Wireless sensor nodes to be powered continuously for long life, good performance and high accuracy in the unattended location of the world. Battery powered sensor nodes are not able to provide continuous power. solar power is the scavenge source, but due to the sun cycle variation, it is important to predict the solar current energy produced according to time. In this paper we use automatic battery charger -discharge system that uses the meta-modeling correlation for prediction of solar energy and processor is designed for powering wireless sensor node. The new method uses simple kriging model and swarms optimization that improves precharge time by 61.54% while speeding up the process by390*. The simulation results are shown by MATLAB, labview, and design of the processor with EDA tools.