针对锅炉过热汽温系统存在的大时延、强非线性和变量耦合等建模难题,建立了一种基于贝叶斯优化时间序列预测神经基扩展分析(neural basis expansion analysis for interpretable time series forecasting,N-BEATS)网络的过热汽温预测模...针对锅炉过热汽温系统存在的大时延、强非线性和变量耦合等建模难题,建立了一种基于贝叶斯优化时间序列预测神经基扩展分析(neural basis expansion analysis for interpretable time series forecasting,N-BEATS)网络的过热汽温预测模型。针对某600 MW超临界火电机组,结合机理分析确定模型的输入和输出变量,通过性能对比实验优化模型的输入/输出时延阶次、Block类型和激活函数,进一步利用贝叶斯优化算法对模型的超参数进行寻优,并与网格搜索、遗传算法的优化效果进行对比。采用该机组仿真运行数据进行建模实验,结果表明所提模型在预测精度方面优于传统优化方法及主流模型。展开更多
This paper expresses the efficient outputs of decisionmaking unit（DMU） as the sum of ＂average outputs＂ forecasted by a GM（1,N） model and ＂increased outputs＂ which reflect the difficulty to realize efficient ou...This paper expresses the efficient outputs of decisionmaking unit（DMU） as the sum of ＂average outputs＂ forecasted by a GM（1,N） model and ＂increased outputs＂ which reflect the difficulty to realize efficient outputs.The increased outputs are solved by linear programming using data envelopment analysis efficiency theories,wherein a new sample is introduced whose inputs are equal to the budget in the issue No.n ＋ 1 and outputs are forecasted by the GM（1,N） model.The shortcoming in the existing methods that the forecasted efficient outputs may be less than the possible actual outputs according to developing trends of input-output rate in the periods of pre-n is overcome.The new prediction method provides decision-makers with more decisionmaking information,and the initial conditions are easy to be given.展开更多
The dissemination of cattle brucellosis in Zhejiang province of China can be attributed to the transport of cattle between cities within the province. In this paper,an n-patch dynamical model is proposed to study the ...The dissemination of cattle brucellosis in Zhejiang province of China can be attributed to the transport of cattle between cities within the province. In this paper,an n-patch dynamical model is proposed to study the effect of cattle dispersal on brucellosis spread. Theoretically,we analyze the dynamical behavior of the muti-patch model. For the 2-patch submodel,sensitivity analyses of the basic reproduction number R0 and the number of the infectious cattle in term of model parameters are carried out. By numerical analysis,it is obtained that the dispersal of susceptible cattle between patches and the centralization of infected cattle to the large scale patch can alleviate the epidemic and are in favor of the control of disease in the whole region.展开更多
We investigate the area distribution of clusters （loops） in the honeycomb O（n） loop model by means of the worm algorithm with n = 0.5, 1, 1.5, and 2. At the critical point, the number of clusters, whose enclosed a...We investigate the area distribution of clusters （loops） in the honeycomb O（n） loop model by means of the worm algorithm with n = 0.5, 1, 1.5, and 2. At the critical point, the number of clusters, whose enclosed area is greater than A, is proportional to A-1 with a proportionality constant C. We confirm numerically that C is universal, and its value agrees well with the predictions based on the Coulomb gas method.展开更多
An N-gram Chinese language model incorporating linguistic rules is presented. By constructing elements lattice, rules information is incorporated in statistical frame. To facilitate the hybrid modeling, novel methods ...An N-gram Chinese language model incorporating linguistic rules is presented. By constructing elements lattice, rules information is incorporated in statistical frame. To facilitate the hybrid modeling, novel methods such as MI-based rule evaluating, weighted rule quantification and element-based n-gram probability approximation are presented. Dynamic Viterbi algorithm is adopted to search the best path in lattice. To strengthen the model, transformation-based error-driven rules learning is adopted. Applying proposed model to Chinese Pinyin-to-character conversion, high performance has been achieved in accuracy, flexibility and robustness simultaneously. Tests show correct rate achieves 94.81% instead of 90.53% using bi-gram Markov model alone. Many long-distance dependency and recursion in language can be processed effectively.展开更多
