Method and system for detecting faults in a charging infrastructure system for electric vehicles

1. A method for detecting an anomalous operating state in a charging infrastructure system for charging batteries, the method comprising:

for a charging process at a charging station:

obtaining, for the charging process, target characteristics of the charging process;

determining, for the charging process, process parameters based on the obtained target characteristics of the charging process;

performing the charging process based on the determined process parameters;

determining a performance metric for the performed charging process based on the obtained target characteristics;

generating a data set for the performed charging process, wherein the data set comprises meta information of the performed charging process, the determined target characteristics of the performed charging process associated with determined variables for the performed charging process and the determined performance metric for the performed charging process; and

storing the generated data set in a database; and

for a plurality of charging processes:

calculating and storing at least one first set of statistic data based on stored data sets in the database for a first time interval and at least one second set of statistic data based on the stored data sets in the database for a second time interval;

comparing the at least one first set of statistic data with the at least one second set of statistic data for each stored data set to compute a set of difference values for each stored dataset; and

determining for each data set, based on the computed set of difference values, whether the charging infrastructure system operates in an anomalous operating state.

2. The method according to claim 1 , wherein the method further comprises:

generating and outputting at least one of a system alert and failure data in case of determining an anomalous operating state of the charging infrastructure system.

3. The method according to claim 1 , wherein the method further comprises:

outputting the determined target characteristics to a user of the charging station, and accepting a user input that changes or selects at least one of the output target characteristics for the charging process.

4. The method according to claim 1 , wherein

the step of determining a performance metric includes determining a customer-satisfaction indicator as performance metric.

5. The method according to claim 1 , wherein the method further comprises:

determining, whether the charging infrastructure system operates in an anomalous operating state, by applying a method of anomaly detection on the first and second sets of statistic data.

6. The method according to claim 5 , wherein

the method of anomaly detection applied on the first and second set of statistic data comprises applying a predefined rule set or a trained machine learning model.

7. The method according to claim 6 , wherein

the method of anomaly detection applies the trained machine learning model, wherein the trained machine learning model includes a mathematical expression, in particular a decision tree, a random forest algorithm, a neural network, or a deep neural network.

8. The method according to claim 1 , wherein

the target characteristics of the charging process include at least one of a target state-of-charge, target charged energy, target minimum state-of-charge, target range, and target departure time.

9. The method according to claim 1 , wherein

determined variables of the charging process include at least one of date, time, identifier of charging station, location of charging station, type of charging station, type of the battery, model of electric vehicle, weather parameters, and termination type of charging process.

10. The method according to claim 1 , wherein the method further comprises:

computing additional sets of statistic data by filtering the determined variables of the charging process with respect to different features and calculating the additional sets of statistic data for the first time interval and the second time interval from the filtered determined variables; and

storing the computed additional sets of statistic data in the database.

11. The method according to claim 10 , wherein the method further comprises:

determining, whether the charging infrastructure system operates in an anomalous operating state based on the computed set of difference values computed on the stored sets of statistic data including the computed additional sets of statistic data.

12. The method according to claim 1 , wherein

the first time interval is shorter than the second time interval.

13. The method according to claim 12 , wherein

the first time interval is shorter than the second time interval by an order of magnitude.

14. The method according to claim 1 , wherein the method further comprises:

adapting at least one of a first interval length of the first time interval and a second interval length of the second time interval based on the plurality of stored data sets or training data sets for the charging infrastructure system.

15. The method according to claim 14 , wherein

at least one of the first interval length and the second interval length is adapted by using machine learning and optimization methods.

16. A non-transitory computer readable medium storing a computer program with program-code to execute steps of:

for a charging process at a charging station:

obtaining, for the charging process, target characteristics of the charging process;

determining, for the charging process, process parameters based on the obtained target characteristics of the charging process;

performing the charging process based on the determined process parameters;

determining a performance metric for the performed charging process based on the obtained target characteristics;

generating a data set for the performed charging process, wherein the data set comprises meta information of the performed charging process, the determined target characteristics of the performed charging process associated with determined variables for the performed charging process and the determined performance metric for the performed charging process; and

storing the generated data set in a database; and

for a plurality of charging processes:

calculating and storing at least one first set of statistic data based on stored data sets in the database for a first time interval and at least one second set of statistic data based on the stored data sets in the database for a second time interval;

comparing the at least one first set of statistic data with the at least one second set of statistic data for each stored data set to compute a set of difference values for each stored dataset; and

determining for each data set, based on the computed set of difference values, whether the charging infrastructure system operates in an anomalous operating state.

17. A system for detecting an anomalous operating state in a charging infrastructure system for charging a battery, the system comprising:

at least one electric charger configured to charge the battery in a charging process;

an interface configured to obtain target characteristics of the charging process; and

at least one processor configured to:

determine for the charging process, process parameters of the charging process based on the obtained target characteristics;

determine a performance metric for the performed charging process;

generate a data set for the performed charging process, wherein the data set comprises meta information of the performed charging process, determined target characteristics of the performed charging process associated with determined variables for the performed charging process and the determined performance metric for the performed charging process;

store the generated data set in a database;

calculate and store at least one first set of statistic data for a first time interval based on a plurality of stored data sets in the database and at least one second set of statistic data for a second time interval based on the plurality of stored data sets in the database;

compare the at least one first set of statistic data with the at least one second set of statistic data for each stored data set to compute difference values for each stored data set; and

determine, based on the computed difference values for each data set, whether the charging infrastructure system operates in an anomalous operating state.