Tilt Optimization
ML-trained dynamic positioning that maximizes energy capture while minimizing mechanical stress. The algorithmic core of AgriVolt tracking.
Dynamic Panel Positioning
Tilt Optimization computes the tilt and azimuth trajectories that maximize energy capture while keeping movement smooth and respecting mechanical constraints.
α(t) - Pitch/Tilt
Time-varying angle from horizontal
β(t) - Azimuth
Time-varying compass direction
Sun Vector
3D direction of incoming radiation
Current Tilt
0.0°
Mode
Tracking
What This Module Does
The algorithmic core that decides how panels should move in time, designed to integrate with the Axis Panels hardware stack.
Solar Tracking
Follow sun path to maximize irradiance
Movement Smoothing
Penalize aggressive motion for hardware longevity
Multi-Mode Support
Fixed, single-axis, and dual-axis configurations
ML Integration
Data-driven policies that improve over time
Optimization Objective
Maximize: Irradiance Integral
Dot product of solar radiation direction and panel surface normal, integrated over daylight hours
Minimize: Movement Penalty
Penalizes rapid angle changes to reduce actuator wear and mechanical stress
Combined Objective
Trade-off between energy capture and mechanical longevity
Optimization Formulation
The optimization balances two competing objectives: maximizing energy capture and minimizing mechanical wear from rapid angle changes.
Irradiance Term
Computes dot product between solar vector and panel normal. Negative/zero contributions discarded.
Movement Penalty
Squared difference between consecutive tilt angles smooths trajectory and reduces actuator wear.
Angle Bounds
Physical limits enforced: tilt range, azimuth stops, mechanical constraints.
Modes of Operation
Different tracking configurations for different project requirements
Tracking Mode Comparison
Recommendation: Single-axis tracking offers the best balance of energy gain (+15%) vs. complexity for most agrivoltaic installations.
Fixed Mounting
Yearly optimization picks the tilt and azimuth that maximize annual energy for the site.
Single-Axis Tracking
Algorithm optimizes tilt trajectory over each day, with a fixed azimuth selected for annual performance.
Dual-Axis Tracking
Both pitch and azimuth can vary in real-time, allowing maximum flexibility but highest complexity.
Optimization Results
Computed trajectories discretized for field hardware implementation
Daily Tracking Trajectory
Optimal panel angles following sun path
Yearly Tilt Optimization
Optimal fixed tilt angle varies by season
Strategy: Select representative day (15th) for each month, compute optimal angles, weight by month length for annual optimization.
Yearly Optimization Strategy
Select Representative Days
Pick 15th of each month as representative for that season
Compute Optimal Trajectories
Run optimization for each representative day with actual weather
Weight by Month Length
Scale daily energy by number of days in that month
Aggregate Annual Performance
Sum weighted monthly values for total yearly optimization
Integration with Hardware & AI
The foundation for continuous improvement through data-driven learning
Hardware Integration
Trajectories discretized at hardware-compatible rates, sent as setpoints to Raspberry Pi cluster controllers.
Local Control
ESP32 nodes implement PID control to follow setpoints. Real-time safety with travel limits and stall detection.
ML Evolution
Historical trajectories, outputs, and environmental data feed ML models. System gets smarter without hardware changes.
Why It Matters
Transparent Control
Tunable algorithms instead of opaque vendor black boxes
Hardware-Stress Balance
Explicit trade-off between energy and mechanical longevity
Defensible Core
Control layer that gets smarter as more sites come online
Ready to Optimize?
Get ML-powered tilt optimization for your agrivoltaic project.