Abstract
Electric Vehicle (EV) is playing a significant role in the distributionenergy management systems since the power consumption level of the EVs is muchhigher than the other regular home appliances. The randomness of the EV driverbehaviors make the optimal charging or discharging scheduling even moredifficult due to the uncertain charging session parameters. To minimize theimpact of behavioral uncertainties, it is critical to develop effective methodsto predict EV load for smart EV energy management. Using the EV smart charginginfrastructures on UCLA campus and city of Santa Monica as testbeds, we havecollected real-world datasets of EV charging behaviors, based on which weproposed an EV user modeling technique which combines statistical analysis andmachine learning approaches. Specifically, unsupervised clustering algorithm,and multilayer perceptron are applied to historical charging record to make theday-ahead EV parking and load prediction. Experimental results withcross-validation show that our model can achieve good performance for chargingcontrol scheduling and online EV load forecasting.