Hengyang, China
Hengyang, China

Time filter

Source Type

Long W.,Central South University | Liang X.-M.,Central South University | Long Z.-Q.,Hengyang Normal College | Qin H.-Y.,Central South University
Kongzhi yu Juece/Control and Decision | Year: 2012

A hybrid evolutionary algorithm based on gradient descends method is proposed to determine the structure of RBF neural network and optimize its parameters. A gradient descend operator is introduced into the evolutionary algorithm. Gradient descend method is carried on search by certain probability for certain elitists of every generation to strengthen the ability of local search. The structure and parameters of RBF network are trained and optimized together by using hybrid evolutionary algorithm. The experimental results show that the RBF neural network has stronger generalization ability.


Long W.,Central South University | Liang X.,Central South University | Long Z.,Hengyang Normal College | Li Z.,Central South University
Dianli Zidonghua Shebei/Electric Power Automation Equipment | Year: 2011

The boiler combustion process of coal-fired power plant is a very complicated MIMO system with high nonlinearity and strong coupling. The LSSVM(Least Square Support Vector Machine) is applied to build the boiler combustion model based on the property test data and the nonlinear MPC(Model Predictive Control) is applied to optimize the control of boiler combustion process. The improved ACO(Ant Colony Optimization) is proposed to solve the nonlinear optimization problem of MPC algorithm, which extracts the target individuals dynamically and stochastically to lead the global search of ant colony while carries out the small step search nearby the optimal ant. Case study indicates its effectiveness.


Long Z.-Q.,Hengyang Normal College
Wuli Xuebao/Acta Physica Sinica | Year: 2011

For a chaotic system with nonlinearity and uncertainty, it is difficult to obtain the satisfactory performance using general control methods. A least square support vector marchine (LSSVM) control method based on particle swarm optimigation(PSO), is proposed for chaos control. Optimizing two parameters of LSSVM model by PSO abilities of the fast convergence and whole optimization, thus aroiding the blindness of man-made choice, the LSSVM-PSO model can enhance the capability of forecasting. The proposed method does not need any analytic model, and it is still effective in the presence of measurement noises. Simulation results with a Logistic mapping and Henon attractor show the effectiveness and feasibility of this method. © 2011 Chinese Physical Society.


Long W.,Central South University | Liang X.-M.,Central South University | Long Z.-Q.,Hengyang Normal College
Zhongnan Daxue Xuebao (Ziran Kexue Ban)/Journal of Central South University (Science and Technology) | Year: 2012

O 2 content in flue gas is a main factor that has great impacts on the safety and economical efficiency of boiler operation. Together with many other complicated factors. Building a model to predict O 2 content in flue gas is a good way to realize the normal operation of boiler. Using the data of boiler operation, a least square support vector machine (LSSVM) model of the boiler oxygen content property was developed based on gas oxygen characteristic. After that, combined with the particle swarm optimization algorithm (PSO), the O 2 content in flue gas of boiler was controlled. Simulation results show that the proposed method can more accurately measure and control the O 2 content in flue gas of boiler, and provide a new way to optimize and control process of boiler combustion in close-loop.


Long W.,Central South University | Liang X.,Central South University | Long Z.,Hengyang Normal College | Li S.,Central South University
Huagong Xuebao/CIESC Journal | Year: 2011

A capacitance model and a steady-state model of oil production process were established based on the historical data accumulated during oil production. A set point optimization problem was formulated based on the proposed model of capacitance and oil production process by minimizing production cost or maximizing profits. A multi-objective constrained optimization evolutionary algorithm based on dynamical selection and replacement strategy was proposed for solving the set point optimization problem of oil production process. The constrained optimization problem was converted into a multi-objective optimization problem with two objectives. During the evolution process, the algorithm was based on multi-objective optimization technique, where the initial population was divided into two sets. A non-dominated individual's conservation bias strategy was used to keep a specific number of infeasible solutions in each generation. The randomly selected individuals in Pareto set were replaced by the remaining non-dominated individuals. Numerical simulation was made to verify the proposed algorithm using a set of data obtained from a heterogeneous reservoir Synfield. © All Rights Reserved.


Long W.,Central South University | Liang X.-M.,Central South University | Long Z.-Q.,Hengyang Normal College | Li Z.-H.,Central South University
Zhongnan Daxue Xuebao (Ziran Kexue Ban)/Journal of Central South University (Science and Technology) | Year: 2011

An optimization method based on the modified ant colony optimization (MACO) algorithm was used to select the two parameters of least square support vector machine (LSSVM) model. In this method, the parameters of LSSVM model were considered the position vector of ants. Target individuals which lead the ant colony to do global rapid search were determined by dynamic and stochastic extraction, and the optimal ant of this generation searched in small step nearly. The optimal parameter value was obtained by MACO and modified ant colony optimization-least square support vector machine (MACO-LSSVM) forecasting model was obtained. The proposed model is applied to the short-term electrical power load forecasting problem. Every hour's load from 2009-08-01 to 2009-08-30 of area in Hunan province was taken as the sample data to be analyzed. The results indicate that the root-mean-square relative error of the proposed method is only 1.71%, which is less than those of BP and SVM model by 1.61% and 1.05%, respectively.


