Subjects: Computer Science >> Integration Theory of Computer Science submitted time 2018-05-20 Cooperative journals: 《计算机应用研究》
Abstract： According to the security and real-time performance of the train control system, the unified modeling language(UML) and the colored Petri net (CPN) model of C2 to C3 transition are established based on the level transition scenario of the CTCS-3 level control system requirement specification. This paper analyzes the influence factors of the safty of train operation and traffic efficiency, that is, the time spent in the transition and the success rate of the level transition, this paper verifies the validity of this modeling method. The results show that the combination of UML and CPN model is suitable for the specification of the requirements of the control system. The hierarchical conversion model can meet the real-time requirements of the system. In order to ensure the success rate of switching, the train running speed is inversely proportional to the switching time, the higher the speed, the shorter the switching time; the higher the train speed, the higher the real-time performance of the system.
Subjects: Computer Science >> Integration Theory of Computer Science submitted time 2018-05-18 Cooperative journals: 《计算机应用研究》
Abstract： Based on the powerful nonlinear mapping ability of kernel learning, proposing a class of kernel learning method for the short-term traffic flow forecasting. Kernel recursive least squares (KRLS) method using Approximate Linear Dependence (ALD) technique can reduce the computational complexity and storage capacity, the KRLS method is an online kernel learning method and is suitable for training on large-scale data sets. Kernel partial least square (KPLS) method utilizes the covariance between input and output variables to extract latent features. Kernel extreme learning machine(KELM) method uses the kernel function to substitute for the unknown nonlinear feature mapping of the hidden layer, in addition, the output weights of the networks can also be analytically determined by using regularization least square algorithm, hence KELM method provides better generalization performance at a much faster learning speed. In order to verify the validity of the proposed kernel learning method, the employed ALD-KRLS, KPLS and KELM methods were respectively applied to different traffic flow forecasting instances in different area, compared to the other methods under the same conditions. Experimental results show that the proposed kernel-based learning methods have higher forecasting accuracy and have improved training speed in the short-term traffic flow forecasting.
Hits
Hits
Downloads
Comment