In order to further improve the accuracy of ultra-short-term prediction of photovoltaic（PV） power generation

a ultra-short-term PV power prediction model based on improved particle swarm optimization（IPSO） algorithm and enhanced brain emotional neural network（EENN） is proposed according to the inherent nonlinear chaotic characteristics of PV power time series. Firstly

the implicit information features of the PV power sequence are projected to the high-dimensional phase space by nonlinear transformation to form a new data space of attractor trajectory. Then

in order to improve the ultra-short-term prediction ability of the model

by considering nonlinear geometric characteristics of continuous attractor trajectories in space

the EENN model is used to map the relation of data in high dimensional space. And the internal weight and threshold of EENN are iteratively optimized by IPSO to improve the EENN model of data mining and forecasting abilities. Finally

based on the measured PV power generation data

the proposed model is verified effectively by single step prediction. The calculation example shows that the proposed prediction model has better prediction effect than traditional model

and it can effectively improve the ultra-short term prediction accuracy of PV power.