Signal processing module

The Signal Processing Tool is a Python library for signal processing, specifically focusing on time and frequency domain analysis. It provides a framework for creating, importing, editing, manipulating and analysing signals. The tool is developed by the Advanced Technology and Research (ATR) Competence Center team of Haskoning as part of the AT&R package.

Warning

The Signal Processing Tool is currently under active development. Functionality may change in future updates, and formal validation is still pending. Please verify your results carefully before using them in critical applications.

Purpose and Key Features

The primary purpose of the Signal Processing Tool is to simplify and streamline the process of signal processing. Key features include:

  • Time Domain Analysis: Tools for creating, importing, editing, manipulating, and analysing time domain signals.

  • Frequency Domain Analysis: Functions for converting time domain signals to frequency domain and analysing frequency spectra.

  • Signal Manipulation: Utilities for filtering, cropping, stitching and combining signals.

  • Visualisation: Functions for plotting time and frequency domain data.

Quick Start Guide

Before you start with the Signal Processing Tool, make sure to install the latest version of the AT&R package. Refer to ATR installation guide.

Start the project with the following code:

from haskoning_atr_tools import ATRProject
project = ATRProject(name='My Project', location='Building location', project_code='P-BH1999')

Basic Usage Example Time Signal

Continuing from the previous step in this section examples are provided for processing time signals.

Step 1 - Create TimeDomainData object

In this step you need to provide or create a signal and add the time domain data to your project. You can create a signal to use, import a signal from a file or use one of the built-in signal creators.

The options for creating the time domain data object are:

  1. Manual creation of signal: Specify the amplitude values and time stamps in lists.

    my_time_domain_data = project.create_time_domain_data(
    name='Manual Time Data',
    amplitude_type='velocity',
    amplitudes=[0.1, 0.2, 0.3, 0.4],
    time_stamps=[0, 1, 2, 3],
    amplitude_units='m/s',
    time_stamps_units='s',
    collection_name='Collection #1')

  2. Import signal: Use a method to import a file with time domain data.

    my_time_domain_data = project.create_time_domain_data_from_sigicom_csv(
    file='path/to/your/sigicom_file.csv',
    name='Sigicom Time Domain Data',
    amplitude_type='velocity',
    amplitude_units='mm/s',
    time_stamps_units='seconds',
    collection_name='Example Collection')

  3. Use built in signal creators: Use a method to create a signal.

    my_time_domain_data = my_project.create_sinusoidal_time_domain_data(
    name='Example Sinusoidal Signal',
    harmonic_amplitudes=[1.0, 0.3],
    harmonic_frequencies=[10.0, 20.0],
    harmonic_phase_shift=[0.0, 45.0],
    duration=5.0,
    sampling_rate=100,
    collection_name='Example Collection',
    amplitude_type='displacement',
    amplitude_units='mm',
    time_stamps_units='seconds')

More options available for the user to create time domain data are described in the advanced section on creating signals.

Step 2 - Edit and Manipulate Time Domain Data

The tool provides different options to edit the input signal. Tools to crop, combine, scale, shift, rotate, filter and other manipulations are available:

cropped_time_domain_data = my_time_domain_data.crop_time_domain(from_front=True, time_amount=2)

More functions available for the user to edit and manipulate time domain data are described in the advanced section edit and manipulate signals.

Step 3 - Convert to Frequency Domain

The most common method to convert time domain data to frequency domain is the Fourier Transform. For discrete signals (e.g., sampled data), we use the Discrete Fourier Transform (DFT), often computed using the Fast Fourier Transform (FFT) algorithm. With the following syntax the time domain data is converted to frequency domain:

frequency_data = my_time_domain_data.convert_to_frequency_domain(window_type='Hanning')

More options available for the user when converting time domain data to frequency domain are described in the advanced section conversion to frequency domain.

Step 4 - Signal Properties

You can extract specific properties of the signal in either the time domain or the frequency domain, depending on your analysis needs. Commonly requested properties include:

  • Significant amplitudes

  • Peak values

  • Root Mean Square (RMS)

  • Dominant frequencies

  • Energy content

For example, to retrieve significant amplitudes from the frequency domain data:

significant_amplitudes = freq_data.find_significant_amplitudes(amplitude_threshold=0.1)

Methods for more properties of the frequency domain data are described in the advanced section details on the time and frequency domain properties.

Step 5 - Plot the Data

The signal processing tool provides built-in functionality to visualise both time domain and frequency domain data. To generate plots, simply call the plot method on the respective data object:

my_time_domain_data.plot()
frequency_data.plot()

All settings available for plotting results of the signal processing tool are described in the advanced section plotting.

Note

The procedure in the quick start guide shows a basic workflow. Different signals can be combined in the collections attribute of Project, where most of the methods are available to apply on all the signals in a collection and create combined plots.

Note

It is also possible to create a spectrum and convert it to time domain data. Refer to the advanced section creating spectra for more information.

This tool is designed to help you extract meaningful insights with ease when you’re working with vibration data, seismic signals, or custom time series. We’re continuously improving the tool and welcome your feedback, ideas, and contributions. If you’ve developed new features, discovered interesting use cases, or have suggestions for enhancements, we’d love to hear from you!

Tip

Get involved: Share your developments, report issues, or contribute via our Azure repository.

Advanced topics

In the quick-guide the basic steps to create and process a time domain signal were provided. More details on each step are provided in the following sections.

Background information

This section provides a brief theoretical overview of signal processing concepts to help you better understand the tool’s methods and capabilities. It also supports selecting the most suitable settings for your specific use case.

References

The following sections provide detailed documentation of the available classes, methods, and functions within the signal processing tool.