Lan Q.,Anhui Communications Vocational and Technical College
Revista Tecnica de la Facultad de Ingenieria Universidad del Zulia | Year: 2016
Character recognition is a topic with great research value and it includes all typical issues in the field of pattern recognition: data acquisition, processing and selection, selection of input sample expression as well as the guided training on the recognizer with sample set. Artificial neural network (ANN) is a mathematical model which simulates the behavior characteristics of neural network and performs distributed and parallel information processing. Along with the progress in neural network and logic techniques, people have used many new methods and means on the study of character recognition. This paper proposes a character recognition method based on improved BP neural network. This method extracts features of training samples according to certain methods, builds the classification analysis system, divides each character to be recognized into 8∗8 blocks for digitalization and represents each block with a vector. The input vector is defined as an input vector matrix and the vector represents a certain character or number. Then, it conducts feature extraction on certain unknown sample to be detected with the same method and the system will judge its classification according to the sample feature extracted. The training of neural network shall train the arrays which represent characters or numbers respectively in the input end under supervision and correspond to the specific location in the output end. The test result on this character recognition method shows that this algorithm has relatively high recognition accuracy, a faster processing speed and strong robustness in the applications.
Zhong L.-S.,East China Jiaotong University |
Yan Z.,Anhui Communications Vocational and Technical College |
Yang H.,East China Jiaotong University |
Qi Y.-P.,East China Jiaotong University |
And 2 more authors.
Tiedao Xuebao/Journal of the China Railway Society | Year: 2013
Along with raising of train speeds, dynamic interaction between train and catenary, track and air intensifies, which sets high requirements for modeling and control of high-speed trains. In this paper, the predictive controller, based on data driven subspace approach, was designed for high-speed trains. The high-speed train multivariate dynamic system was established on the basis of the state space model. The high-speed train predictive subspace model was designed from input and output data. The method of designing the predictive controller was analyzed in detail and the predictive control algorithm was given. Through numerical simulation of high-speed trains, the effectiveness of the proposed control method is proved.
Wu F.,Lanzhou Jiaotong University |
Ma C.-X.,Lanzhou Jiaotong University |
Li Z.-C.,Anhui Communications Vocational and Technical College
Jiaotong Yunshu Xitong Gongcheng Yu Xinxi/Journal of Transportation Systems Engineering and Information Technology | Year: 2011
Because of the high cost of building urban compressed natural gas (CNG) station, reasonable arrangement of CNG filling stations plays significant role. Considering the arrangement problem usually involves several complex factors, the paper develops an assignment model based on minimum travel cost and maximum enterprise economic benefit in view of the investor, user and urban planning. The improved hybrid particle swarm algorithm is also designed for solution. Simulation experiment results show the effectiveness of the algorithm in solving this kind of location arrangement problems. The quantity, type and location of the urban CNG station are determined. It thus enables the resource is reasonably and fully utilized, which may provide reference for planning the urban CNG station.
Wang H.,Hefei University of Technology |
Yang L.,Anhui Communications Vocational and Technical college |
Hu Y.,Hefei University of Technology
INTERNOISE 2014 - 43rd International Congress on Noise Control Engineering: Improving the World Through Noise Control | Year: 2014
The coupled dynamics between the vehicle chassis suspension system and electrical power steering system(EPS) is analyzed, to establish the full-vehicle nonlinear model, EPS model, tire model and road input model. The tire's complex nonlinear model is approximated by utilizing the least square method, so as to obtain the integrated system model with the 22th-order. To simplify the nonlinear controller design, the nonlinear dynamics model is separated into two parts of nonlinear part and linear part. The state feedback optimal controller is designed for the linear part model, and the linear compensator based on the deviation separation is for the nonlinear part model, which can ensure the closed-loop control system is exponential asymptotic stability at the equilibrium point. The human-vehicle function allocation is adopted to adjust the two subsystems' output weights based on the fuzzy rules, to restrain the suspension roll motion and adaptively compensating driver's steering torque. The considerable simulations are carried out, and the results demonstrate that the suspension and EPS coordinated control system by applying bilinear control can obtain better performance than another nonlinear control method; the human-vehicle function allocation can further improve the vehicle whole-region control performance.
