Qiao Jun

Bio: Qiao Jun is an academic researcher. The author has contributed to research in topics: Artificial neural network & Radial basis function. The author has an hindex of 1, co-authored 1 publications receiving 8 citations.

Position calculation models by neural computing and online learning methods for high-speed train

Abstract: For high-speed trains, high precision of train positioning is important to guarantee train safety and operational efficiency. By analyzing the operational data of Beijing---Shanghai high-speed railway, we find that the currently used average speed model (ASM) is not good enough as the relative error is about 2.5 %. To reduce the positioning error, we respectively establish three models for calculating train positions by advanced neural computing methods, including back-propagation (BP), radial basis function (RBF) and adaptive network-based fuzzy inference system (ANFIS). Furthermore, six indices are defined to evaluate the performance of the three established models. Compared with ASM, the positioning error can be reduced by about 50 % by neural computing models. Then, to increase the robustness of neural computing models and real-time response, online learning methods are developed to update the parameters in the last layer of neural computing models by the gradient descent method. With the online learning methods, the positioning error of neural computing models can be further reduced by about 10 %. Among the three models, the ANFIS model is the best in both training and testing. The BP model is better than the RBF model in training, but worse in testing. In a word, the three models can reduce the half number of transponders to save the cost under the same positioning error or reduce the positioning error about 50 % in the case of the same number of transponders.

Complex-Valued Ordinary Differential Equation Modeling for Time Series Identification

Abstract: Time series identification is one of the key approaches to dealing with time series data and discovering the change rules. Therefore, time series forecasting can be treated as one of the most challenging issues in this field. In order to improve the forecasting performance, we propose a novel time series prediction model based on a complex-valued ordinary differential equation (CVODE) to predict time series. A multi expression programming (MEP) algorithm is utilized to optimize the structure of the CVODE model. So as to achieve the optimal complex-valued coefficients, a novel optimization algorithm based on a complex-valued crow search algorithm (CVCSA) is proposed. The chaotic Mackey-Glass time series, small-time scale traffic measurements, Nasdaq-100 index, and Shanghai stock exchange composite index are utilized to evaluate the performance of our method. The results prove that our proposed method could predict more accurately than state-of-the-art real-valued neural networks and an ordinary differential equation (ODE). The CVCSA has faster convergence speed and stronger optimization ability than the crow search algorithm (CSA) and particle swarm optimization (PSO).

Research of small current grounding fault location algorithm in distribution grid based on RBF

Abstract: Single phase grounded fault of small current often occurrs in distribution network. In order to assure consumer an uninterruptible power supply. These are necessary: increase the diagnosis precision of single phase grounded fault in distribution network, locate the fault point of small current grounded and cut off the fault. This paper proposes an fault location algorithm based on RBFneural network. Some datas are analysed which are collected by the feedback terminal device in distribution network. The analysis results show that the change of zero-sequence current is most evident. Therefore the zero-sequence current's value is consider as the input value of RBF neural network, the fault location of small current grounded is analyzed that based on the sample's trainning of existing zero-sequence current parameter. In the same time the ground fault of small current realize the real-time self-adapting location. In this paper, the MATLAB is used to do the simulation, the simulation results is close to the expect result. It shows the network can real-time accurately proceed test for the small current grounded fault of distribution network. In addition, field test demonstrates that the fault location of online state is feasible.

Application of Adaptive RBF-SMC for Electro-Hydraulic Position Servo System

Abstract: Electro-hydraulic servo system was hard to control with traditional control strategy and RBF-SMC (Radial Basis Function neural networks-Sliding Mode Control) controller was designed for this system. The mathematical model of the electro-hydraulic servo system was analyzed and the neural sliding mode controller was designed, the control law of sliding mode control was based on linearization feedback techniques and estimate parameters with RBF neural network. The simulation shows RBF neural networks can learning the uncertainties and disturbance, RBF-SMC has good control performance of reduces chattering and parameters estimation.

Active Disturbance Rejection Course Control for USV Based on RBF Neural Network

Abstract: With the development and utilization of marine resources and the development of the marine environment, Unmanned Surface Vessel (USV) are playing an increasingly important role. This paper takes the Dalian Maritime University "LanXin" as the research object. Based on the traditional Norrbin model, uses the ADRC algorithm and the ADRC algorithm based on the RBF neural network to design the corresponding USV course controllers. Then the corresponding simulation experiments are performed on the above controllers and the effects are compared. Finally, in this paper, the "LanXin" was used as the object, and the corresponding ship experiments were carried out in the sea area of Dalian Cross-sea Bridge. Through the analysis of the experimental data, it can be seen that the ADRC controller based on RBF neural network has better control effect.