⚙️ How the Calculation Engine Works (APRIORIENGINE)

The calculation engine in the TotalPropagatedError Django application computes Total Propagated Uncertainty (TPU) for hydrographic surveys. It is driven by a custom Python module named APRIORIENGINE, which models both Total Horizontal Uncertainty (THU) and Total Vertical Uncertainty (TVU).

🧠 1. Core Design

• Input ingestion from the Django form
• Configuration of all relevant sensor and survey parameters
• Uncertainty computation using propagation models
• Output generation in both tabular and visual (Plotly) form
• Session-based export for reproducibility

📥 2. Data Input & Initialization

User inputs from the TPU web form are POSTed and used to configure the engine:

engine = apriori.APRIORIENGINE()
engine.positionsensorname = request.POST['PositionSensor']
engine.patchtestreliabilityPITCH = float(request.POST['patchtestreliabilityPITCH'])
engine.sectorangle = int(request.POST['SectorAngle'])

Key inputs include:

• Project metadata (name, duration, user, etc.)
• Sensor choices (MBES, Position, Attitude, Tide)
• Platform installation offsets (IMU, GNSS)
• Environmental variables (roll, pitch, heave, SVP)

🔧 3. Configuration Phase

After populating input values, the engine calls:

engine.configure()

This performs:

• Sanity checks on input ranges
• Derivation of internal variables
• Preparation of the computation matrix (e.g., beam geometry, ping spacing)

📊 4. Core Uncertainty Computation

The main computational steps are:

engine.thu()

• GNSS horizontal accuracy
• Sensor installation offsets
• Vessel motion (roll, yaw)
• Timing & latency
• Patch test quality

engine.tvu()

• Heave
• Pitch
• GNSS height errors
• Tide model (if included)
• Draft and squat

These are calculated using Root-Sum-Square (RSS) propagation:

total_error = √(error₁² + error₂² + ... + errorₙ²)

📏 5. Specification Compliance

The engine includes built-in standards for acceptable uncertainty levels.

thuacceptance = engine.testagainstspecification(engine.permittedhorizontalaccuracies, engine.thucombinedtotal)
tvuacceptance = engine.testagainstspecification(engine.permittedverticalaccuracies, engine.tvucombinedtotal)

📈 6. Angle-Specific TVU Sweep

Sample loop for computing depth vs. angle TVU values:

for angle in range(-engine.sectorangle, engine.sectorangle):
    tvuangle.sectorangle = angle
    tvuangle.tvu()
    meanresults.append([angle, tvuangle.tvucombinedtotalsummary[1][2]])

📑 7. Output Tables and Plots

• inputtable – original configuration
• outputtable – calculated THU/TVU
• thuplot, tvuplot – error visualizations
• beamgeometryplot, swathplot – diagnostic graphs

💾 8. Saving and Exporting

Saved using the current session:

engine.saveconfiguration(request.session.session_key)

Includes:

• .ini config snapshot
• .csv summaries
• .html and .png plots
• Packaged ZIP download

♻️ 9. Session Isolation

All state is tied to the Django session key, ensuring multi-user concurrency and secure, isolated computations.

✅ Summary

• Accepts detailed survey inputs
• Computes propagated THU and TVU
• Validates against standards
• Exports data and visualizations
• Session-based for isolation and reproducibility