Time filter

Source Type

Han P.,North China Electrical Power University | Gao F.,North China Electrical Power University | Gao F.,Hebei University | Zhai Y.-J.,North China Electrical Power University | And 2 more authors.
International Journal of Advancements in Computing Technology | Year: 2012

A novel regression algorithm of multi-output least squares support vector machine is proposed for the modeling of multi-output systems. The fitting error of each dimension output and the fitting error of overall output are defined. And combined with the equality constraints of least squares support vector machine, the linear equation of multi-output least squares support vector machine is deduced. A multioutput combustion system model for utility boiler is established based on above algorithm and the prediction of NOx emission, carbon content of fly ash, exhaust flue gas temperature and desuperheating water of reheater is achieved. In addition, the parameters of the multi-output model are optimized by particle swarm optimization. The performance of the multi-output least squares support vector machine model and multiple single-output least squares support vector machine are compared. The effectiveness of the multi-output model is verified and it provides a good foundation for combustion optimization.


Gao F.,North China Electrical Power University | Gao F.,Hebei University | Zhai Y.-J.,North China Electrical Power University | Zhuo Y.,Tianjin SDIC Jinneng Electrical Power Co. | And 2 more authors.
Dongli Gongcheng Xuebao/Journal of Chinese Society of Power Engineering | Year: 2012

The boiler combustion system is a complex multi-input & multi-output system. To achieve the objectives of high efficiency and low pollutant emission for the system in one model, some improvements were carried out on the standard regression algorithm of least squares support vector machine(LSSVM). Based on the field combustion adjustment data of a 1000 MW ultra supercritical boiler, a sharing LSSVM model was established by taking both the boiler thermal efficiency and mass concentration of NOx emission as the output variables, with which an improved particle swarm algorithm was selected to optimize the operating conditions of the boiler. Results show that with the sharing LSSVM model, the mean prediction error for boiler thermal efficiency and mass concentration of NOx emission can reach 0.028% and 2.16% respectively, the searched parameter combination of high thermal efficiency and low NOx emission may serve as a reference for optimization of boiler operation.


An L.,North China Electrical Power University | Wang P.,North China Electrical Power University | Jiang G.,North China Electrical Power University | Li G.,Tianjin SDIC Jinneng Electrical Power Co. | Shen G.,North China Electrical Power University
Zhongguo Dianji Gongcheng Xuebao/Proceedings of the Chinese Society of Electrical Engineering | Year: 2012

The interference of strong background noise and reflected by the surface of wall and tube rows in industrial boiler furnace, renders the generalized cross-correlation (GCC) peaks of tube leak sound pressure signals received by sensors array ambiguous, even the stable peak couldn't be searched at all, which leads to leak position may fail to accurately be fixed. The 600 MW supercritical boiler model was established, the leakage source propagation process of reflection and attenuation in boiler furnace was simulated by enhanced acoustic simulator for engineers (EASE), the approximate signal to noise ratio (SNR) was obtained through it; The reverberation time was calculated with the modified Sabine and squared impulse response integration method by the simulation. The time delay estimation algorithm: phase transformation (PHAT), maximum likelihood (ML), phase transform for noise (PTN, derived from PHAT) and SWITCH were evaluated and results revealed the superiority of SWITCH and PTN methods in reverberant and noisy boiler background. Although SWITCH is outperformed by PTN slightly, but the prior knowledge of reverberant energy to direct energy ratio and frequencies onset detection is required in PTN method. © 2012 Chin. Soc. for Elec. Eng.


An L.,North China Electrical Power University | Wang P.,North China Electrical Power University | Jiang G.,North China Electrical Power University | Li G.,Tianjin SDIC Jinneng Electrical Power Co. | Shen G.,North China Electrical Power University
Zhongguo Dianji Gongcheng Xuebao/Proceedings of the Chinese Society of Electrical Engineering | Year: 2010

The main requirement of a modern power plant is an efficient and reliable operation; hence, the source of boiler tube leakage must be precisely localized. In this study, the quaternary acoustic array was used in the furnace, a set of hyperboloidal equations was established, and the sensors distributed for capturing the leakage signal were given. The experiment on leakage localization shows that a stable and sharp peak is obtained through an approximation of the maximum likelihood estimator which signal-to-noise ratio function on frequency domain performs well in inhibition of the background noise. In order to avoid initial guess and distant source assumption, this study employed adaptive Gaussian mutation operator, which was implemented in the genetic algorithm (GA), to provide an explicit solution. Compared with the quasi-Newton method, GA is more robust and accuracy, although it requires more time to converge. The time delay estimator error greatly influences the accuracy of the localization. When the time delay estimator error is below 0.1 ms, the quaternary plane array localization error is within the permitted range of 1 m, which is derived through numerical calculation. © 2010 Chin. Soc. for Elec. Eng.


Liu D.,North China Electrical Power University | Zhao M.,Yunnan Electric Power Research Institute | Lin Y.-J.,North China Electrical Power University | Shi R.-J.,Tianjin Sdic Jinneng Electrical Power Co.
Procedia Engineering | Year: 2011

This paper analyzes a certain power plant ultra-supercritical unit feed water control system, and has setting, optimized control system. Through analysis of operation, when Load on a large scale changes and intermediate point temperature control, this unit gets a good control effect. The control system offers ultra-supercritical units feed water control scheme for reference. © 2011 Published by Elsevier Ltd.


Yang M.,Northeast Dianli University | Sun Y.,Northeast Dianli University | Mu G.,Northeast Dianli University | Yan G.,Northeast Dianli University | Zhang M.,Tianjin SDIC Jinneng Electrical Power Co.
Taiyangneng Xuebao/Acta Energiae Solaris Sinica | Year: 2015

Using wavelet theory to establish model for wind power multi-step prediction, aimed at the predictive value deteriorate with the increasing of prediction steps in the model, the real-time wavelet decomposition theory to predict different frequency component using different predictive methods was introduced, then the superimposed multi-step predictive value was got. According to the index of wind power real-time prediction of wind farm published by the National Energy Administration, the wind power data measured from wind farm of Jilin province were used to verify the validity of the method. ©, 2015, Science Press. All right reserved.


Xu H.,North China Electrical Power University | Yan S.-L.,North China Electrical Power University | Wu Q.-Y.,Tianjin SDIC Jinneng Electrical Power Co. | Li Y.-H.,North China Electrical Power University
Dongli Gongcheng Xuebao/Journal of Chinese Society of Power Engineering | Year: 2012

For the purpose of energy saving, it is meaningful to quantitatively analyze the influence of upper terminal temperature difference of a heater on the coal consumption rate in a coal-fired power plant. To overcome the deficiency of previous calculation methods, a general strength matrix calculation model was established with gradient operator based on the unified physical and mathematic model in thermal economy analysis, which was later on applied to the calculation for a 600 MW unit under typical working conditions. Results show that the model has simple structure, less calculation, high precision and strong universality; the strength matrix can be directly used to quantitatively and qualitatively analyze the influence of the upper terminal temperature difference on the coal consumption rate; the influence of terminal temperature difference on thermal economy of power plant has a relationship not only to the value of the temperature difference itself, but also to the location and structure of the heater, therefore it is suggested that great attention should be paid to the influence of terminal temperature differences with high strength factors on the unit economy.

Loading Tianjin SDIC Jinneng Electrical Power Co. collaborators
Loading Tianjin SDIC Jinneng Electrical Power Co. collaborators