We investigate tricritical behavior of the O(n) model in two dimensions by means of transfer-matrix and finite-size scaling methods. For this purpose we consider an O(n) symmetric spin model on the honeycomb lattice w...We investigate tricritical behavior of the O(n) model in two dimensions by means of transfer-matrix and finite-size scaling methods. For this purpose we consider an O(n) symmetric spin model on the honeycomb lattice with vacancies; the tricritical behavior is associated with the percolation threshold of the vacancies. The vacancies are represented by face variables on the elementary hexagons of thelattice. We apply a mapping of the spin degrees of freedom model on a non-intersecting-loop model, in which the number n of spin components assumes the role of a continuously variable parameter. This loop model serves as a suitable basis for the construction of the transfer matrix.Our results reveal the existence of a tricritical line, parametrized by n, which connects the known universality classes of the tricritical Ising model and the theta point describing the collapse of a polymer. On the other side of theIsing point,the tricritical line extends to the n = 2 point describing a tricritical O(2) model.展开更多
In order to numerically simulate daily nitrous oxide （N2O） emission from a rice-winter wheat rotation cropping system, a process-based site model was developed （referred to as IAP-N-GAS） to track the movement and ...In order to numerically simulate daily nitrous oxide （N2O） emission from a rice-winter wheat rotation cropping system, a process-based site model was developed （referred to as IAP-N-GAS） to track the movement and transformation of several forms of nitrogen in the agro-eeosystem, which is affected by climate, soil, crop growth and management practices. The simulation of daily N2O fluxes, along with key daily environmental variables, was validated with three-year observations conducted in East China. The validation demonstrated that the model simulated well daily solar radiation, soil temperature and moisture, and also captured the dynamics and magnitude of accumulated rice aboveground biomass and mineral nitrogen in the soil. The simulated daily N2O emissions over all three years investigated were generally in good agreement with field observations. Particularly well simulated were the peak N2O emissions induced by fertilizations, rainfall events or mid-season drainages. The model simulation also represented closely the inter-annuM variation in N2O emission. These validations imply that the model has the capability to capture the general characteristics of N2O emission from a typical rice-wheat rotation agro-ecosystem. Sensitivity analyses revealed that the simulated N2O emission is most sensitive to the fertilizer application rate and the soil organic matter content, but it is much less sensitive to variations in soil pH and texture, temperature, precipitation and crop residue incorporation rate under local conditions.展开更多
文摘针对锅炉过热汽温系统存在的大时延、强非线性和变量耦合等建模难题,建立了一种基于贝叶斯优化时间序列预测神经基扩展分析(neural basis expansion analysis for interpretable time series forecasting,N-BEATS)网络的过热汽温预测模型。针对某600 MW超临界火电机组,结合机理分析确定模型的输入和输出变量,通过性能对比实验优化模型的输入/输出时延阶次、Block类型和激活函数,进一步利用贝叶斯优化算法对模型的超参数进行寻优,并与网格搜索、遗传算法的优化效果进行对比。采用该机组仿真运行数据进行建模实验,结果表明所提模型在预测精度方面优于传统优化方法及主流模型。
基金supported by the Research Start Funds for Introducing High-level Talents of North China University of Water Resources and Electric Power
文摘This paper expresses the efficient outputs of decisionmaking unit（DMU） as the sum of ＂average outputs＂ forecasted by a GM（1,N） model and ＂increased outputs＂ which reflect the difficulty to realize efficient outputs.The increased outputs are solved by linear programming using data envelopment analysis efficiency theories,wherein a new sample is introduced whose inputs are equal to the budget in the issue No.n ＋ 1 and outputs are forecasted by the GM（1,N） model.The shortcoming in the existing methods that the forecasted efficient outputs may be less than the possible actual outputs according to developing trends of input-output rate in the periods of pre-n is overcome.The new prediction method provides decision-makers with more decisionmaking information,and the initial conditions are easy to be given.
