摘要
在港口及工业领域中,由于轮胎式起重机高强度的作业需求和复杂工况,对驾驶员疲劳状态的实时监控和安全管理提出了严格要求。为应对疲劳驾驶所带来的安全隐患,文中提出一种基于虚拟现实(VR)与数字孪生技术的疲劳检测系统。首先,通过3DS Max与Unity引擎构建高精度的起重机模型及动态交互场景;其次,利用ZeroMQ实现Unity与Python之间的跨平台通信,以支持高并发、低延迟的实时数据处理;最后,集成基于深度学习的多模态面部特征分析技术,高效识别驾驶员的疲劳状态并提供反馈。通过虚拟现实平台,疲劳监测过程得以可视化展现,增强了操作的沉浸感和交互性。实验结果表明:响应时间稳定在0.1~0.2 s,相较于Flask框架和TCP/UDP框架,ZeroMQ在高并发条件下表现出更低的延迟和更高的吞吐量,其非阻塞通信机制和动态负载均衡策略提升了系统的实时性和鲁棒性,为轮胎式起重机的智能化管理和操作安全性提升提供了技术参考。
Wheel cranes are widely used in ports and industrial fields.Given the high-intensity operational demands and complex working conditions,there is a pressing need for real-time monitoring and safety management to address driver fatigue.To mitigate the safety risks associated with fatigue driving,a fatigue detection system based on virtual reality(VR)and digital twin technology is proposed.Firstly,a high-precision crane model and dynamic interactive scene were constructed using 3DS Max and the Unity engine.Secondly,cross-platform communication between Unity and Python was enabled through ZeroMQ,supporting real-time data processing with high concurrency and low latency.Finally,multimodal facial feature analysis technology based on deep learning was integrated to effectively identify the driver’s fatigue state and provide immediate feedback.Through the virtual reality platform,the fatigue monitoring process was visually displayed,enhancing the immersion and interactivity of the operation.Experimental results show that the system’s response time remains stable between 0.1 and 0.2 seconds.Compared to the Flask framework and TCP/UDP frameworks,ZeroMQ demonstrates lower latency and higher throughput under high-concurrency conditions.Its non-blocking communication mechanism and dynamic load-balancing strategy significantly improve the system’s real-time performance and robustness.This research serves as a valuable reference for the intelligent management and operational safety enhancement of wheel cranes.
作者
李果朋
马思群
Li Guopeng;Ma Siqun
出处
《起重运输机械》
2025年第14期54-63,共10页
Hoisting and Conveying Machinery
基金
大连市科技创新基金项目(2020JJ27FZ126)
辽宁省自然科学基金资助项目(20180550474)。
