Heart Rate Regulation with Different Heart rate Reference Profiles for Electric Bicycle Riders

In this paper, we propose a new nonlinear control strategy for electric bicycles that adjusts the motor assistance automatically to maintain a desired heart rate level of the cyclist. The control algorithm uses the current heart rate, torque and cadence of the rider as inputs, and adjusts the assistance of the motor adaptively. The controller consists of feedforward as well as feedback components: Whereas the heart rate is the feedback component, the system's feedforward component is the rider's estimated torque output at the desired heart rate level. Environmental changes like slope or headwind are modeled as disturbances that the controller has to reject. We conduct test rides with one subject at different intensity levels to develop and validate a nonlinear heart rate response model of the bicycle rider. The model uses the power output of the cyclist as an input and takes immediate and gradual responses of the heart rate into account. We use a commercial electric bicycle with a front wheel hub motor to test the controller design. The bicycle is modified to measure torque output and pedalling frequency of the rider and to control the motor torque. The heart rate is measured by a chest strap. Data processing and controller design were performed on a laptop. We present simulation as well as experimental results for different reference heart rate profiles to verify the performance of the controller.