Microgrid transient stability evaluation method based on long short-term memory network

Abstract

A microgrid transient stability evaluation method based on a long short-term memory network comprises the following steps of: constructing a microgrid transient stability evaluation network; training a microgrid transient stability evaluation network to obtain a network weight; and carrying out online evaluation of the transient stability evaluation network of a micro-grid, wherein the online evaluation is implemented by the substeps that: the transient stability evaluation network and the network weight are burnt into a control unit of a virtual synchronous inverter for information interaction; when the micro-grid operates normally, the transient stability evaluation network does not perform evaluation; when the micro-grid detects that a fault occurs, the transient stability evaluation network performs online evaluation, predicts a virtual power angle value of a future T/2 time length according to virtual power angle data of the T/2 time length obtained at the current moment, and judges transient stability by an inverter control unit; and when the predicted virtual power angle value is within a set change range, the transient stability evaluation network enters a non-evaluation state again. According to the microgrid transient stability evaluation method, the debugging work of a neural network can be greatly simplified, and the transient stability of the micro-grid is judged in advance.

Description

technical field [0001] The invention relates to the technical field of power systems, in particular to a method for assessing transient stability of a microgrid based on a long-short-term memory network. Background technique [0002] As an important way for power electronic equipment to be connected to the power system friendly and absorb new energy, microgrid has attracted more and more attention from academia and industry. Different from rotating generators, power electronic equipment has a small inertia and the electrical quantity changes too fast, which is not conducive to the stable operation of the microgrid. To this end, the academic and industrial circles have proposed control strategies for power electronic equipment with virtual inertia characteristics, such as virtual synchronous inverters that simulate the external characteristics of synchronous generators, to improve the stable operation of microgrids. But it also brings to the microgrid the synchronous stabili...

AI Technical Summary

Problems solved by technology

However, the current data-driven and machine learning-based methods have the following defects: (1) rely heavily on prior knowledge to select input features and adjust network parameters

Existing methods need to pre-screen representative features based on manual experience, and determine the hyperparameters of the network through multiple experiments, which is inefficient, highly random, and cumbersome to operate

(2) The required characteristic quantity needs to install a large number of sensors and communication equipment in the network, such as collecting the voltage amplitude phase, active and reactive power values ​​of each bus, and sending them to the system control center through communication lines to generate dispatching control after calculation The command is sent to the power generation unit, the construction cost is high and the real-time performance is poor

The need for fast and accurate transient stability assessment of microgrids has not been well met

Images