In the context of high penetration of renewables, the need to build dynamic models of power system components based on accessible measurement data has become urgent. To address this challenge, firstly, a neural ordinary differential equations (ODE) module and a neural differential-algebraic equations (DAE) module are proposed to form a data-driven modeling framework that accurately captures components' dynamic characteristics and flexibly adapts to various interface settings. Secondly, analytical models and data-driven models learned by the neural ODE and DAE modules are integrated together and simulated simultaneously using unified transient stability simulation methods. Finally, the neural ODE and DAE modules are implemented with Python and made public on GitHub. Using the portal measurements, three simple but representative cases of excitation controller modeling, photovoltaic power plant modeling, and equivalent load modeling of a regional power network are carried out in the IEEE-39 system and 2383wp system. Neural dynamic model-integrated simulations are compared with the original model-based ones to verify the feasibility and potentiality of the proposed neural ODE and DAE modules.

With the rapid development of artificial intelligence (AI), it is foreseeable that the accuracy and efficiency of dynamic analysis for future power system will be greatly improved by the integration of dynamic simulators and AI. To explore the interaction mechanism of power system dynamic simulations and AI, a general design for AI-oriented power system dynamic simulators is proposed, which consists of a high-performance simulator with neural network supportability and flexible external and internal application programming interfaces (APIs). With the support of APIs, simulation-assisted AI and AI-assisted simulation form a comprehensive interaction mechanism between power system dynamic simulations and AI. A prototype of this design is implemented and made public based on a highly efficient electromechanical simulator. Tests of this prototype are carried out in four scenarios including sample generation, AI-based stability prediction, data-driven dynamic component modeling, and AI-aided stability control, which prove the validity, flexibility, and efficiency of the design and implementation for AI-oriented power system dynamic simulators.

Parallel computing is an effective tool to improve the efficiency of transient stability simulations. The method of forming the bordered block diagonal form matrix (BBDF method) is widely used. However, as the number of subnets increases, the computational complexity of the cut-node network and the communication overhead become large quickly, which limits the efficiency of the method. A new fully parallel BBDF (FBBDF) method is proposed to solve this problem. During the iterative solutions of network equations, the FBBDF method parallelizes the solutions of subnets and the solution of the cut-node network by utilizing the result of the cut-node network at previous iteration step for the solutions of subnets at current iteration step. The calculation time can be reduced from the total solution time of the largest subnet and the cut-node network to the largest solution time of subnets and the cut-node network. The whole network is divided via the factorization path graph. The test results show that the FBBDF method requires less computing time than the classic BBDF method and can increase the speedup of parallel computing to more than 16x with 40 threads in a 24886-node actual power system.

In many China's enterprise‐owned power grids (EPGs), the frequently changing operation state and the poor reliability of some outdated facilities may lead to inaccuracy or even malfunctions of the traditional pre‐decision‐making emergency control systems, which have already caused several blackouts in EPGs. To solve this problem, a real‐time decision‐making emergency control system based on the detailed time‐domain transient stability simulation is designed. Based on the trajectory characteristics of the stability restoration process of EPGs, the stability restoration criteria are proposed to significantly shorten the simulation time needed to judge the elimination of a certain stability problem. Fast decision‐making algorithms for frequency instability, rotor angle instability and branch overload with relatively low computational cost are introduced. Control strategies can be determined within a specific time after faults are detected. The corresponding countermeasures are directly executed when the strategy is obtained. The system configuration is designed to be capable of dealing with complicated and unanticipated faults, such as cascading outages and unexpected grid splitting. The test results of a 31‐node practical EPG show that the proposed algorithms and the system configuration are practical and efficient. This work is a meaningful exploration of the real‐time decision‐making emergency control system.

In the power system parallel transient stability simulation (TSS), the parallel overhead will rise significantly with the increase in concurrency, which will seriously affect the efficiency of parallel computing. However, there are relatively few studies on the parallel overhead in the power system research field at present. In this paper, the geneses of the parallel overhead are introduced from the point of view of computer technology. The composition of the parallel overhead of the bordered block diagonal form (BBDF) method and the nested BBDF (NBBDF) method of the power system parallel TSS is analyzed. The time consumption of computing tasks and the parallel overhead of the BBDF method and the NBBDF method implemented by MPI and OpenMP under different thread scheduling patterns are tested in detail in 2383wp system and two practical power systems (13490-node and 24886-node). On this basis, an efficient implementation of the NBBDF method based on subnet-core mapping and mixed programming of MPI and OpenMP is suggested, which can further reduce the parallel overhead and increase the speedup ratio to more than 17x with 64 cores in a 24886-node system. At last, some rules are summarized based on the test results.