Neural network based inversion global sliding mode fuzzy control method for active power filter

Abstract

The invention discloses a neural network based inversion global sliding mode fuzzy control method for an active power filter. A neural network based inversion global sliding mode fuzzy control system is composed of an inversion PID global sliding mode controller, a neural network dynamic property estimator and a fuzzy uncertainty approximator. Global sliding mode control can overcome shortcomings of robustness insufficiency of an arrival mode in traditional sliding mode control and accelerates system response so as to enable the system to have robustness in the whole process of response. Due to approximation of switch function items by adopting fuzzy control, switch items of sliding mode control are converted to continuous fuzzy control output, so that chattering in sliding mode control is relieved and comparatively high self-adaptive tracking capability is achieved. Therefore, through combination of inversion PID global sliding mode control and neural network fuzzy control, advantages of inversion PID global sliding mode control and neural network fuzzy control can be fully played, transient performance and robustness of the sliding mode control system can be improved and chattering in variable structure control is reduced.

Description

technical field [0001] The invention relates to a neural network-based active power filter inversion global sliding mode fuzzy control method. Background technique [0002] With the popularization and application of modern power electronic technology, there are more and more various power electronic devices. The working performance and service life of equipment and electrical equipment, and even endanger the safety of the power system. Currently, external filters are mainly used for treatment. Filters are divided into passive filters and active filters. Since the passive filter can only compensate specific harmonics and other defects, the current research on the control of electric energy problems mainly focuses on the active filter. The active filter can track and compensate the harmonics with changing frequency and amplitude. It can not only compensate each harmonic, but also suppress flicker and compensate reactive power. At the same time, the filtering characteristics ...

AI Technical Summary

Problems solved by technology

However, the robustness of adaptive control to external disturbances is very low, and it is easy to make the system unstable.

[0004] This shows that above-mentioned existing active power filter obviously still has inconvenience and defect in use, and urgently needs to be further improved

In order to solve the problems existing in the use of existing active power filters, relevant manufacturers have tried their best to find a solution, but no suitable design has been developed for a long time

Images