We currently only have observations for when EVs are being driven or are charging. This means there is a lot of 'missing' data when it comes to calculating 'population' average kW charging demand etc. We need to impute missing 1 minute observations (given them 0 kW charging) and re-calculate sample means. Obviously we can't impute state_of_charge - although we cold assume some decay curve between the last and next real observation?
We currently only have observations for when EVs are being driven or are charging. This means there is a lot of 'missing' data when it comes to calculating 'population' average kW charging demand etc. We need to impute missing 1 minute observations (given them 0 kW charging) and re-calculate sample means. Obviously we can't impute state_of_charge - although we cold assume some decay curve between the last and next real observation?