Electric vehicles as flexible loads: Algorithms to optimize aggregate behavior

This paper considers load shifting for electric vehicles (EVs) to reduce the peak value of the total power consumption. Large numbers of EV charging requests are classified into relatively small number of load types so that the computational effort remains unchanged as number of loads increases. A twolayer optimization scheme is proposed to deal with tasks that may demand sequentially varying power levels, which further reduces the computational burden. These ideas are potentially useful in harnessing flexibility in electric loads.