Complementary Filter for Two wheel System

sensor fusion technique: multiple sensors are used

Complementary filter
: Combines two sensor of having different frequency responses

1. Idea
: to use appropriate filters for typical sensor characteristics to suppress the corresponding noises.
: 일반적인 센서 특성에 적합한 필터를 사용하여 해당 노이즈를 억제하는 것

• Sensor characteristics? 

    the complementary filter consists of two filters

1.  lowpass filter for the tilt sensor 
2.  highpass filter for the gyro sensor

        But, gyro sensor has a drift when integration is performed

 

    Therefore, a combination of two sensors is required to estimate an angle

     which is a complementary filter.

 

2. Filter Algorithm 

complementary filter

: lowpass filter(for the tilt sensor) + highpass filter(for the gyro sensor)

Lowpass filter for tilt

 

Highpass filter for gyro

 

                                                                                <T is a time constant>

 

 

• Filter Structure

 

 

 

    : Angle data in all ranges of the frequency are summed together to be one

 

 

 

 

    : psi ^ = estimates the balancing angle

    : ηH   =  High frequency noise

    : ηL   =  Low frequency noise

 

 

 

    : psi ^ = can be obtained by 'psi'

    : psi  = filtering the real angle 

    : Since noises are filtered out, becomes close to the real angle

    : Whether psi  is a low or high-frequency signal, 

      the corresponding filters are active to cancel noises in low and high frequencies

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

        Offset Values:  obtained from empirical studies for calibration