Exponential Filter

 In Arduino Libraries

The ExponentialFilter class implements a simple linear recursive exponential filter for the Arduino. It provides a simple way to smooth noisy measurements from analog sensors without using as much memory as a moving average filter.

Every time you provide a new value (xn), the exponential filter updates a smoothed value (yn):

yn = w × xn + (1 – w) × yn – 1

Here:

  • yn is the output of the filter at a moment in time n
  • xn is the new input value at a moment in time n
  • yn – 1 is the previous output value of the filter
  • w is the weighting factor in the range [0, 100]. High values of w (90, for example) favor new data over old data. The output responds quickly to changes in the input but is not smoothed much. Low values of w (10, for example) favor old data over new data. The filter output is heavily smoothed and responds slowly to changes (noisy or not) in the input .

Functions

The ExponentialFilter is a template class that takes a single parameter: the type of measurement to filter. The current filtered value is multiplied by 10 internally to improve precision of the filter. Normally use either long (to save code space if you haven’t already used float’s) or float (for more accurate results) as the filter parameter.

The constructor takes two arguments:

  1. The initial filter weight. This should be between 0 and 100
  2. The initial value. Typically 0.

Declare a new exponential filter using:

Filter(NewValue)

Applies the filter to a new value. See example below.

SetWeight(NewWeight)

Changes the weight applied when filtering. This value should be between 0 and 100.

High weights (90, for example) favor new data over old data. So, the output responds quickly to changes in the input and is not smoothed much.

Low values of (10, for example) favor old data over new data. So, the output is heavily smoothed and the filter responds slowly to changes (noisy or not) in the input.

GetWeight()

Returns the current value of the filter weight parameter.

Current()

Returns the current value of the filter output.

SetCurrent(NewValue)

Forces the current value to the new value supplied. Typically used to initialize the filter from a measurement.

Example

This example reads the channel-0 ADC on an Arduino. Both the raw measurement and a heavily filtered value are sent to MegunoLink for plotting.

The noisy output from the analog-to-digital converter (green series) is heavily filtered here (red series):

ADC data smoothed with an exponential filter

The noisy measurements from the ADC (green) are smoothed by an exponential filter (red)

Availability

The exponential filter is included in our MegunoLink Arduino Library and can be installed using the Arduino integration tool.

Recent Posts
Showing 3 comments
  • Matěj Suchánek
    Reply

    Cool! Where can I get it?

    • Matěj Suchánek
      Reply

      Nevermind, found it! 😀

pingbacks / trackbacks

Leave a Comment

Contact Us

Send us an email and we will get back to you shortly.

Not readable? Change text.

Start typing and press Enter to search