Yang L.,Hefei University of Technology |
Yang L.,Anhui Communications Vocational and Technical College |
Chen W.,Hefei University of Technology |
Gao Z.,Hefei University of Technology |
Chen Y.,Hefei University of Technology
Nongye Jixie Xuebao/Transactions of the Chinese Society for Agricultural Machinery | Year: 2014
On the basis of experiments on force-speed characteristics of solenoid valve damper, a slide model control strategy based on saturated input was put forward for a quarter vehicle semi-active suspension with solenoid valve damper, which considered the nonlinear properties of vehicle semi-active suspension and output saturation properties of solenoid valve damper with adjustable damping force. A quarter vehicle nonlinear model with semi-active suspension and a suspension reference model with simplified input were built. A slide model controller was designed for a quarter vehicle nonlinear model. Meanwhile, considering the damping force output saturation property of solenoid valve damper, an assistant analysis system was designed to generate control compensation signal in order to compensate the saturation of the slide model controller. The results of simulation under Matlab/Simulink and experiments showed that the slide model controller based on saturated input could effectively eliminate the influence of output saturation of solenoid valve damper, and make the performance indexes, such as vehicle body vertical acceleration and suspension working space, close to the corresponding ideal output of a suspension reference model. It could also optimize the semi-active suspension design and control so as to improve vehicle ride comfort.
Chen L.,Hefei University of Technology |
Zhang J.,Anhui Communications Vocational and Technical College
Advanced Materials Research | Year: 2011
Earthquake ground motions were reasonally selected for the nonlinear dynamic time-history analysis conducted for a two-tower long-span cable-stayed bridge. The correlation between frequency content of earthquake ground motions and probabilistic seismic demands reveals that the geometric mean and dispersion of response spectra from earthquake ground motions have significant effects on probabilistic seismic demand assessment of long-span bridge stuctures and these effects are related to the difference of the shape of geometric mean spectra in the important period ranges where cumulative modal mass participation is significant. Response spectra of selected earthquake ground motions should match well with target spectra in the important period ranges. If input ground motions are reasonally selected, analytic results can be obtained more precisely and effectively and more amplitude parameters can be selected as intensity measures. © (2011) Trans Tech Publications, Switzerland.
Li X.,Hefei University of Technology |
Wang B.,Anhui Communications Vocational and Technical College
Journal of Earthquake Engineering and Engineering Vibration | Year: 2016
Optimal control is a method to solve the optimal vibration control input by using maximum principle, optimal filtering or dynamic programming optimization. Controller designed based on optimal control theory can only achieve sub-optimal control, because optimal control is built on the ideal mathematical model and the actual structure is often a reduced order model with a variety of constraints. The emergence and development of the iterative learning control theory provides a new method for structure vibration control. But its control effect and convergence restrict its application. With respective advantages of iterative learning control and quadratic optimal control, we combine them and obtain a new control strategy, which is named linear quadratic iterative learning control (LQILC). The new mixed control strategy enhances the stability and robustness of the iterative learning control system, improves the speed of convergence and the control effect of the quadratic optimal control. Using the linear quadratic iterative learning control to control the Emerson Memorial Bridge against earthquake and calculate the benchmark performance indicators. The result show that the new control strategies are able to effectively control the Emerson Memorial Bridge against earthquake and the control effect is improved.
Kan P.,Anhui Communications Vocational and Technical College |
Qiu L.,Hefei University of Technology |
Qian L.,Hefei University of Technology
IEEE Transportation Electrification Conference and Expo, ITEC Asia-Pacific 2014 - Conference Proceedings | Year: 2014
In this paper, control strategy of a plug-in 4 wheel drive car is researched. According to torque request of the car, the steady work state is classified into 10 modes, which consists of 7 drive modes and 3 brake modes. Besides, torque distribution strategy among the power components which realizes basic energy management and torque coordination control which aims at reducing shock extent when mode changes are put forward and discussed in detail. In the end, Matlab/Simulink is used to create a forward simulation model of the car. The simulation results indicate that the strategy is effective and feasible. © 2014 IEEE.
Wang F.-S.,Anhui Communications Vocational and Technical College
Chang'an Daxue Xuebao (Ziran Kexue Ban)/Journal of Chang'an University (Natural Science Edition) | Year: 2011
In order to forecasting the embankment settlement accurately, based on the consolidation theory of soil, the additional stress factor is introduced to improve the traditional delamination summation-method of the total embankment settlement, and the improved calculation theory and formula are put forward. A new calculation and prediction model is built based on the improved delamination summation-method and practical survey. The calculation program Settlement_Calculation for the calculation and prediction of embankment settlement is wrote based on the improved delamination summation-method and practical survey data. The survey and theory calculation for the test section of a secondary road in Anhui province is carried out adopting the calculation method and program. The results show that the error of the prediction value compared to the one of survey is less than 2.0%.
Lou Z.,Anhui Communications Vocational and Technical College
Applied Mechanics and Materials | Year: 2012
This paper introduces a digital signal processor based system design, through the computer acquisition individuals EMG(Electromyographic)signal data to monitor the dynamic activity of muscles, and the estimation of normal and pathological conditions of the acquired data of the power spectrum. In the collection system, the input signal of the two differential amplifier signal preprocessing and obtain the EMG signal linear feature. Obtain electromyographic signal is then digitized, and transmitted to the computer analysis. The optimization system for detection and Analysis on the clinical diagnosis of human functional status and rehabilitation has important significance. © (2012) Trans Tech Publications.