Abstracts:This letter proposes a novel quality-diversity learning (QDL) method for multi-alternatives unit commitment (UC) optimization. Existing UC methods focus solely on finding a single global optimum, neglecting insights from alternative solutions with competitive performance. In contrast, QDL maintains a multi-cell agent archive populated with multiple high-performing UC policies, each sharing the same objective while evolving to explore distinct behavioral regions, enabling simultaneous optimization of solution quality and diversity. The resulting diverse solutions catering to various dispatch preferences not only enhance operational preparedness, but also allow rapid retrieval of alternatives if feasibility tests fail. Case studies on several standard test systems confirm the effectiveness of the method.

Abstracts:Binary variables in unit commitment (UC) problems invalidate gradient-based directional information, often causing computational bottlenecks. Existing binary algorithms ignore a tendency of these variables towards 0 or 1, which affects efficiency. To improve performance, this letter formalizes this tendency as preference and leverages it to guide the optimization process. A solution-set-based global optimization algorithm is introduced to handle to non-convexity arising from complex operational constraints. The results show that the guided algorithm has improved efficiency and robust global convergence ability.

Abstracts:This letter proposes a novel online spatiotemporal ensemble learning framework that can rapidly adapt to load pattern changes caused by abnormal events. Unlike existing online learning approaches that focus solely on temporal dependencies, the proposed method also exploits spatial correlations across different regions to achieve fast convergence. An online complementary learning network that can instantly adapt to new patterns while recalling similar historical knowledge is first built as the basic forecast expert to extract spatial and temporal features. The two information streams are then combined using an online convex programming framework, which is further solved by exponentiated gradient descent and reinforcement learning methods. Experiments on real-world electricity load datasets from the COVID-19 period demonstrate the proposed method's effectiveness.

Abstracts:This letter presents a GPU-accelerated implementation of the Current Injection Method for Power Flow (CIM-PF) using a Sparse Inverse Factor Matrix-Vector Multiplication (SIF-MVM) linear solver. Unlike conventional LU-based solvers that are inefficient on GPU architectures due to their sequential nature, the proposed method leverages GPU parallelism through precomputed sparse inverse LU factors and parallel matrix-vector operations. The approach is evaluated on large-scale distribution systems with up to 168,435 nodes. For the largest test case, the solver achieves over 4× acceleration in single-scenario power flow computations and up to 8× acceleration in batch simulations involving 1,000 distinct current injection scenarios, compared to CPU-based methods. These results demonstrate the method’s suitability for high-throughput applications such as time-series analysis, probabilistic studies, and large-scale planning simulations.

Abstracts:This letter proposes a novel power flow (PF) linearization method for accurately enforcing network constraints in optimal dispatch problems. Unlike conventional linearization methods that focus on reducing PF solution errors, the proposed method aims to enhance the decision feasibility of network-constrained dispatch problems modeled with linear PF equations. A data-driven framework based on hybrid regression and classification is developed to determine coefficients of the linear PF equation. This problem is equivalent to minimizing a weighted sum of the root-mean-square error and hinge loss, which compels the linear PF model to enforce network constraints accurately. Simulations with various system scales verify that the proposed PF linearization method outperforms existing ones in terms of decision feasibility and optimality.

Abstracts:Participation factors (PFs) quantify the interaction between system modes and state variables, and they play a crucial role in various applications such as modal analysis, model reduction, and control design. With increasing system complexity, especially due to power electronic devices and renewable integration, the need for scalable and high-order nonlinear PF (NPF) computation has become more critical. This paper presents an efficient tensor-based method for calculating NPFs up to an arbitrary order. Traditional computation of PFs directly from normal form theory is computationally expensive—even for second-order PFs—and becomes infeasible for higher orders due to memory constraints. To address this, a tensor contraction–based approach is introduced that enables the calculation of high-order PFs using a batching strategy. The batch sizes are dynamically determined based on the available computational resources, allowing scalable and memory-efficient computation.

Abstracts:This letter proposes an electro-thermal aging coordination (ETAC) concept to enhance the system’s operational performance by exploiting the current-carrying capacity potential during the thermal aging process in system dispatch decisions. First, the equivalent operating duration of the overhead conductor at its long-term permissible operating temperature (LPOT), which yields the same degree of thermal aging as that caused by the operating temperature at the dispatch period, is formulated based on the Morgan tensile strength loss model. Then, the thermal aging constraints are formulated, and the ETAC model is established by introducing the thermal aging constraints into the electrothermal coordination model, replacing the conventional LPOT constraint. Finally, case studies on a six-node system are conducted to validate the effectiveness of the ETAC concept.

Abstracts:The supplementary damping controller (SDC) at the doubly-fed induction generator (DFIG) can suppress angular oscillations among synchronous generators (SGs). The existing studies see wind farm as equivalent DFIG, ignoring interaction among SDCs and DFIGs, not suitable to coordinate SDCs in heterogeneous wind farm (HWF) with only some DFIGs having SDCs. This paper proposes a dynamic coordination model to SDCs to suppress the oscillations considering the interaction due to HWF. First, the interactive currents of DFIGs are decomposed to components related to stator voltages and SDCs. Non-square interactive matrix caused by SDCs is derived to quantify impact of SDCs on non-identical DFIGs. Then by using the non-square internal model control (IMC), a two-stage model is proposed to coordinate the gains of SDCs with the interactive matrix to avoid uniform setting. Finally, to further improve suppression effect, interactive sensitivity with respect to SDC’s gain is derived, based on which the analytical expression of the pseudo inverse of interactive matrix is newly derived by equating it to numerical solution derived by Moore-Penrose inverse and equating the sensitivity of the former to that of the latter, to avoid uncertain solutions of coordination model. Simulation results verify control effect of the proposed model.

Abstracts:This letter introduces a novel quantum Newtonian power flow (QNPF) algorithm. The QNPF features a more general quantum circuit that can process non-Hermitian matrices with fewer required qubits and eliminates the need for iterative optimization of gate parameters. Our contributions include: 1) Developing a quantum state-based Newton's power flow framework to enhance accuracy, convergence, and versatility; 2) Integrating quantum singular value transformation to efficiently solve each iteration of QNPF with scalable quantum circuits; and 3) Devising a block-rescaling technique to ensure computational accuracy in ill-conditioned cases. Test results validate the accuracy, scalability and numerical stability of QNPF, underscoring its potential to advance quantum power flow computation.

Abstracts:Machine learning (ML)-based transient stability assessment (TSA) provides extraordinary accuracy performance while limited by potential misjudgment risks. To address this issue, this letter originally develops a generic scene-dependent credibility evaluation (SCE) framework. The variance upper bound of ML model prediction error is inferred using an improved localized generalization error estimation (ILGEE) method, and the probability density of system stability is furtherly described as a Gaussian distribution incorporating Neumann boundary condition. Then the scene-dependent credibility index (SCI) is ultimately derived and defined as the information entropy implying the uncertainty of TSA results. Case studies verify the validity of the SCE framework and demonstrate the promising 100% accurate TSA performance with critical proposed SCI as 0.93.