The virtual synchronous generator (VSG) improves the robustness of the inverter-interfaced distributed generator (IIDG) against instability by introducing a virtual inertia. However, the transient response of the active power and the angular frequency conflict with each other for the IIDG with fixed inertia control. It is necessary to adopt adaptive control to improve overall performances of power and frequency as the operating condition changes. This paper analyzes the impact of the inertia on power and angular frequency. A dual-adaptivity inertia control strategy is proposed to offer a responsive and stable frequency support and also achieve the balance between power regulation and frequency regulation according to different operating conditions. The principle of parameter design is given to obtain the range of adaptivity. Quantitative assessment considering the cumulative effect of the output deviation and its duration is also presented to evaluate the proposed strategy intuitively. The strategy is further verified based on PSCAD/EMTDC and a hardware-in-loop experiment platform based on RTDS. Results confirm that the proposed strategy not only achieves rapid frequency response with slight dynamic deviations under disturbances but also strikes a balance between the frequency and power and leads to improved overall control.

Constrained optimization problems have appeared in a wide variety of challenging real-world problems, where constraints often capture the physics of the underlying system. Classic methods for solving these problems relied on iterative algorithms that explored the feasible domain in the search for the best solution. These iterative methods often became the computational bottleneck in decision-making and adversely impacted time-sensitive applications. Recently, neural approximators have shown promise as a replacement for the iterative solvers that can output the optimal solution in a single feed-forward providing rapid solutions to optimization problems. However, enforcing constraints through neural networks remains an open challenge. In this paper, we have developed a neural approximator that maps the inputs to an optimization problem with hard linear constraints to a feasible solution that is nearly optimal. Our proposed approach consists of five main steps: 1) reducing the original problem to optimization on a set of independent variables, 2) finding a gauge function that maps the ℓ _∞ -norm unit ball to the feasible set of the reduced problem, 3) learning a neural approximator that maps the optimization’s inputs to a virtual optimal point in the ℓ _∞ -norm unit ball, and 4) gauge mapping to project the virtual optimal point in the ℓ _∞ -norm unit ball onto the feasible space, then 5) finding the values of the dependent variables from the independent variable to recover the solution to the original problem. We can guarantee hard feasibility through this sequence of steps. Unlike the current learning-assisted solutions, our method is free of parameter-tuning (compared to penalty-based methods) and removes iterations altogether. We have demonstrated the performance of our proposed method in quadratic programming in the context of optimal power dispatch (critical to the resiliency of our electric grid) and constrained non-convex optimization in the context of image registration problems. Our results have supported our theoretical findings and demonstrate superior performance in terms of computational time, optimality, and the feasibility of the solution compared to existing approaches.

Inverter-interfaced distributed generators (IIDGs) have been widely applied due to their control flexibility. The stability problems of IIDGs under large signal disturbances, such as large load variations and feeder faults, will cause serious impacts on the system. The virtual synchronous generator (VSG) control is an effective scheme for IIDGs to increase transient stability. However, the existing linearized stability models of IIDGs are limited to small disturbances. Hence, this paper proposes a Lyapunov approach based on non-linearized models to assess the large signal stability of VSG-IIDG. The electrostatic machine model is introduced to establish the equivalent nonlinear model. On the basis of Popov’s theory, a Lyapunov function is derived to calculate the transient stability domain. The stability mechanism is revealed by depicting the stability domain using the locus of the angle and the angular frequency. Large signal stability of the VSG-IIDG is quantified according to the boundary of the stability domain. Effects and sensitivity analysis of the key parameters including the cable impedance, the load power, and the virtual inertia on the stability of the VSG-IIDG are also presented. The simulations are performed in PSCAD/EMTDC and the results demonstrate the proposed large signal stability assessment method.

Leveraging machine learning to facilitate the optimization process is an emerging field that promises to bypass the fundamental computational bottleneck caused by classic iterative solvers in critical applications requiring near-real-time optimization. Most existing approaches focus on learning data-driven optimizers that lead to fewer iterations in solving an optimization. In this paper, we take a different approach and propose to replace the iterative solvers altogether with a trainable parametric set function, that outputs the optimal arguments/parameters of an optimization problem in a single feed-forward. We denote our method as Learning to Optimize the Optimization Process ( LOOP ). We show the feasibility of learning such parametric (set) functions to solve various classic optimization problems, including linear/nonlinear regression, principal component analysis, transport-based coreset, and quadratic programming in supply management applications. In addition, we propose two alternative approaches for learning such parametric functions, with and without a solver in the LOOP . Finally, through various numerical experiments, we show that the trained solvers could be orders of magnitude faster than the classic iterative solvers while providing near-optimal solutions.

Wireless Mesh Network is a type of wireless networks in which demands of bandwidth for users has mobility. The backbone node placement of wireless mesh networks in a dynamic scenario is investigated, and the TSDPSO algorithm is used to adapt the dynamic environment, which updates node deployment location to adapt to changes in demand if it detects environmental changes at the beginning of the cycle time. In order to meet the demands of bandwidth for users and network connectivity, particle swarm optimization algorithm is employed to select the gateway location, then nodes to the backbone network is added constantly until all requirement is covered. The experimental results show that algorithm could get effective solution in dynamic environment.