Federated Learning for Automotive

Use case 1: EV battery range prediction model update using Federated Learning

Model inputs

1. State of Charge (SOC) (%)
2. Battery Voltage (V)
3. Battery Temperature (C)
4. Current Vehicle Speed (km/h or mph)
5. Average Speed over the last 5 minutes
6. Current draw (A) (can be inferred or included directly)
7. State of Health (SOH) (%) (optional, useful for long-term accuracy)

Model output

• Range (distance in km/miles)

Simulated data for training

Input data generated as normalized (zero mean, unit variance) tensor

EVdata = torch.randn(NUM_SAMPLES, INPUT_DIM) # 100 samples, 10 features each

Output data is range in between (MIN_RANGE, MAX_RANGE)

MIN_RANGE = 0
MAX_RANGE = 100
EVrange = torch.FloatTensor(NUM_SAMPLES,1 ).uniform_(MIN_RANGE, MAX_RANGE) 

Model training (without Federated Learning)

python3 model_trainer.py

It is expected that model weights will not converge during training as randomized data is used.

Model training using flower FL library with pytorch model

Following steps are required to run the demo for EV range model training using Federated Learning

Code changes

1. Use flwr new to create a template director
2. Update model architecture and data source for the use case in ./ev_range_pred/task.py
   • Replace the CIFAR dataset with random data
   • Replace flower_datasets with torch DataLoader

Above steps are completed in ./ev_range_pred directory.

Running the demo on localhost

Run below commands in different terminals as given in deployment tutorial.

Terminal 1 - Run superlink

flower-superlink --insecure

Terminal 2 - Run supernode

flower-supernode \
     --insecure \
    --superlink 127.0.0.1:9092 \
    --clientappio-api-address 127.0.0.1:9094 \
    --node-config "partition-id=0 num-partitions=2"

Terminal 3 - Run client (to be ported to run on the vehicle edge ECU)

flower-superexec \
    --insecure \
    --plugin-type clientapp \
    --appio-api-address zupernode-1:9094