Build a Bifacial Solar Array Albedo Tracking Optimizer — implementation

## Build
Model rear-side irradiance on bifacial photovoltaic arrays against seasonal ground albedo data using pvlib. Generate optimal daily tracking angles using Optuna to maximize combined front and rear energy yield.

## User
Solar Farm Operations Engineer

## Workflow
1. Ingest seasonal ground albedo telemetry, weather forecasts, and array layout parameters.
2. Calculate front and rear irradiance utilizing pvlib bifacial performance models.
3. Identify tracking angles that maximize total panel generation under diffuse light.
4. Synthesize an updated daily tracking dispatch schedule for array actuators.

## Interface
A web dashboard displaying seasonal ground albedo trends, diffuse light fractions, and recommended dual-axis tracking adjustments.

## Data
Create synthetic normalized records modeling site ground albedo, diffuse fraction, solar zenith, and bifacial panel specifications in SQLite. These serve as simplified synthetic app records, not complete official schema files.
Starter examples below are synthetic. They are not downloaded records.

## Source
[PVLib solar modeling](https://raw.githubusercontent.com/pvlib/pvlib-python/main/README.md)
Applies pvlib bifacial modeling features to calculate front and rear plane-of-array irradiance.

## Tools
optuna, pvlib, fastapi, pydantic.
Optimizes tracking tilt angles across PV strings subject to combined front and rear generation non-linearities and albedo shifts.
Fast reasoning API or local 8B parameter model for explaining tracking trade-offs and rear-side yield margins.

## Deliverables
Submit runnable source, pinned dependencies, setup commands, a working interface, example outputs, and a recorded end-to-end demo. Include tests and an action log.

## Acceptance
- Calculates combined front and rear plane-of-array irradiance.
- Enforces actuator tracking limits and collision constraints.
- Outputs hourly tracking angles that increase theoretical yield over standard front-only tracking algorithms.
- Requires human validation before updating actual field controllers.

## Starter Examples
```json
[
  {
    "input": {
      "site_id": "BIFACIAL-01",
      "ground_albedo": 0.25,
      "diffuse_fraction": 0.4,
      "solar_zenith": 35
    },
    "expected_output": {
      "optimal_tilt_angle": 30,
      "front_irradiance_w_m2": 700,
      "rear_irradiance_w_m2": 85,
      "total_yield_gain_percent": 4.2
    }
  },
  {
    "input": {
      "site_id": "BIFACIAL-02",
      "ground_albedo": 0.6,
      "diffuse_fraction": 0.8,
      "solar_zenith": 45
    },
    "expected_output": {
      "optimal_tilt_angle": 15,
      "front_irradiance_w_m2": 400,
      "rear_irradiance_w_m2": 150,
      "total_yield_gain_percent": 12.5
    }
  }
]
```

## Safety
All tracking routes are simulated suggestions; no actual robotic actuators are triggered without human engineering verification.
Use mock external writes. Require human review before any real deployment.