Skip to content

IMU

The IMU class in MMFS represents an advanced inertial sensor interface designed to support both raw sensor reporting and fused orientation estimation. It processes data from gyroscopes, accelerometers, and optionally magnetometers, and internally runs a quaternion-based complementary filter to estimate device orientation.

Unlike basic sensors, IMUs operate in a multi-frame context — reporting both local-frame measurements and global-frame acceleration — and may expose both raw and filtered signals. Orientation is represented using quaternions but is often interpreted in Euler angles for ease of telemetry.

Overview

An IMU in MMFS provides:

  • Angular velocity (from gyroscope)
  • Linear acceleration (from accelerometer)
  • Magnetic field strength (from magnetometer)
  • Quaternion orientation (internally computed)
  • Gravity-compensated acceleration (global frame)

Inheritance Structure

IMU inherits from Sensor and expects the user to override two core methods:

1
2
bool init() override;
void read() override;
  • init() sets up the hardware driver(s), configures sensors, and prepares communication.

  • read() must populate the following fields with current measurements:

  • measuredAcc — linear acceleration in body frame

  • measuredGyro — angular velocity in deg/s
  • measuredMag — magnetic field vector (optional)

The class handles filtering and fusion internally during update().

Orientation Fusion

The IMU class performs sensor fusion using a quaternion-based complementary filter. This combines fast-but-drifting gyro data with stable-but-noisy accelerometer estimates of gravity direction. Magnetometers can further improve yaw stability.

The computed orientation is stored in:

1
Quaternion orientation;

Global-frame acceleration (i.e. gravity-compensated) is computed using:

1
Vector<3> getAccelerationGlobal();

and is derived by rotating the measured acceleration vector into the world frame.

Exposed Data Columns

The IMU automatically exposes the following:

  • accX, accY, accZ — linear acceleration (m/s²)
  • gyroX, gyroY, gyroZ — angular velocity (deg/s)
  • magX, magY, magZ — magnetic field (μT)
  • oriX, oriY, oriZ, oriW — quaternion orientation

These are registered for logging and telemetry via DataReporter and packed for transmission.

Bias Correction Mode

Bias correction enables zeroing gyro and acceleration offsets over time. If enabled before flight, the IMU will average recent values to zero-out drift while stationary.

1
2
imu.setBiasCorrectionMode(true);
imu.markLiftoff(); // Locks in calibration at launch

Avoid enabling this while the vehicle is moving.

Usage Flow

1
2
3
4
5
6
7
8
MyIMU imu;
imu.begin();

loop() {
    imu.update();
    Quaternion q = imu.getOrientation();
    Vector<3> a_global = imu.getAccelerationGlobal();
}

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
class MyIMU : public IMU {
    MyIMULib hw;

    bool init() override {
        return hw.begin();
    }

    void read() override {
        measuredAcc = { hw.ax(), hw.ay(), hw.az() };
        measuredGyro = { hw.gx(), hw.gy(), hw.gz() };
        measuredMag = { hw.mx(), hw.my(), hw.mz() }; // optional
    }
};

Built-in Implementations

BMI088 + LIS3MDL

  • Combines separate accelerometer, gyro, and magnetometer
  • Uses MMFS’s internal filter and fusion logic
  • Best for flexible, high-performance inertial applications

BNO055

  • Integrated 9DOF sensor with onboard fusion
  • MMFS defers to the sensor’s internal orientation logic
  • Returns pre-filtered quaternion and acceleration data

Advanced Controls

The following parameters can be adjusted to control noise filtering sensitivity during static startup:

1
2
imu.setAccelBestFilteringAtStatic(val); // threshold for accel stability
imu.setMagBestFilteringAtStatic(val);   // threshold for mag stability

Use these to tune how long the IMU waits before accepting a bias-corrected origin.

Summary

  • Implements Sensor with IMU-specific orientation logic
  • Expects raw accel, gyro, and optional mag readings
  • Computes quaternion orientation using a complementary filter
  • Supports packed telemetry and gravity-compensated acceleration
  • Used directly or as a base class for sensor-specific drivers

Written by ChatGPT. May not be fully accurate; verify against source.