Personnel dynamic risk assessment method and system based on neural network

Abstract

The invention provides a personnel dynamic risk assessment method and system based on a neural network. The method comprises the following steps: obtaining personnel static data and dynamic data and carrying out data coding; for the encoded data, obtaining an evaluation result by using a neural network; wherein the evaluation result is obtained by using the neural network, and the method comprises the steps of constructing a multi-task classification deep learning network, extracting deep features of static data and dynamic data by using a multi-layer full-connection network, extracting time sequence features among continuous multiple dynamic evaluation data by using a multi-layer gate control loop unit (GRU), and finally obtaining the evaluation result by using the full-connection network and softmax operation' and outputting risk levels of different extreme behaviors in a probabilistic manner. The risk levels of three types of extreme behaviors of the personnel can be timely and accurately grasped, key factors influencing the risk levels are mined, and data are expanded and perfected. The evaluation result of the invention is beneficial to the classification and level-to-level management of personnel, provides a targeted supervision and correction scheme, reduces the occurrence of extreme events, and guarantees the safety and stability.

Description

technical field [0001] The invention relates to the technical field of artificial intelligence, in particular to a neural network-based personnel dynamic risk assessment method and system. Background technique [0002] The statements in this section merely provide background information related to the present invention and do not necessarily constitute prior art. [0003] The deep learning network adopts the supervised training method to mine the deep space-time characteristics of the data, construct the functional relationship between the data and the label, and realize the classification of the unlabeled data. In a high-pressure airtight environment, personnel often hide their true psychological conditions. By collecting multi-dimensional data, including static data and dynamic data, during the process of admission and supervision and correction, the deep learning network is used to realize data fusion analysis and extract data in depth. The space-time characteristics can...

AI Technical Summary

Problems solved by technology

[0005] The existing guarding methods generally use observation combined with scales to regularly evaluate the extreme behavior levels of personnel, which has a large workload and low accuracy

In addition, due to the fact that a large amount of data has been collected, the utilization of the data is not sufficient

Risk assessment of personnel based on static factors can identify high-risk individuals, but cannot monitor dynamic changes in personnel's extreme behavior

Images