Understanding Optimization Results¶

This guide explains how to interpret HAEO's optimization results and use them effectively.

Optimization Sensors¶

HAEO creates three main network sensors:

Optimization Cost¶

Entity ID: sensor.{network_name}_optimization_cost

Total cost over the optimization horizon in dollars.

Lower is better: HAEO minimizes this value
Includes: Import costs, export revenue, storage costs
Unit: $ (or your configured currency)

Artificial Costs

The optimization cost includes small artificial costs for battery discharge and solar generation. These encourage the optimizer to use stored energy and solar power effectively. The actual monetary cost to you is primarily the grid import/export prices.

Optimization Status¶

Entity ID: sensor.{network_name}_optimization_status

Current optimization state:

optimal: Solution found successfully
feasible: Solution found but may not be optimal
infeasible: No solution exists (check constraints)
unbounded: Problem is unbounded (configuration error)
undefined: Solver error

Optimization Duration¶

Entity ID: sensor.{network_name}_optimization_duration

Time taken to solve the optimization in seconds. If this value climbs higher than you expect, adjust the horizon or period length, simplify the network, or try another solver. Review the horizon guidance before changing that value.

Element Sensors¶

For each configured element, HAEO creates sensors:

Power Sensors¶

Current optimal power at this time step (kW).

Positive: Producing/discharging/importing
Negative: Consuming/charging/exporting

Energy Sensors (Batteries)¶

Current energy level (kWh).

SOC Sensors (Batteries)¶

Current state of charge (%).

Forecast Attributes¶

All sensors include forecast attributes with future values:

attributes:
  forecast:
    - datetime: "2025-10-11T12:00:00+00:00"
      value: 1.23
    - datetime: "2025-10-11T12:05:00+00:00"
      value: 1.17
    # ... more timestamped values

Use these in automations or dashboards to visualize the optimal schedule.

Using Results in Automations¶

Example: Control Battery Based on Optimization¶

automation:
  - alias: "Follow HAEO Battery Schedule"
    trigger:
      - platform: state
        entity_id: sensor.main_battery_power
    action:
      - service: battery.set_power
        data:
          power: "{{ states('sensor.main_battery_power') | float }}"

Performance Considerations¶

Optimization duration¶

Monitor the optimization duration sensor to keep solve times comfortable. Reduce the horizon, increase the period length, simplify the network, or try a different solver if it starts trending upward. Follow the horizon guidance whenever you adjust the planning window.

Update Frequency¶

HAEO re-optimizes periodically. Balance:

More frequent: Better response to changes, higher CPU usage
Less frequent: Lower CPU usage, may miss price changes

Interpreting Cost¶

The optimization cost represents the total forecasted cost over the full horizon, not just the immediate step. Track changes in this value when you adjust configuration parameters to confirm the optimiser is producing the expected behaviour.

Next Steps¶

Explore these guides to act on the optimization outputs.

Review a complete example

See how the optimization outputs drive real-world decisions.

Sigenergy example

Build automations from the results

Turn recommended power schedules into actionable automations.

Automation patterns

Monitor data updates

Understand how new sensor data triggers optimizations.

Data update guide