【作者】 杨霄；

【导师】 雷波；

【作者基本信息】 西南交通大学 ， 供热、供燃气、通风及空调工程， 2004， 硕士

【摘要】 竖井送排式纵向通风作为一种经济节能的通风方式在长大公路隧道通风中被广泛使用。由于公路隧道通风系统具有很强的非线形特征，传统的线性控制理论难以获得精确的数学模型，因此模糊控制、神经网络控制及其两者相结合的神经模糊控制等现代控制方法成为公路隧道通风控制的趋势，但模糊控制及神经模糊控制在一定程度上存在不能及时根据外界条件的变化，有效地修改控制模型的问题。本文建立了公路隧道竖井送排式纵向通风神经网络在线控制系统的一般方法。 对于竖井送排纵向通风隧道，为了达到相同的通风效果，投入运行的通风设备可以有多种组合，但能耗却是各不相同的。本文采用非线性规划的方法来寻求不同交通条件和隧道污染状况下各控制时段投入运行的通风设备的最优组合，通过数值模拟计算，得出比较合理的结果，表明采用非线性规划方法求解竖井送排纵向通风的通风设备的组合是可行的。 探讨了神经网络在线控制器建立的一般流程，建立了隧道竖井送排式纵向通风的神经网络在线控制器；利用非线性规划方法获得的通风设备运行状况和通风过程的动态数值模拟取得样本数据对神经网络在线控制器初始化；然后建立神经网络在线控制器的隧道送排纵向通风的仿真程序，在MATLAB下对不同交通流下的隧道通风进行了模拟计算；最后，分析了神经网络在线控制器系统的适应性，表明采用神经网络在线控制方法对控制目标值、车辆基准排放量及交通流的变化均能实现有效控制。 通过模拟仿真各种交通状况下的污染物分布和通风设备运行情况，结果表明神经网络在线控制方法对长大隧道竖井送排纵向通风具有良好的控制效果，可以克服目前隧道通风智能控制方法不能及时适应外界变化问题。在长大公路隧道竖井送排式纵向通风控制中采用神经网络在线控制方法具有实际应用价值，本文的研究成果对公路隧道通风控制设计具有参考价值。更多还原

【Abstract】 Being energy saving, shaft blowing and exhausting longitudinal ventilation is widely used in long and large road tunnel ventilation. Because road tunnel ventilation system has strong non-linear characteristic and it is difficult to gain the precise mathematical model by using the traditional linear control theory, the modem control methods become the trend for the road tunnel ventilation control, such as the fuzzy control, neural networks control and FNNC. However, there are some difficulties with the establishment of the fuzzy membership functions and the rule base. This paper involves the research of the application of shaft blowing and exhausting longitudinal ventilation on-line control method.To get the same ventilation effect, various combinations of ventilation equipments are put into operation, but their energy consumption are different. This paper adopts the method of non-linear programming to seek the best combination of ventilation equipments under different traffic and tunnel contamination conditions and at different periods of time. Through the numerical simulation, the conclusion is drawn that it is feasible to solve the problem of combination of ventilation equipments with the method of non-linear programming.This paper also studies the general steps to establish neural networks on-line controller and establishes the neural networks on-line controller of shaft blowing and exhausting longitudinal ventilation. Through the combination of non-linear programming and numerical simulation, the sample data is obtained to initialize neural networks on-line controller. Then establish the computer simulation program of tunnel blowing and exhausting longitudinal ventilation and simulate tunnel ventilations under various traffic flows to evaluate the control effect with MATLAB. Finally, this paper discusses the adaptability of neural networks on-line control system.The simulation of pollutants distribution and the operations of ventilation equipments under various traffic conditions shows that long and large road tunnels can be effectively controlled with the method of shaft blowing and exhausting longitudinal ventilation. It can solve the problem of the inadaptability to the circumstances of the present tunnel ventilation intelligent control method. Neural network on-line control method is very practical in long and large road tunnel shaft blowing and exhausting longitudinal ventilation. This paper provides important reference for the road tunnel ventilation control and design.更多还原