored and collected in real time by selecting a smart wearable device integrated with multiple sensors. On this basis, the key physiological index parameters are analyzed and screened, which are taken as input parameters, and the construction risk prediction results are taken as output. The BP neural network model and support vector machine (SVM) are, respectively, used to establish the safety risk prediction model of high-rise construction workers based on key indicators, to quantitatively assess the construction risk of the construction workers in the process of high-rise construction. The results showed that heart rate and blood pressure had the greatest impact on the construction safety of the construction worker, followed by the duration of work, age, working period, and gender. Compared with the BP neural network, the risk prediction model established by SVM can obtain more accurate prediction results under the condition of a smaller training data set. The presented research can not only effectively reduce the health threats caused by the physical and psychological effects faced by construction personnel when working at altitude and ensure construction safety, but also further enrich the application scenarios of multi-sensor data-driven equipment and expand its application in the construction field.