In this study,a machine learning based method is proposed for creating synthetic eventful phasor measurement unit(PMU)data under time-varying load conditions.The proposed method leverages generative adversarial networ...In this study,a machine learning based method is proposed for creating synthetic eventful phasor measurement unit(PMU)data under time-varying load conditions.The proposed method leverages generative adversarial networks to create quasi-steady states for the power system under slowly-varying load conditions and incorporates a framework of neural ordinary differential equations(ODEs)to capture the transient behaviors of the system during voltage oscillation events.A numerical example of a large power grid suggests that this method can create realistic synthetic eventful PMU voltage measurements based on the associated real PMU data without any knowledge of the underlying nonlinear dynamic equations.The results demonstrate that the synthetic voltage measurements have the key characteristics of real system behavior on distinct time scales.展开更多
Power systems around the world have been registering a degenerating inertial response in view of the growth of inverter-based resources along with the withdrawal of conventional coal units.Therefore,there is a need fo...Power systems around the world have been registering a degenerating inertial response in view of the growth of inverter-based resources along with the withdrawal of conventional coal units.Therefore,there is a need for swift frequency support and its control,preferably by means of power electronic-interfaced storage devices,owing to their beneficial capabilities.Despite being particularly efficient,pragmatically,the traditional model-based non-linear control techniques are not highly popular in power system control design,primarily due to the complications faced in obtaining accurately suitable models for certain power system components.Lately,the modelfree Koopman operator-based model predictive control(KMPC)has proven to be highly conducive for data-driven non-linear control design.The principle behind KMPC is to change the coordinates in a manner to get an approximately linear model,which can then be controlled using a linear model predictive control.In this study,we employed time-delayed embedding of measurements to reconstruct a new set of preferable coordinates,thereby suggesting an approach for finding the optimal number of time lags and the embedding dimensions which are the key parameters of this algorithm.The efficacy of this KMPC framework is established by adopting a decentralized frequency control problem through a decoupled synchronous machine system,which we proposed for both the Kundur two-area system as well as the IEEE 39-bus test system.展开更多
基金supported by the National Science Foundation(No.OAC-1934675,No.ECCS-2035688,No.ECCS-1611301)。
文摘In this study,a machine learning based method is proposed for creating synthetic eventful phasor measurement unit(PMU)data under time-varying load conditions.The proposed method leverages generative adversarial networks to create quasi-steady states for the power system under slowly-varying load conditions and incorporates a framework of neural ordinary differential equations(ODEs)to capture the transient behaviors of the system during voltage oscillation events.A numerical example of a large power grid suggests that this method can create realistic synthetic eventful PMU voltage measurements based on the associated real PMU data without any knowledge of the underlying nonlinear dynamic equations.The results demonstrate that the synthetic voltage measurements have the key characteristics of real system behavior on distinct time scales.
文摘Power systems around the world have been registering a degenerating inertial response in view of the growth of inverter-based resources along with the withdrawal of conventional coal units.Therefore,there is a need for swift frequency support and its control,preferably by means of power electronic-interfaced storage devices,owing to their beneficial capabilities.Despite being particularly efficient,pragmatically,the traditional model-based non-linear control techniques are not highly popular in power system control design,primarily due to the complications faced in obtaining accurately suitable models for certain power system components.Lately,the modelfree Koopman operator-based model predictive control(KMPC)has proven to be highly conducive for data-driven non-linear control design.The principle behind KMPC is to change the coordinates in a manner to get an approximately linear model,which can then be controlled using a linear model predictive control.In this study,we employed time-delayed embedding of measurements to reconstruct a new set of preferable coordinates,thereby suggesting an approach for finding the optimal number of time lags and the embedding dimensions which are the key parameters of this algorithm.The efficacy of this KMPC framework is established by adopting a decentralized frequency control problem through a decoupled synchronous machine system,which we proposed for both the Kundur two-area system as well as the IEEE 39-bus test system.