为满足掘进工作面通风系统智能化的发展需求，针对传统局部通风系统无法实时监测及智能调控风筒出风口风流状态而导致风速场分配不合理，死角区瓦斯积聚严重和粉尘污染等安全隐患问题，提出一种基于数字孪生技术的掘进工作面出风口风流智能调控系统，优化风流场分布。分析建立了系统实现的整体框架、运行流程及关键技术，利用Zigbee自组网功能进行巷道风速、瓦斯体积分数及粉尘质量浓度等实时数据的监测采集，通过ARIMA时间序列预测模型对下一时刻瓦斯体积分数及粉尘质量浓度进行智能预测分析，并引入小生境四段式编码遗传算法提取相应出风口风流智能调控规则，结合GPRS无线传输技术实现出风口风流状态的智能调控，在此基础上，运用Unity3D构建系统虚拟模型并实现了物理实体与虚拟孪生体的映射交互。通过设计搭建的数字孪生系统实验测试平台，验证了系统实时监测、决策评价、智能调控、虚实融合等关键功能技术可行性，并结合具体案例对系统智能调控效果进行了对比分析，结果表明:调控后司机位置处风速明显上升，上隅角瓦斯体积分数由0.623%降低为0.306%，降低率达50.8%;司机位置粉尘质量浓度值由1 180 mg/m3降低到695 mg/m3，单点降尘率达41.1%，回风侧行人高度沿程粉尘质量浓度由430 mg/m3降低为150 mg/m3，降尘率达65%，进一步优化了掘进工作面巷道风速、瓦斯及粉尘场的运移分布。

In order to meet the intelligent development needs of the ventilation system in the heading face,aiming at overcoming the problems that the traditional local ventilation system cannot monitor and intelli-gently control the airflow state of the air outlet in real time,which leads to the unreasonable distribution of the wind speed field,the serious gas accumulation in the dead corner area and the safety hazards such as dust pollution,an airflow intelligent control system based on the digital twin technology at the air outlet of the heading face is proposed to optimize the airflow field distribution.The overall framework,operation process and key technologies of the system implementation is established,using the Zigbee self-organizing network function to monitor and collect real-time data such as wind speed,gas and dust concentration in the roadway.The ARIMA time series prediction model is used to intelligently predict and analyze the gas and dust concentration at the next moment,moreover the niche four-segment coding genetic algorithm is introduced to extract the corresponding intelligent control rules of airflow,and combined with GPRS wireless transmission technology to realize the intelligent control of airflow status of air outlet.On this basis,Unity3D is used to construct the virtual model of the system and realize the mapping interaction between the physical entity and the virtual twin.Through the design and construction of the digital twin system experimental test platform,the technical feasibility of system’s key function such as real-time monitoring,decision-making evaluation,intelligent control,and virtual-real integration are verified,besides combined with specific cases to compare and analyze the intelligent control effect of system.The results shows that after adjustment,the wind speed at the driver’s position increases significantly.The gas concentration at the top corner is reduced from 0.623% to 0.306%,and the reduction rate reaches 50.8%.The dust concentration at the driver’s position is reduced from 1 180 mg/m3 to 695 mg/m3,and the single-point dust reduction rate reached 41.1%,the dust concentration along the pedestrian height at the return air side is reduced from 430 mg/m3 to 150 mg/m3,and the dust re-duction rate reaches 65%,which further optimize the distribution of wind speed,gas and dust field in the heading face.

1 系统实现的整体框架及运行流程

  1.1 整体框架模型及组成

  1.2 系统运行流程

2 系统实现的关键技术

  2.1 基于ZigBee自组网功能的实时监测技术

  2.2 基于ARIMA模型的瓦斯体积分数及粉尘质量浓度智能预测技术

  2.3 基于小生境四段式编码遗传算法的风流调控规则获取技术

  2.4 基于GPRS的系统无线传输调控技术

  2.5 基于Unity3D的数字化建模技术

3 系统实现及验证

  3.1 系统实现功能验证

  3.2 智能调控效果验证

4 结论