Forecasting charging time of electric vehicles

A system configured to forecast charging time of electric vehicles, the system comprising a plurality of charging stations communicatively coupled to a computer device or computer system over a communication network, wherein the plurality of charging station are configured to obtain and send charging event data to the computer device or computer system, and wherein the computer device or computer system is configured:

• to receive charging event data from the plurality of charging stations over a data communication network, the charging event data received from a charging station of the plurality of charging stations comprises at least one of operating conditions of the charging station, configuration of the charging station, and temperature of components of the charging station,
• to filter received charging event data to discard at least one piece of the charging event data or disable use of at least one piece of the charging event data,
• to process the remaining charging event data after said filtering to generate a data collection comprising sample data,
• to obtain from the data collection a data cluster representing sample data associated with a selected electric vehicle cluster,
• to generate a charging time forecasting model for the selected electric vehicle cluster on basis of the obtained data cluster, wherein the charging time forecasting model is generated by a machine learning method performed by the computer device or computer system, the method comprising:
• training a plurality of charging time forecasting models based on a first portion of the data cluster,
• selecting the best performing charging time forecasting model among the plurality of trained charging time forecasting models based on an evaluation performed on basis of a second portion of the data cluster, and
• testing the selected charging time forecasting model based on a third portion of the data cluster,
• to repeat said obtaining the data cluster and generating the charging time forecasting model for a plurality of different data clusters and for generating a charging time forecasting model for each of a plurality of electric vehicle clusters, and
• to provide the plurality of generated charging time forecasting models to the plurality of charging stations,
• and wherein the charging station is configured:
• upon detecting initiation of a new charging event of an electric vehicle, to identify the electric vehicle or the vehicle cluster of the electric vehicle, and to determine the selected vehicle cluster of the electric vehicle,
• to select the charging time forecasting model to be applied on the new charging event based on the selected vehicle cluster, and
• to apply the selected charging time forecasting model on the charging event of the electric vehicle comprised in the selected electric vehicle cluster.