# Regarding Complementary Filter for Accelerometer- Gyroscope sensor fusion

When you talk of filter, how can it be defined in terms of two constant weighing coefficient? Instead it should be defined in terms of first or second order transfer functions (as per normal literature on filters).

My application is finding tilt of a cantilever structure. I am using gyro and accelerometer sensor fusion. What I found was the tilt angle can be found using complementary filter as: tilt= xgyro + (1-x) accl; where, gyro and accl are angles calculated from gyroscope and accelerometer output signals respectively, while x is the weight factor, whose value usually is $0.95-0.98$.

My question is how this algorithm acts as a filter. Though its easy to understand it weighs gyro output higher under motion condition, how does it undermine gyro output as compared to accelerometer under static condition.

Can anyone explain how exactly filtering occurs in case of Complementary filter? Thanks

• Can you please edit your question for clarity? "Two constant weighing coefficient filters" have different capabilities than first and second order transfer function filters and complementary filters. If the question is "How does the complementary filter works?" can you be a bit more specific? What are you trying to achieve? What have you done so far? – A_A Oct 7 '16 at 9:47