The Research Of Tunnel Ventilation Control System Based On Frequency Conversion Speed Regulation

With China’s highway construction continues to expand, rapid increase of highwaytunnel mileage directly result in an increase of operating costs of tunnel ventilation system,energy conservation has become an important means of reducing energy waste of tunnelventilation. Since the tunnel longitudinal ventilation system has such characteristics asnon-linear, time-delay and time-varying, it is difficult to establish a precise mathematicalmodel. In this paper, frequency control technology is applied to control tunnel ventilation andenergy conservation system, reduce excessive electricity loss and equipment loss, therebyreduce operating costs.In this paper, traditional tunnel ventilation control technology and frequency controltechnology are compared to illustrate the feasibility of frequency control technology in tunnelventilation control system. Based on related research, highway tunnel ventilation frequencycontrol system is designed and mathematical modeled. According to the relevant parametersand technical indicators of Baiyun tunnel, algorithms and simulation of tunnel ventilationfrequency control system is studied. Then, by detecting the traffic volume and vehicle speedof Baiyun tunnel, SUMO software is used to conduct simulation. Based on the simulationresults, required air volume of CO, VI and peculiar smell is respectively calculated andcompared to get the ultimate required air volume of tunnel. Output power and rotational speedof Jet Fans are regulated by frequency control technology. The energy consumption oftraditional jet fan multistage control and frequency control is compared, superiority of thetunnel ventilation frequency control system is verified.At last, BP neural network control algorithm is employed to control tunnel ventilationfrequency control system, the real-time value of CO, VI and pollutant concentration are usedas input samples, the output power of jet fans is taken as the target output samples. UseMATLAB Neural Network Toolbox for training and simulation, experimental results showthat the average error between the actual output power of jet jans and frequency of neuralnetwork is less than0.2, the established neural network model can fit the frequency ventilation control model well, and have relatively strong generalization ability.