Summary
The authors conceive a notion of reactive control that supersedes the time-triggered approach by leveraging the characteristics of existing control logic and of the hardware it runs on. Under reactive control, control decisions are taken only upon recognizing the need to.
Background
Current Drone Architecture
Ground-control station (GCS) configure mission parameters, such as coordinates to cover and actions to take at each waypoint Autopilot Software implements low-level control (PID) autonomously driving the drone to the next desired coordinate
Low-level Control
- Most PID controllers on drones are tuned so that only the Proportional component bear an influence
- PID controllers run in time-triggered fashion: every T time units, sensors are probed, control decisions are computed, and commands are sent to the actuators.
- Authors observe that control output doesn’t change often
Reactive Control
Authors argue for reactive control.
Rather than periodically triggering the control logic, we constantly monitor the navigation sensors and run the control logic only upon recognizing the need to. As a result, reactive control dynamically adapts the control rate.
Benefits of Reactive control
- enables more timely and adaptive control decisions
- spares unnecessary processing, improving the utilization of the hardware
- lessens the need to over-provision control rates to handle extreme situations
Challenges
- What is a “significant” change in the sensor input depends on several factors, including the accuracy of sensor hardware, the physical characteristics of the drone, the control logic, and the granularity of actuator output
- An indication for running the control logic may originate from different sensors, at different rates, and asynchronously with respect to each other
- Reactive control must run on resource-constrained embedded hardware
Solutions
- Use a logistic regression based On-line approach to determine how large changes in a sensor constitutes “significant”. A small boot time at the beginning of runtime is needed to learn how large of sensor reading changes result in control output changes for each sensor.
- Sample sensors at the highest frequency available and use hyperperiod to group correlated sensor changes together. A fail-safe mode is employed to run the control logic anyhow every 0.1 second.
- A specially designed reactive programming model is implemented to avoid “callback hell”
Result
