Convective storms and lightning are among the most important weather phenomena that are challenging to forecast.In this study,a novel multi-task learning(MTL)encoder-decoder U-net neural network was developed to forec...Convective storms and lightning are among the most important weather phenomena that are challenging to forecast.In this study,a novel multi-task learning(MTL)encoder-decoder U-net neural network was developed to forecast convective storms and lightning with lead times for up to 90 min,using GOES-16 geostationary satellite infrared brightness temperatures(IRBTs),lightning flashes from Geostationary Lightning Mapper(GLM),and vertically integrated liquid(VIL)from Next Generation Weather Radar(NEXRAD).To cope with the heavily skewed distribution of lightning data,a spatiotemporal exponent-weighted loss function and log-transformed lightning normalization approach were developed.The effects of MTL,single-task learning(STL),and IRBTs as auxiliary input features on convection and lightning nowcasting were investigated.The results showed that normalizing the heavily skew-distributed lightning data along with a log-transformation dramatically outperforms the min-max normalization method for nowcasting an intense lightning event.The MTL model significantly outperformed the STL model for both lightning nowcasting and VIL nowcasting,particularly for intense lightning events.The MTL also helped delay the lightning forecast performance decay with the lead times.Furthermore,incorporating satellite IRBTs as auxiliary input features substantially improved lightning nowcasting,but produced little difference in VIL forecasting.Finally,the MTL model performed better for forecasting both lightning and the VIL of organized convective storms than for isolated cells.展开更多
As the proportion of renewable energy(RE)increases,the inertia and the primary frequency regulation(FR)capability of the power system decrease.Thus,ensuring frequency security in the scheduling model has become a new ...As the proportion of renewable energy(RE)increases,the inertia and the primary frequency regulation(FR)capability of the power system decrease.Thus,ensuring frequency security in the scheduling model has become a new technical requirement in power systems with a high share of RE.Due to a shortage of conventional synchronous generators,the frequency support of multi-source converters has become an indispensable part of the system frequency resources,especially variable-speed wind turbine generation(WTG)and battery energy storage(BES).Quantitative expression of the FR capability of multi-source converters is necessary to construct frequency-constrained scheduling model.However,the frequency support performance of these converter-interfaced devices is related to their working states,operation modes,and parameters,and the complex coupling of these factors has not been fully exploited in existing models.In this study,we propose an integrated frequency-constrained scheduling model considering the coordination of FR capabilities from multi-source converters.Switchable FR control strategies and variable FR parameters for WTG with or without reserved power are modeled,and multi-target allocation of BES capacity between tracking dispatch instruction and emergency FR is analyzed.Then,the variable FR capabilities of WTG and BES are embedded into the integrated frequency-constrained scheduling model.The nonlinear constraints for frequency security are precisely linearized through an improved iteration-based strategy.The effectiveness of the proposed model is verified in a modified IEEE 24-bus standard system.The results suggest that the coordinated participation of BES and WTG in FR can effectively reduce the cost of the scheduling model while meeting frequency security constraints.展开更多
基金supported by the Science and Technology Grant No.520120210003,Jibei Electric Power Company of the State Grid Corporation of China。
文摘Convective storms and lightning are among the most important weather phenomena that are challenging to forecast.In this study,a novel multi-task learning(MTL)encoder-decoder U-net neural network was developed to forecast convective storms and lightning with lead times for up to 90 min,using GOES-16 geostationary satellite infrared brightness temperatures(IRBTs),lightning flashes from Geostationary Lightning Mapper(GLM),and vertically integrated liquid(VIL)from Next Generation Weather Radar(NEXRAD).To cope with the heavily skewed distribution of lightning data,a spatiotemporal exponent-weighted loss function and log-transformed lightning normalization approach were developed.The effects of MTL,single-task learning(STL),and IRBTs as auxiliary input features on convection and lightning nowcasting were investigated.The results showed that normalizing the heavily skew-distributed lightning data along with a log-transformation dramatically outperforms the min-max normalization method for nowcasting an intense lightning event.The MTL model significantly outperformed the STL model for both lightning nowcasting and VIL nowcasting,particularly for intense lightning events.The MTL also helped delay the lightning forecast performance decay with the lead times.Furthermore,incorporating satellite IRBTs as auxiliary input features substantially improved lightning nowcasting,but produced little difference in VIL forecasting.Finally,the MTL model performed better for forecasting both lightning and the VIL of organized convective storms than for isolated cells.
基金supported by the National Key Research and Development Program of China(No.2021YFB2400500)the Science and Technology Project of State Grid Corporation of China“Fast control of photovoltaic and wind power plant for transient frequency/voltage support”.
文摘As the proportion of renewable energy(RE)increases,the inertia and the primary frequency regulation(FR)capability of the power system decrease.Thus,ensuring frequency security in the scheduling model has become a new technical requirement in power systems with a high share of RE.Due to a shortage of conventional synchronous generators,the frequency support of multi-source converters has become an indispensable part of the system frequency resources,especially variable-speed wind turbine generation(WTG)and battery energy storage(BES).Quantitative expression of the FR capability of multi-source converters is necessary to construct frequency-constrained scheduling model.However,the frequency support performance of these converter-interfaced devices is related to their working states,operation modes,and parameters,and the complex coupling of these factors has not been fully exploited in existing models.In this study,we propose an integrated frequency-constrained scheduling model considering the coordination of FR capabilities from multi-source converters.Switchable FR control strategies and variable FR parameters for WTG with or without reserved power are modeled,and multi-target allocation of BES capacity between tracking dispatch instruction and emergency FR is analyzed.Then,the variable FR capabilities of WTG and BES are embedded into the integrated frequency-constrained scheduling model.The nonlinear constraints for frequency security are precisely linearized through an improved iteration-based strategy.The effectiveness of the proposed model is verified in a modified IEEE 24-bus standard system.The results suggest that the coordinated participation of BES and WTG in FR can effectively reduce the cost of the scheduling model while meeting frequency security constraints.