Entity

Time filter

Source Type


Yongzhi J.,Key Laboratory of Magnetic Suspension Technology and Maglev Vehicle | Jian X.,Southwest Jiaotong University
2011 IEEE International Conference on Cyber Technology in Automation, Control, and Intelligent Systems, CYBER 2011 | Year: 2011

The gap sensor plays an important role for electromagnetic levitation system which is a critical component of highspeed maglev train. Artificial neural network is a promising area in the development of intelligent sensors. In this paper, we present an model of gap sensor based on radial basis function (RBF) neural network. The proposed model based RBF scheme incorporates intelligence into the sensor. It is revealed from the simulation studies that this gap sensor model can provide correct gap within 0.3mm error over a range of temperature variations from 20 C to 80 C. The experimental results show that the compensated gap signal meets the requirement of levitation control system. © 2011 IEEE. Source


Liu D.,Southwest Jiaotong University | Feng Q.,Southwest Jiaotong University | Jiang Q.,Southwest Jiaotong University | Jiang Q.,Key Laboratory of Magnetic Suspension Technology and Maglev Vehicle
Xinan Jiaotong Daxue Xuebao/Journal of Southwest Jiaotong University | Year: 2010

To reduce the effect of nonlinearization on maglev gap control, the PSO (particle swarm optimization) algorithm was used to optimize the parameters of a maglev controller, and an improved algorithm was proposed based on the linear decreasing weight particle swarm optimization (LDW-PSO). In order to improve the optimization speed and convergence performance, neighborhood topologies, stagnation detection and global best perturbation were adopted to build the improved algorithm. The simulation and experiment results show that the output overshoot of an optimized PID (proportional-integral-derivative) controller based on the improved algorithm is 45% smaller than that of a traditional PID maglev controller. Source


Jing Y.,Key Laboratory of Magnetic Suspension Technology and Maglev Vehicle | Zhang K.,Key Laboratory of Magnetic Suspension Technology and Maglev Vehicle | Xiao J.,Southwest Jiaotong University
Proceedings - 2011 8th International Conference on Fuzzy Systems and Knowledge Discovery, FSKD 2011 | Year: 2011

In this brief, we propose a fuzzy neural network (FNN) modeling approach which is applied for the modeling of gap sensor in the high-speed maglev train. The gap sensor plays an important role for electro-magnetic levitation system which is a critical component of high-speed maglev train. Artificial neural network is a promising area in the development of intelligent sensors. In this paper, we present a model of gap sensor based on fuzzy neural network. The proposed model based fuzzy network scheme incorporates intelligence into the sensor. The fuzzy neural network, as an inverse model compensator if connected in series to the output terminal of the gap sensor, would estimate the correct true gap in a range of temperature after proper training. We trained the network by gradient descent learning algorithm with momentum. It is revealed from the simulation studies that this gap sensor model can provide correct gap within the error less than 0.4mm over a range of temperature variations from 20 C to 80C and within 0.2mm only considering the work gap 8mm to 12mm. The experimental results show that the compensated gap signal meets the requirement of levitation control system. © 2011 IEEE. Source


Lu L.,Key Laboratory of Magnetic Suspension Technology and Maglev Vehicle | Lu L.,Southwest Jiaotong University | Zhao H.,Key Laboratory of Magnetic Suspension Technology and Maglev Vehicle | Zhao H.,Southwest Jiaotong University
Journal of Sound and Vibration | Year: 2016

The filtered-x least mean lp-norm (FxLMP) algorithm is proven to be useful for nonlinear active noise control (NANC) systems. However, its performance deteriorates when the impulsive noises are presented in NANC systems. To surmount this shortcoming, a new nonlinear adaptive algorithm based on Volterra expansion model (VFxlogLMP) is developed in this paper, which is derived by minimizing the lp-norm of logarithmic cost. It is found that the FxLMP and VFxlogLMP require to select an appropriate value of p according to the prior information on noise characteristics, which prohibit their practical applications. Based on VFxlogLMP algorithm, we proposed a continuous lp-norm algorithm with logarithmic cost (VFxlogCLMP), which does not need the parameter selection and thresholds estimation. Benefiting from the various error norms for 1≤p≤2, it remains the robustness of VFxlogLMP. Moreover, the convergence behavior of VFxlogCLMP for moving average secondary paths and stochastic input signals is performed. Compared to the existing algorithms, two versions of the proposed algorithms have much better convergence and stability in impulsive noise environments. © 2015 Elsevier Ltd. Source


Wu R.,Southwest Jiaotong University | Wang Y.,Key Laboratory of Magnetic Suspension Technology and Maglev Vehicle | Yan Z.,Key Laboratory of Magnetic Suspension Technology and Maglev Vehicle | He Z.,Southwest Jiaotong University | Wang L.,Southwest Jiaotong University
IEEE Transactions on Plasma Science | Year: 2013

Higher energy density makes inductive energy storage more promising than capacitive storage for pulsed power supplies in industrial and military fields. To realize high amplitude of pulsed current and relieve stress of opening switch, this paper proposes a novel inductive pulsed power supply consisting of high-temperature superconducting pulse power transformer and ZnO-based nonlinear resistor. First, working processes and laboratory setup are described in detail. Then, simulation using the software SIMPLORER is built to show major pulse characteristics and comparisons of two different nonlinear resistors. For verifying the feasibility of this mode, high-current testing is carried out and the results show that large amplitude of pulsed current 3 kA with energy transfer efficiency 60% is achieved, and the ZnO-based nonlinear resistor can help to limit the voltage of the opening switch to a small constant below its clamping value as current is interrupted. © 1973-2012 IEEE. Source

Discover hidden collaborations