Battery Optimizer

One battery's dynamic programme, solved slowly once and then read fast from anywhere. Hover the state-of-charge chart: the path is the policy replayed from that point, with no re-solve.

—saved per day vs no battery

Slow tier · solve the DP

—

every slot × every state × every action → a value function V[t, s]

dp_battery.solve_battery

Fast tier · read the policy

—

one decision: a look-ahead against the stored V

policy.action · policy.rollout

Value of a stored kWh

—

λ = dV/ds where the pointer is

policy.marginal_value

Battery capacity40 kWh

Max charge power5 kW

Max discharge power5 kW

Charge and discharge

Reserve (% of capacity)

DP grid (states × actions)

Import prices (CSV or pasted numbers)

Upload CSV

Inputs: price (left axis) and load (kW, right). Drag the points. buysellload

State of charge, with the policy as a flow field animate shift+click add a minimum · drag move · double-click remove chargedischarge

Power (kW) battery (+ charging)grid import (battery + load)

Running in your browser, no server. Same solver as the tests run: src/home_energy_optimizer/battery/webapi.py over dp_battery and policy.