IV. CONCLUSION
In this paper, the energy consumption of public buildings is
analyzed by using the self-correlation of the time series, thus
the dimension of the input variables is determined. Then we use
artificial fish-swarm algorithm to optimize the weights and
thresholds of BP neural network. The results show that this
method improves the convergence speed of the network, and
overcomes the defects of the BP neural network which is easy
to fall into local extremum. At the same time, this method has
some adaptability to the local optimization, and it does not
exist the problem of complex encoding. It also has strong
robustness. Therefore, it provides a powerful guarantee for the
energy saving control of public buildings.
In order to improve the accuracy of prediction, the number
of training samples and iterations can be increased. How to
predict the standard energy consumption based on the
prediction results of artificial fish-swarm algorithm, and when
to update the weights and thresholds of BP network, will be a
future research direction