Hardware-in-the-Loop for Characterization of Embedded State Estimation for Flying Microrobots

Abstract

Autonomous flapping-wing micro-aerial vehicles (FWMAV) have a host of potential applications such as environmental monitoring, artificial pollination, and search and rescue operations. One of the challenges for achieving these applications is the implementation of an onboard sensor suite due to the small size and limited payload capacity of FWMAVs. The current solution for accurate state estimation is the use of offboard motion capture cameras, thus restricting vehicle operation to a special flight arena. In addition, the small payload capacity and highly non-linear oscillating dynamics of FWMAVs makes state estimation using onboard sensors challenging due to limited compute power and sensor noise. In this paper, we develop a novel hardware-in-the-loop (HWIL) testing pipeline that recreates flight trajectories of the Harvard RoboBee, a 100mg FWMAV. We apply this testing pipeline to evaluate a potential suite of sensors for robust altitude and attitude estimation by implementing and characterizing a Complimentary Extended Kalman Filter. The HWIL system includes a mechanical noise generator, such that both trajectories and oscillatinos can be emulated and evaluated. Our onboard sensing package works towards the future goal of enabling fully autonomous control for micro-aerial vehicles.

Figures (7)

• Figure 1: (a) The Harvard RoboBee, the target FWMAV, photographed with the IMU and ToF components utilized in the hardware experiments. (b) A RoboBee flight trajectory that the robot arm reproduces in (c), while carrying the IMU and ToF sensors to collect data. (d) The sensor package, mounted to the distal end of the arm, which contains the sensor components shown in (a) for state estimation.
• Figure 2: Block diagram illustrating the overall pipeline for the evaluation of state estimation algorithms. Based on RoboBee flight trajectory data, which is captured with a suite of Vicon motion capture cameras, a candidate onboard sensor package is evaluated by reproducing the same flight trajectories in hardware (or simulation). Subsequently, the sensor data is fed into the state estimation algorithm which utilizes a complementary filter and EKF to estimate the orientation and altitude of the vehicle.
• Figure 3: The Harvard RoboBee; the state described via the roll ($\phi$), pitch ($\theta$), yaw ($\psi$), and altitude ($\zeta$) in the world coordinate frame $\{W\}$, and the control inputs $\tau_x$, $\tau_y$, $\tau_z$, and $F_T$ are described in the body coordinate frame $\{B\}$.
• Figure 4: Prototype PCBs and schematic for interfacing with the Invensense ICM-20948 nine-axis IMU and the VL6180 ToF sensor for hardware-in-the-loop testing.
• Figure 5: Characterization of RoboBee flight dynamics and device used to replicate induced noise in our hardware-in-the-loop experiments. (a) The frequency spectrum of the translational acceleration along the $x$ and $y$ axes for 80 open-loop RoboBee flight experiments, from external motion capture (Vicon) data. The frequency spectrum displays two primary modes of interest that contribute to oscillations: the "body" mode and "flapping wing" mode. (b) The vibration unit is designed to reproduce the body oscillation dynamics observed in (a) using a low frequency vibration motor. (c) A PWM-based control circuit is used to adjust the speed of the vibration motor to allow us to approximate the magnitude of the oscillations shown in (a). (d) The amplitude of oscillations as a function of PWM duty cycle relative to the amplitude of the body oscillation mode.