Input: training set Z = ( x k , y k ) ; learning rate

Algorithm process:


1. Randomly initializing weights and thresholds within the range of (0, 1);

2. Repeat;

3. For all ( x k , y k ) Z do

4. Calculating the output of the current sample from the current parameters and the training set;

5. Calculating a gradient term of the output layer neuron from the mean square error and the output of the current sample;

6. Calculating the gradient term of the hidden layer neuron according to the error;

7. Update weights and thresholds;

8. End for

9. Until reaches the stop condition;

Output: charging station planning model, weighted and threshold-determined multilayer feedforward neural network