Dynamic security assessment method based on principal component analysis and convolutional neural network

Abstract

The invention relates to a dynamic security assessment method based on principal component analysis and a convolutional neural network, which specifically comprises the following steps of: 1, obtaining a system operation data sample, constructing a dynamic security index, and forming a corresponding initial sample set; 2, generating an efficient sample set; 3, updating the efficient sample set, and completing the updating of an evaluation model; and 4, based on the real-time operation data of a power system, completing evaluation of the real-time dynamic security state of the power system by using the continuously updated dynamic security evaluation model to obtain a final online dynamic security evaluation result. According to the online dynamic security model of the power system, rapid,efficient and accurate prediction and evaluation can be provided for the power system, system maintenance and safety measure prevention work of power personnel are facilitated, the safety and stability of operation of the power system are improved, and the reliability of power supply is improved.

Description

technical field [0001] The invention relates to the field of dynamic security assessment of power systems, in particular to a dynamic security assessment method based on principal component analysis and convolutional neural network. Background technique [0002] With the ever-changing reorganization of the power system and the large-scale development of wide-area interconnection of the system and the increasing penetration of renewable energy (especially wind power), this makes it difficult to predict the operating status of the power system. When the power system is forming At the same time as wide-area interconnection, the scope affected by major disturbance accidents is also expanded, and the risk of major blackouts is also increased. The scale of the power system continues to expand, making the operating environment of the power system more complex and changeable, and potential emergencies increase the dynamic unsafe risk of the power grid. Therefore, the safe operation...

AI Technical Summary

Problems solved by technology

[0004] (1) The traditional mechanism analysis method mainly relies on off-line calculation, and the time-domain simulation method relies on the accuracy of modeling, and the calculation amount is huge and the calculation time is long; due to the continuous expansion of the scale of modern power systems, this method is difficult for large-scale Sample analysis, it is difficult to meet the requirements of real-time dynamic security assessment for calculation speed, unable to provide stability margin information and many other defects;

[0005] (2) When the traditional data-driven method is applied to the dynamic security assessment of the power system, it often does not consider various factors that may exist in the actual power grid operation, such as the efficiency of the training sample set, does not evaluate the assessment results, and cannot evaluate the dynamic security information. provide visualization

At the same time, the training time is too long and it is difficult to apply to large-scale data

In the actual operation of the power system, some unexpected situations often appear, and the traditional dynamic security assessment model is difficult to evaluate these unexpected situations

[0006] As mentioned above, traditional methods have been difficult to meet the needs of dynamic security assessment of modern power systems, and there is an urgent need for a real-time assessment method that can meet high adaptability and high precision.

Images