UUV Noise Web

UUV Noise Web is a local web prototype for semi-empirical UUV target radiated-noise simulation. It wraps a MATLAB noise-source model with a lightweight Python backend and a browser-based frontend.

This project is a research and engineering prototype. The default parameters are illustrative and are not calibrated to a real UUV. Generated signals should not be treated as certified acoustic signatures of any specific vehicle.

What It Does

• Runs point, line, surface, and volume equivalent-source UUV noise cases.
• Generates UUV target radiated-noise previews with machinery broadband noise, motor/shaft tonal lines, propeller blade-pass-frequency tonal lines, modulated propeller/cavitation broadband noise, DEMON preview, and hull-flow broadband noise.
• Produces PNG figures, WAV audio previews, CSV tables, MAT result files, and TXT descriptions.
• Provides a local web interface for parameter input, result display, WAV playback, and file download.

Project Structure

uuv_noise_web/
  backend/        Python local HTTP backend
  frontend/       HTML/CSS/JavaScript browser UI
  matlab_model/   MATLAB simulation kernel
  jobs/           Example and runtime simulation outputs
  README.md       Project introduction and usage

Requirements

• Windows is recommended for the current local prototype.
• Python 3.10 or newer.
• MATLAB with Signal Processing Toolbox functions used by the plotting routines.
• The matlab command should be available in PowerShell/CMD PATH.

No Python web framework is required in this first version. The backend uses Python standard-library HTTP tools.

Quick Start

In PowerShell:

cd uuv_noise_web
python .\backend\server.py

Then open:

http://127.0.0.1:8765

You can also double-click:

start_local_web.bat

Output Files

Each simulation creates a task folder under jobs/<job_id>/.

Typical output files:

• uuv_source_signal.wav: 1 m equivalent source-signal preview.
• uuv_received_target.wav: simplified propagated target-signal preview.
• uuv_received_mix.wav: target plus ambient-noise preview.
• uuv_source_spectrum.png: source-level spectrum.
• uuv_waveforms.png: waveform preview.
• uuv_lofar.png: LOFAR-style low-frequency display.
• uuv_demon.png: DEMON envelope-spectrum preview.
• uuv_source_geometry_3d.png: equivalent-source geometry.
• uuv_source_spectrum.csv: spectrum data table.
• uuv_tonal_lines.csv: tonal-line table.
• uuv_source_geometry.csv: source-element coordinates and weights.
• uuv_source_model_result.mat: full MATLAB result.
• uuv_noise_description.txt: readable description of the generated signal.

The repository includes one successful demo job in jobs/20260818-091054-722e7c1b/.

Architecture

Browser UI
  -> frontend/app.js sends JSON parameters
Python backend
  -> backend/server.py creates a job folder and calls MATLAB
MATLAB model
  -> matlab_model/web_run_case.m runs the source model
Job output
  -> PNG/WAV/CSV/MAT/TXT files are served back to the browser

Notes

This version focuses on getting the full local workflow running. It is not yet a production web system. Future versions can replace the standard-library backend with FastAPI, add user accounts, add a real job queue, and connect the source model to Bellhop/RAM/Kraken propagation workflows.