Main Content

convertVibration

Compute baseline-corrected and filtered acceleration, velocity, and displacement signals from vibration measurements using a single sensor output from either an accelerometer, velocity sensor, or displacement sensor

Since R2024a

collapse all in page

Syntax

[A,V,D] = convertVibration(X,Fs)
[A,V,D] = convertVibration(T,Variable=var)
[___] = convertVibration(___,Name=Value)
[___,Options] = convertVibration(___)
convertVibration(___)

Description

The convertVibration function converts accelerometer signals into baseline-corrected and filtered velocity and displacement signals. This function is especially useful when you plan to perform machine vibration monitoring using ISO 10816 and 20816 vibration standards. You can also use convertVibration to convert displacement or velocity measurements into the three signal forms.

The ISO standards are provided for RMS velocity measurements (in mm/s) and displacement measurements (in μm), but not for accelerometer measurements (in g or mm/s2). convertVibration uses a combination of high-pass filtering, integration, and baseline correction to accurately convert accelerometer measurements into both velocity and displacement signals. The function uses a similar algorithm to convert position signals to acceleration, but with differentiation instead of integration. Converting velocity signals to acceleration and position uses a combination of differentiation and integration.

The units in the convertVibration output signals are consistent with the units of the input signals and the integration or differentiation functions that the function performs.

  • Integration (such as acceleration to velocity): Input-signal unit is multiplied by seconds, so, for example, m/s2 becomes m/s.

  • Differentiation: (such as displacement to velocity): Input-signal unit is divided by seconds, so, for example, m becomes m/s.

If the input is in units of g, you must multiply the output by 9.81 to convert the results to m/s (velocity) or m (displacement).

Return Converted Signals

[A,V,D] = convertVibration(X,Fs) returns baseline-corrected signals for acceleration A, velocity V, and displacement D from the measured acceleration data X.

example

[A,V,D] = convertVibration(T,Variable=var) uses measured acceleration data from the variable var in timetable T.

example

[___] = convertVibration(___,Name=Value) allows you to specify additional parameters using one or more name-value arguments. For example, if your data in T represents a velocity signal rather than an acceleration signal, use the name-value argument Type="velocity".

example

[___,Options] = convertVibration(___) also returns the Options structure containing various parameter values that the software used to perform the conversion, such as sampling frequency Fs and lowpass filter cutoff frequency Fmax.

You can use this syntax with any of the previous input and output argument combinations.

example

Plot Converted Signals

convertVibration(___) plots the converted signals.

example

Examples

collapse all

Open Live Script

Load the data, which consists of a true acceleration signal acc_true, the measured acceleration acc_meas, which includes noise, offset, and drift, and the sampling frequency Fs.

load acceldata acc_meas acc_true Fs

Generate a time vector from Fs. 

t = (0:1/Fs:0.4);

Plot the true and corrupted signals together.

plot(t,acc_meas,t,acc_true)
title('Acceleration Signal')
legend('Corrupted','True')

Figure contains an axes object. The axes object with title Acceleration Signal contains 2 objects of type line. These objects represent Corrupted, True.

Use convertVibration to apply baseline correction to the accelerometer signal, and convert the signal into velocity and displacement signals. Use the syntax that plots all three signals.

convertVibration(acc_meas, Fs);

Figure contains 3 axes objects. Axes object 1 with title Displacement, xlabel Time, ylabel Displacement contains an object of type line. This object represents Estimated Signal. Axes object 2 with title Velocity, xlabel Time, ylabel Velocity contains an object of type line. This object represents Estimated Signal. Axes object 3 with title Acceleration, xlabel Time, ylabel Acceleration contains 2 objects of type line. These objects represent Corrected Signal, Input Signal.

The conversion corrects the acceleration signal and produces stable displacement and velocity signals.

Save the signals by specifying output arguments.

[A,V,D] = convertVibration(acc_meas,Fs);
Open Live Script

Load the data, which consists of a true velocity signal vel_true, the measured velocity vel_meas, which includes noise, offset, and drift, and the sampling frequency Fs. The signals are stored in timetables.

load veldata vel_meas_tt vel_true_tt Fs

Generate a time vector from Fs.

t = (0:1/Fs:0.2)';

Plot the true and corrupted signals together.

plot(t,vel_meas_tt.vel,t,vel_true_tt.vel)
title('Velocity Signal')
legend('Corrupted','True');

Figure contains an axes object. The axes object with title Velocity Signal contains 2 objects of type line. These objects represent Corrupted, True.

