Simulation and Real‐Time Implementation for Power Quality Enhancement Systems Using Adaptive Neural Network Strategy

Neural Network-Assisted Model Predictive Control of a Modular Multilevel Converter

This paper describes an advanced algorithm for current control in a modular multilevel power converter based on deadbeat model predictive control, which is further enhanced by applying an artificial neural network.

Finite-time synchronization of neuronal networks based on double-power sliding mode control

Finite-time synchronization control of discrete neuronal networks with piecewise nonlinear characteristics is a critical and challenging issue in complex network dynamic research. This paper investigates the finite-time synchronization problem of discrete neuronal networks constructed by the Rulkov map, and proposes a novel double-power finite-time sliding mode control (SMC) strategy. Different from single-power SMC schemes with inherent performance trade-offs, the designed control law integrates super-power and sub-power terms organically. The super-power term accelerates the convergence of sliding variables when system errors are large, while the sub-power term guarantees precise finite-time convergence of system states near the equilibrium point, effectively solving the contradiction between transient response speed and steady-state synchronization accuracy in traditional SMC. A terminal sliding surface matching the discrete network dynamic characteristics is established for each neuron node, and the control law is derived based on system error dynamics with a coupling compensation term added to adapt to the dynamic adjustment of large-scale network coupling relationships. Based on Lyapunov stability theory and discrete finite-time stability theorems, the finite-time convergence of both the reaching phase and sliding phase of the control system is strictly proved, and the explicit upper bound of network synchronization time is deduced, which quantitatively reveals the correlation between synchronization efficiency and system initial states as well as control parameters. Numerical simulations on an 80-node randomly connected network demonstrate the effectiveness of the proposed scheme in achieving rapid error convergence and accurate synchronization, and a systematic parameter study reveals the dominant influence of the low-power gain and the high-power exponent on the synchronization time.

Adaptive control strategy for VSG virtual inertia based on RBF neural network

This paper presents a hierarchical control scheme for virtual synchronous generators (VSGs) to mitigate power oscillations and steady-state errors induced by power command variations and grid frequency fluctuations. Unlike conventional adaptive methods that operate at the parameter level, the proposed architecture addresses the inherent structural coupling between damping and droop characteristics from a control-structure perspective. It integrates two complementary modules: a lower-level differential feedforward compensation (DFC) loop that restructures the active-power control channel to achieve structural decoupling between damping and droop, thereby eliminating steady-state deviations under frequency disturbances; and an upper-level RBF neural network that uses frequency deviation and its rate of change to dynamically regulate the virtual inertia J-the sole optimized parameter-reducing the control dimension while enhancing transient response and frequency stability. The uniform ultimate boundedness of the closed-loop system is proved via Lyapunov theory. The strategy is validated through four typical disturbance scenarios via Starsim hardware-in-the-loop experiments, along with comprehensive sensitivity analyses covering system parameters (short-circuit ratio from 1 to 5) and controller parameters (DFC coefficient and filter time constant deviations), as well as RBF initialization uncertainties. Results demonstrate that the DFC-RBF-VSG achieves zero overshoot, settling time within 0.1 s, zero steady-state error under frequency disturbances, total harmonic distortion (THD) reduced to 2.44%–2.95%, and peak inrush current limited within 300 A, consistently outperforming conventional Fixed-VSG, linear-adaptive VSG (Linear-VSG), fuzzy-based adaptive VSG (Fuzzy-VSG), and feedback-compensated VSG (FBC-VSG). The method strikes a favorable balance among dynamic responsiveness, steady-state precision, and robustness, offering an effective solution for high-penetration renewable energy systems.

A novel robust and fast control algorithm for wireless power transfer chargers based on the sliding-mode theory and fixed-time reaching process

ADAPTIVE NEURAL NETWORK-ASSISTED LINEAR QUADRATIC REGULATOR CONTROL WITH FULL-ORDER EXTENDED STATE OBSERVER FOR ACTIVE SUSPENSION SYSTEMS

This paper proposes integrating a Linear Quadratic Regulator (LQR) control scheme with an adaptive Neural Network (NN) updating law to improve ride comfort performance. The dynamic states of the suspension system are effectively estimated using a Full-Order Extended State Observer (FOESO) rather than through direct sensor measurements, thereby reducing implementation costs and minimizing the influence of sensor noise. Numerical simulations are conducted under two different road profile scenarios based on ISO standards, while accounting for lumped uncertainties in the system dynamic modeling. The results show that the Root Mean Square (RMS) body displacement is reduced to 4.854 mm in the first case with ISO C-class road disturbance, and the RMS body acceleration decreases to 0.220 m/s² in the second case with ISO D-class road disturbance; both are significantly lower than those achieved by conventional controllers. Furthermore, the dynamic states are estimated with high accuracy, confirming the effectiveness of the proposed control strategy in suspension system regulation.

Neural network-based energy-input shaping combined control for anti-sway and positioning of crane systems under non-zero initial conditions

A control method based on the combination of energy control, input shaping, and neural network is proposed, in order to solve the problem of anti-swing and positioning of portal bridge cranes under non-zero initial conditions.