基金supported by the National Natural Science Foundation of China under Grant（11331009,11171314,11147015,11301490） the National Youth Natural Science Foundation（11201434）+1 种基金 the Specialized Research Fund for the Doctoral Program of Higher Education（20121420130001） the Research Project Supported by Shanxi Scholarship Council of China（2013-3）
文摘The dissemination of cattle brucellosis in Zhejiang province of China can be attributed to the transport of cattle between cities within the province. In this paper,an n-patch dynamical model is proposed to study the effect of cattle dispersal on brucellosis spread. Theoretically,we analyze the dynamical behavior of the muti-patch model. For the 2-patch submodel,sensitivity analyses of the basic reproduction number R0 and the number of the infectious cattle in term of model parameters are carried out. By numerical analysis,it is obtained that the dispersal of susceptible cattle between patches and the centralization of infected cattle to the large scale patch can alleviate the epidemic and are in favor of the control of disease in the whole region.
基金Project supported by the National Natural Science Foundation of China (Grant No. 10975127)the Specialized Research Fund for the Doctoral Program of Higher Education, China (Grant No. 20113402110040)
文摘We investigate the area distribution of clusters （loops） in the honeycomb O（n） loop model by means of the worm algorithm with n = 0.5, 1, 1.5, and 2. At the critical point, the number of clusters, whose enclosed area is greater than A, is proportional to A-1 with a proportionality constant C. We confirm numerically that C is universal, and its value agrees well with the predictions based on the Coulomb gas method.
文摘An N-gram Chinese language model incorporating linguistic rules is presented. By constructing elements lattice, rules information is incorporated in statistical frame. To facilitate the hybrid modeling, novel methods such as MI-based rule evaluating, weighted rule quantification and element-based n-gram probability approximation are presented. Dynamic Viterbi algorithm is adopted to search the best path in lattice. To strengthen the model, transformation-based error-driven rules learning is adopted. Applying proposed model to Chinese Pinyin-to-character conversion, high performance has been achieved in accuracy, flexibility and robustness simultaneously. Tests show correct rate achieves 94.81% instead of 90.53% using bi-gram Markov model alone. Many long-distance dependency and recursion in language can be processed effectively.
文摘We investigate tricritical behavior of the O(n) model in two dimensions by means of transfer-matrix and finite-size scaling methods. For this purpose we consider an O(n) symmetric spin model on the honeycomb lattice with vacancies; the tricritical behavior is associated with the percolation threshold of the vacancies. The vacancies are represented by face variables on the elementary hexagons of thelattice. We apply a mapping of the spin degrees of freedom model on a non-intersecting-loop model, in which the number n of spin components assumes the role of a continuously variable parameter. This loop model serves as a suitable basis for the construction of the transfer matrix.Our results reveal the existence of a tricritical line, parametrized by n, which connects the known universality classes of the tricritical Ising model and the theta point describing the collapse of a polymer. On the other side of theIsing point,the tricritical line extends to the n = 2 point describing a tricritical O(2) model.
基金supported by the Chinese Academy of Sciences (KZCX2-YW-204, KSCX3-SW-440, KZCX1-SW-01)the National Natural Science Foundation of China (40425010, 40331014)+1 种基金the European Union (NitroEurope IP 017841)the Helmholtz Society via the Sino-German Joint Laboratory project ENTRANCE
文摘In order to numerically simulate daily nitrous oxide （N2O） emission from a rice-winter wheat rotation cropping system, a process-based site model was developed （referred to as IAP-N-GAS） to track the movement and transformation of several forms of nitrogen in the agro-eeosystem, which is affected by climate, soil, crop growth and management practices. The simulation of daily N2O fluxes, along with key daily environmental variables, was validated with three-year observations conducted in East China. The validation demonstrated that the model simulated well daily solar radiation, soil temperature and moisture, and also captured the dynamics and magnitude of accumulated rice aboveground biomass and mineral nitrogen in the soil. The simulated daily N2O emissions over all three years investigated were generally in good agreement with field observations. Particularly well simulated were the peak N2O emissions induced by fertilizations, rainfall events or mid-season drainages. The model simulation also represented closely the inter-annuM variation in N2O emission. These validations imply that the model has the capability to capture the general characteristics of N2O emission from a typical rice-wheat rotation agro-ecosystem. Sensitivity analyses revealed that the simulated N2O emission is most sensitive to the fertilizer application rate and the soil organic matter content, but it is much less sensitive to variations in soil pH and texture, temperature, precipitation and crop residue incorporation rate under local conditions.
