With the development of the industry, the physical model of controlled object tends to be complicated and unknown. It is particularly important to estimate the state variables of a nonlinear system when the model is unknown. This paper proposes a state estimation method based on adaptive fusion of multiple kernel functions to improve the accuracy of system state estimation. First, a dynamic neural network is used to build the system state model, where the kernel function node is constructed by a weighted linear combination of multiple local kernel functions and global kernel functions. Then, the state of the system and the weight of the kernel functions are put together to form an augmented state vector, which can be estimated in real time by using high-degree cubature Kalman filter. The high-degree cubature Kalman filter performs adaptive fusion of the kernel function weights according to specific samples, which makes the neural network function approximate the real system model, and the state estimate follows the real value. Finally, the simulation results verify the feasibility and effectiveness of the proposed algorithm.
Abstract:With some of the intermittent new energy and large nonlinear loads, grid voltage unbalance, harmonics, and frequency deviation are increasing year by year. The voltage source converter (VSC) is seriously affected by the various unexpected factors, and the presence of grid impedance makes the situation worse. In order to make the VSC track the nonideal grid quickly and accurately, this paper proposes a frequency-adaptive grid-virtual-flux synchronization by multiple second-order generalized integrators (MSOGI-GVFS). Key expressions of the MSOGI-GVFS and its frequency response characteristics are described in this paper. A second-order generalized integrator configured as a quadrature signal generator generates a specific-frequency virtual flux. A harmonics decoupling network achieves fundamental and harmonic components of the virtual flux. The positive-and negative-sequence components are separated by multiple positive-and negative-sequence calculators. A frequency-locked loop is used to track the grid angular frequency. Finally, after compensating the voltage and the flux on the grid impedance and the filtering inductor, it accurately achieves the estimation of the grid virtual flux in the highly polluted grid environment. This method may reduce the voltage sensors, eliminate the influence of grid impedance, and track the grid frequency quickly, which contributes to the stability of the VSC. The good performance of MSOGI-GVFS is verified by simulation and experimental results.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.