Zhao Y.-H.,University of South China | Liu L.-Y.,University of South China | Huang S.,Hengyang Normal College | Liu C.-X.,Nuclear Power Institute of China | And 2 more authors.
Yuanzineng Kexue Jishu/Atomic Energy Science and Technology | Year: 2014

For the measurement of neutron fluence rate in reactor and reliable assurance of γ count measurement of activated 55Mn-58Ni alloy irradiated foils in reactor, 9-channel amplifier-discriminator was developed. The main technical parameter test and application test show that the gain of each channel amplifier-discriminator is continuously adjustable from 1 to 21, the threshold of each discriminator circuit is continuously adjustable, the maximum count rate and sensitivity of discriminator circuit are high, and the system has stable property and excellent anti-interference. In conclusion, relevant technical parameters can guarantee the real-time and long-term stable measurement of neutron fluence rate relative distribution in reactor, with the technical parameters that gain stability of amplifier is less than 1%, the minimum input pulse width of discriminator circuit is greater than 0.1 μs, and the maximum count rate of discriminator is less than 4×106 s-1.


Zhao X.,University of South China | Huang S.,Hengyang Normal College | Liu L.,University of South China | Liu C.,Nuclear Power Institute of China | Zong S.,Nuclear Power Institute of China
Hedongli Gongcheng/Nuclear Power Engineering | Year: 2014

To realize the relative distribution measurement of neutron flux in reactor core, the measurement device for multi-channel neutron flux relative distribution has been developed, based on the radioactive counting of activated detection slice measured by NaI detector. The measurement device is made up of PTMC12 data-acquisition board, industrial PC and MNFDAS software, which can automatically achieve the timing and counting function in both cyclical and non-cyclical models, meanwhile the measuring results can be saved as both graphic and data files. The test shows that the stability of the measurement device is excellent with relative deviation within 1%, which can guarantee the requirement of real-time and long-term stable measurement of relative distribution measurement of neutron flux in reactor.


Hu J.,University of South China | Hu J.,Hunan Normal University | Ding Y.,Hunan Normal University | Qian S.,Hunan Normal University | Tang X.,Hengyang Normal College
Ultrasonics | Year: 2013

The control problem in ultrasound therapy is to destroy the tumor tissue while not harming the intervening healthy tissue with a desired temperature elevation. The objective of this research is to present a robust and feasible method to control the temperature distribution and the temperature elevation in treatment region within the prescribed time, which can improve the curative effect and decrease the treatment time for heating large tumor (≥2.0 cm in diameter). An adaptive self-tuning-regulator (STR) controller has been introduced into this control method by adding a time factor with a recursive algorithm, and the speed of sound and absorption coefficient of the medium is considered as a function of temperature during heating. The presented control method is tested for a self-focused concave spherical transducer (0.5 MHz, 9 cm aperture, 8.0 cm focal length) through numerical simulations with three control temperatures of 43 °C, 50 °C and 55 °C. The results suggest that this control system has adaptive ability for variable parameters and has a rapid response to the temperature and acoustic power output in the prescribed time for the hyperthermia interest. There is no overshoot during temperature elevation and no oscillation after reaching the desired temperatures. It is found that the same results can be obtained for different frequencies and temperature elevations. This method can obtain an ellipsoid-shaped ablation region, which is meaningful for the treatment of large tumor. © 2012 Elsevier B.V. All rights reserved.


Long W.,Guizhou University of Finance and Economics | Long W.,Central South University | Liang X.-M.,Central South University | Long Z.-Q.,Hengyang Normal College | Yan G.,Central South University
Xitong Gongcheng Lilun yu Shijian/System Engineering Theory and Practice | Year: 2013

To solve the problems of the uncertain parameters of LSSVM and the low forecasting precision of single method, the learning algorithm of grey least squares support vector machines combined forecasting model optimized by particle swarm algorithm is proposed. Optimize two parameters of LSSVM model study by particle swarm algorithm's abilities of the fast convergence and whole optimization. It can escape from the blindness of man-made choice. First, the combinational results of initial forecasts are put as the input and the corresponding actual values are put as the output of LSSVM. Then we can get combinational model of the grey and the least squares support vector machine based on particle swarm algorithm by training it. The proposed combinational model can enhance the efficiency and the capability of forecasting. Actual data from 1985 to 2006 of area in Sanjiang plain is taken as the sample data. A combinational model based on PSO-LSSVM and GM(1, 1) model is proposed. Predict precision of the model is examined by two ways, and the results show that it is more precise than the other methods.

Loading Hengyang Normal College collaborators
Loading Hengyang Normal College collaborators