Power system transient stability assessment based on the multiple paralleled convolutional neural network and gated recurrent unit

Protection and Control of Modern Power Systems

Table 9 The parameter setting of MP CNN + GRU on the IEEE 39-bus system

CNN branch	Quantity/function
Time sequence dimension	34
Input feature dimension	170
Convolution layer	3/ReLU
Convolution kernel size	5 × 5
Convolution kernel Quantity	64
Pooling layer	3
Pooling size	3 × 3
Pooling method	Max-pooling
Full connection layer units	128
Dropout rate	0.2

Full connection layer	Quantity/function
Layers	3/ReLU
Units on the first layer	128
Units on the second layer	64
Units on the last layer	1
Loss function	Improved FL
Optimizer	Adam
Initial learning rate	4 × 10^–3
The first decay rate (β₁)	0.8
The second decay rate (β₂)	0.5
Epsilon (ε)	2 × 10^–9
Total parameters of the model	3078