Use convertVibration to apply baseline correction to the velocity signal, convert and correct the acceleration and displacement signals, and return all three signals. Set the cutoff frequency Fmax to 500 to reduce the effect of high-frequency noise. Return all three signal types as well as information on the options used.

[a,v,d,options] = convertVibration(vel_meas_tt,Var="vel",Type="velocity",FMax=500);

Examine the options you used.

options
options = struct with fields:
        Fmin: 20
        Fmax: 500
        Type: "velocity"
    Variable: "vel"
          Fs: 10000

options contains information on the frequency range, the signal type, the variable name, and the sampling frequency.

Plot the three signals.

convertVibration(vel_meas_tt,Var="vel",Type="velocity",FMax=500);

Figure contains 3 axes objects. Axes object 1 with title Displacement, xlabel Time, ylabel Displacement contains an object of type line. This object represents Estimated Signal. Axes object 2 with title Velocity, xlabel Time, ylabel Velocity contains 2 objects of type line. These objects represent Corrected Signal, Input Signal. Axes object 3 with title Acceleration, xlabel Time, ylabel Acceleration contains an object of type line. This object represents Estimated Signal.

Compare the corrected velocity with the true velocity.

figure
plot(t,v,t,vel_true_tt.vel)
legend('Corrected Velocity','True Velocity')

Figure contains an axes object. The axes object contains 2 objects of type line. These objects represent Corrected Velocity, True Velocity.

The corrected velocity has a much better fit than the measured velocity to the true velocity.

Input Arguments

collapse all

Measured vibration data, specified as numeric vector containing double-precision or single-precision data.

Data sampling rate in Hz, specified as a positive numeric scalar.

Timetable containing measured vibration data, specified as a timetable that can contain multiple data variables.

Timetable variable in T to convert, specified as a string.

Name-Value Arguments

Specify optional pairs of arguments as Name1=Value1,...,NameN=ValueN, where Name is the argument name and Value is the corresponding value. Name-value arguments must appear after other arguments, but the order of the pairs does not matter.

Example: convertVibration(v_tt,Var=vel,Type="velocity") converts velocity measurements stored in the variable vel of the timetable v_tt into corrected velocity, acceleration, and position signals and plots them.

Type of input signal, specified as one of the following values: "acceleration", "velocity", or "displacement".

Lower end of the frequency range used for high-pass filtering, specified as a scalar in Hz. The software uses highpass filtering to remove DC offset and drift due to low-frequency noise. High-pass filtering also improves computations involving integration (for example, when obtaining velocity from an acceleration signal). The value of Fmin must be less than the Nyquist frequency Fs/2.

Upper end of the frequency range used for low-pass filtering, specified as a scalar in Hz. The software uses lowpass filtering to remove high-frequency noise. Low-pass filtering also improves computations involving differentiation (for example, when obtaining acceleration from a velocity signal). The value of Fmax must be less than the Nyquist frequency Fs/2.

Output Arguments

collapse all

Baseline-corrected acceleration signal, returned as a numeric vector. A can be computed directly from an accelerometer signal or converted from a velocity or displacement sensor signal.

Baseline-corrected velocity signal, returned as a numeric vector. V can be computed directly from a velocity sensor signal or converted from an accelerometer or a displacement sensor signal.

Baseline-corrected displacement signal, returned as a numeric vector. D can be computed directly from a displacement sensor signal or converted from an accelerometer or a velocity sensor signal.

Options used for conversion, specified as a structure.

Algorithms

The convertVibration algorithm is based on the numerical integrator design described in [1]. The integrator contains three stages, each stage of which includes some combination and sequence of offset or baseline calibration, highpass filtering, and integration. The result is a stable velocity signal and displacement signal. The following block diagram, adapted from [1], illustrates the general algorithm flow for an accelerometer input.

Block diagram of vibration conversion that starts with accelerometer input on the top left corner and corrected acceleration, velocity, and position outputs along the top row

A similar algorithm converts position data to the three corrected outputs, but uses differentiation instead of integration. For converting velocity data, convertVibration uses a combination of integration and differentiation.

References

[1] S. Thenozhi, W. Yu and R. Garrido, "A novel numerical integrator for structural health monitoring." 2012 5th International Symposium on Resilient Control Systems, Salt Lake City, UT, USA, 2012, pp. 92-97, doi: 10.1109/ISRCS.2012.6309300.

Version History

Introduced in R2024a

See Also

External Websites