Iterative Learning Control For Internal Pressure Of Trains Under The Excitation Of The Morphologically Similar Tunnel Pressure Waves

When the high-speed train passes through the tunnel rapidly,a tunnel pressure wave will be generated by the tunnel-vehicle coupling.The tunnel pressure wave will induce into the carriage and cause the discomfort of the passengers.The numerous tunnels,high-density tunnels and long tunnels are becoming a new and important feature of the high-speed railway as the continuous growth of the construction of the high-speed railway in mountainous areas,for instance,the Sichuan-Tibet railway.However,the traditional passive control strategy by shutting down the air ducts for a fixed period may fail to handle the control of the riding comfort and the air quality inside carriage,as a result of ignorance of mismatching of the control strategy with the tunnel pressure waves.Focusing on the problem of mismatching with the tunnel pressure wave when traditional passive control is applied,researches about the strategy and the algorithm of iterative learning control(ILC)matching with the tunnel pressure wave is proposed.Firstly,the model of the internal pressure under multi-factors is established.Then,focusing on the varying-amplitude and varying-scale characteristics of the morphologically similar signal,ILC algorithms are formulated.Finally,the control strategy matching with the tunnel pressure waves are formulated and the ILC algorithm is applied into controlling of the internal pressure,then ILC algorithm matching with the tunnel pressure wave is proposed.In this dissertation,the main contributions concerning the internal pressure model,ILC strategy and algorithm and applicational algorithm for internal pressure control are as follows,(1)Combining the theory,simulation and experiments,the models of the internal pressure are established and analysis of the internal pressure under multiple factors is conducted.Firstly,by surveying the mechanism of the carbody deformation,the airtightness gaps and the ventilation ducts,theoretical models of the internal pressure caused by one of the factors are established.Then,based on the analysis of the coupling of carbody deformation,airtightness gaps and the ventilation fans,by applying the mass conservation and the ideal gas equation,a non-linear and a linear model of the internal pressure under multi-factors are established,under the consideration of the enhancing the precision of the simultation and reducing the complexity of algorithm design respectively.Finally,by applying the multi-factor model,the relationships between the deformation,the airtightness gaps or the air ducts and the opening degree of the pressure protection valve are analysed by simulation,which will guide the design of the control system.(2)The ideology of orthogonal projection,and the ideology of condition matching and performance iteration are applied to deal with the varying-amplitude and varying-scale characteristics of the morphologically similar signal respectively.On the basis of convensional ILC algorithm,the ILC algorithms for morphologically similar signals are formulated.Therotical surveys are conducted and then convergence analysis,simulations and experiments of the algorithm are performed.Results shows a better tracking performance for morphologically-similar signals when the formulated ILC algorithm is applied,comparing with the convensional ILC algorithm.(3)The repeativity of the train passes through the tunnel is applied and the ILC algorithm for internal pressure control under the morphologically similar tunnel pressure wave is surveyed.Firstly,an iterative learning control system for internal pressure control is established and the control strategy based on ILC is proposed.Then,taking the tunnel pressure wave as well as its amplitude and scale into consideration,on the basis of conventional ILC and the ILC algorithms for morphologically similar signal proposed in this dissertation,the ILC algorithm for internal pressure control under the excitation of the morphologically similar tunnel pressure wave is put forward.Finally,simulations for internal pressure control are conducted.Results show that: comparing with the traditional passive control algorithm,ILC algorithm will be able to match with the tunnel pressure wave and then alleviate the impact of the unproper opening time on the internal pressure.(4)Experiments and validations are conducterd on the test rig of the internal pressure control in Mechano-electronic Measurement and Control Laboraterial Center,Southwest Jiaotong University.Results furtherly validate that the ILC algorithm matching with the tunnel pressure wave will result in a better performance than traditional passive control algorithm.