Construction method and application of dominant instability mode recognition model of power system
A power system and construction method technology, applied in the field of power system dominant instability pattern recognition model construction, can solve the problems of increasing labeling sample costs, expensive labeling costs, and increasing costs, so as to improve recognition accuracy, reduce costs, and reduce dependencies. Effect
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
Method used
Image
Examples
Embodiment 1
[0051] The present invention adopts a method based on semi-supervised learning to build a power system dominant instability pattern recognition model. The method can construct a complex mapping relationship from original data to dominant instability patterns, and a well-trained model can quickly and accurately judge the dominant instability pattern of the system. The stability mode can effectively distinguish the two types of instability, voltage instability and power angle instability, and provide a basis for the subsequent formulation of control decision tables in the analysis of simulation data.
[0052] It should be noted that power angle stability can be subdivided into large disturbance power angle stability and small disturbance power angle stability according to the size of the disturbance. Similarly, voltage instability can also be divided into small disturbance voltage stability (static voltage stability) and large disturbance power angle stability. The disturbance vo...
Embodiment 2
[0098] Further analysis found that the application of disturbance in the model is random, which will cause the model to not make good use of the information of unlabeled samples, and the model should be able to adapt to small input changes, and if the model is overfitting, then there will be small A perturbation of the input causes a large change in the output. Therefore, during the training process, adding disturbances to the input features of unlabeled samples makes the model have the ability to resist disturbances, and at the same time strengthens the decision boundary, making the sample distribution near the decision boundary sparse, that is, strengthening the classification ability of the model. The virtual adversarial training method introduced in this embodiment is to find a direction that causes the largest change in the model output due to small disturbances, which is called the maximum disturbance direction. Applying a certain disturbance in this direction is better f...
Embodiment 3
[0111]A machine-readable storage medium. The machine-readable storage medium stores machine-executable instructions. When the machine-executable instructions are called and executed by a processor, the machine-executable instructions prompt the processor to implement the dominant failure in Embodiment 1 or 2. A construction method for stable pattern recognition and / or a dominant unstable pattern recognition method. Related technical features are the same as those in Embodiment 1 or 2.
PUM
Login to View More Abstract
Description
Claims
Application Information
Login to View More 


