为提高热负荷预测的精度,提升基于负荷预测的供热系统调控效果,提出一种基于特征融合的供热系统预测调控方法。首先,采用偏自相关函数(partial autocorrelation function,PACF)、Pearson相关系数和最大信息系数(maximum information coe...为提高热负荷预测的精度,提升基于负荷预测的供热系统调控效果,提出一种基于特征融合的供热系统预测调控方法。首先,采用偏自相关函数(partial autocorrelation function,PACF)、Pearson相关系数和最大信息系数(maximum information coefficient,MIC)相结合的特征选择方法来确定预测模型的基本特征;然后,使用线性回归融合、指数融合和主成分分析融合对基本特征进行融合,应用递归MLR预测确定最佳融合方法,进一步对比在最佳融合策略下递归MLR、PSO-SVR、CNN和XGBoost中效果最优的预测方法;最后,将辨识出的融合方法和预测模型方法用于实际热力站调控。结果显示,基于线性回归融合的XGboost预测方法效果最好,可以提升训练精度并减少计算时间,同时可以有效指导调控,节热率达到4%以上。展开更多
The atmospheric surface layer of the tropical coastal ocean is commonly very unstable and experiences weakwind conditions.How the latent(LE)and sensible(H)heat fluxes behave under such conditions are unclear because o...The atmospheric surface layer of the tropical coastal ocean is commonly very unstable and experiences weakwind conditions.How the latent(LE)and sensible(H)heat fluxes behave under such conditions are unclear because of the lack of observation stations in the tropics.Thus,this study aims to analyze LE and H and the microclimate parameters influencing them.The authors deployed an eddy covariance system in a tropical coastal region for seven months.The microclimate parameters investigated were wind speed(U),vapor pressure deficit(Δe),temperature difference(ΔT),wind-vapor pressure deficit(UΔe),wind-temperature difference(UΔT),and atmospheric stability(z/L),where z is height and L is the Monin–Obukhov length.On the daily time scale,the results show that LE was more associated with U thanΔe,while H was more related toΔT than U.Cross-wavelet analysis revealed the strong coherence in the LE-U relationship for periods between one and two days,and for H–ΔT,0.5 to 1 day.Correlation and regression analyses confirmed the time series analyses results,where strong positive correlation coefficients(r)were obtained between LE and U(r=0.494)and H andΔT(r=0.365).Compared to other water bodies,the transfer coefficient of moisture(CE N)was found to be small(=0.40×10^(-3))and independent of stability;conversely,the transfer coefficient of heat(CH N)was closer to literature values(=1.00×10^(-3))and a function of stability.展开更多
文摘为提高热负荷预测的精度,提升基于负荷预测的供热系统调控效果,提出一种基于特征融合的供热系统预测调控方法。首先,采用偏自相关函数(partial autocorrelation function,PACF)、Pearson相关系数和最大信息系数(maximum information coefficient,MIC)相结合的特征选择方法来确定预测模型的基本特征;然后,使用线性回归融合、指数融合和主成分分析融合对基本特征进行融合,应用递归MLR预测确定最佳融合方法,进一步对比在最佳融合策略下递归MLR、PSO-SVR、CNN和XGBoost中效果最优的预测方法;最后,将辨识出的融合方法和预测模型方法用于实际热力站调控。结果显示,基于线性回归融合的XGboost预测方法效果最好,可以提升训练精度并减少计算时间,同时可以有效指导调控,节热率达到4%以上。
基金supported by a PETRONAS-Academia Collabora-tion Dialogue 2022 Grant[Grant number PACD 2022]from PETRONAS Research Sdn.Bhd。
文摘The atmospheric surface layer of the tropical coastal ocean is commonly very unstable and experiences weakwind conditions.How the latent(LE)and sensible(H)heat fluxes behave under such conditions are unclear because of the lack of observation stations in the tropics.Thus,this study aims to analyze LE and H and the microclimate parameters influencing them.The authors deployed an eddy covariance system in a tropical coastal region for seven months.The microclimate parameters investigated were wind speed(U),vapor pressure deficit(Δe),temperature difference(ΔT),wind-vapor pressure deficit(UΔe),wind-temperature difference(UΔT),and atmospheric stability(z/L),where z is height and L is the Monin–Obukhov length.On the daily time scale,the results show that LE was more associated with U thanΔe,while H was more related toΔT than U.Cross-wavelet analysis revealed the strong coherence in the LE-U relationship for periods between one and two days,and for H–ΔT,0.5 to 1 day.Correlation and regression analyses confirmed the time series analyses results,where strong positive correlation coefficients(r)were obtained between LE and U(r=0.494)and H andΔT(r=0.365).Compared to other water bodies,the transfer coefficient of moisture(CE N)was found to be small(=0.40×10^(-3))and independent of stability;conversely,the transfer coefficient of heat(CH N)was closer to literature values(=1.00×10^(-3))and a function of stability.
