EWRouter example

This notebook uses OptiWindNet to route the collector system cables via EWRouter.

⚡ EWRouter (Esau-Williams Heuristic)

EWRouter is a fast and lightweight heuristic router based on the Esau-Williams algorithm. It builds a spanning tree that connects turbines to substations while respecting cable capacity constraints. This method is especially suitable for electrical network optimization in quick prototyping and large-scale turbine layouts where optimization speed is more important than global optimality.

EWRouter is the default router used in WindFarmNetwork. If no router is specified when creating a WindFarmNetwork instance, .optimize() will automatically run with EWRouter.

🔧 Constructor: EWRouter(...)

Required arguments:

• None -> You can instantiate EWRouter() without any arguments.

Optional arguments:

Argument	Type	Description
method	str	Heuristic variant (see table below, default: "biased_EW")
maxiter	int	Maximum number of iterations (default: 10000). Controls how long the heuristic explores the search space.
feeder_route	str	Routing style:• "segmented" (default)• "straight"
bias_margin	float	Fractional margin within which edges are treated as equivalent (used by "biased_EW", "radial_EW", "ringed"). Defaults to 0.02, except "ringed" defaults to 0.25 so its return-to-root feeder tie-break has room to act.
verbose	bool	If True, prints potential warnings during routing (default: False).

Method	Topology	Description
'esau_williams'	branched	Closest to original Esau-Williams (modified to prevent crossings)
'biased_EW'	branched	EW with root-ward bias on quasi-ties
'rootlust'	branched	EW with capacity-dependent root-ward bias
'radial_EW'	radial	EW with radial subtrees (simple paths)
'ringed'	ringed	Clarke-Wright like savings formula

✅ Example

# Implicit use of EWRouter as default
wf = WindFarmNetwork(cables=..., turbinesC=..., substationsC=...)
wf.optimize()  # Uses EWRouter by default

# Explicitly use with custom options
router = EWRouter(feeder_route='straight', maxiter=5000)
wf = WindFarmNetwork(cables=..., turbinesC=..., substationsC=..., router=router)
wf.optimize()

# Ringed (closed-loop) topology
router = EWRouter(method='ringed')
wf = WindFarmNetwork(cables=..., turbinesC=..., substationsC=..., router=router)
wf.optimize()

While EWRouter is very efficient, it’s a heuristic and does not guarantee globally optimal solutions. Use it when you need speed or when solving large problems where exact methods are too slow.

from optiwindnet.api import WindFarmNetwork, EWRouter

create an instance of wfn using .from_pbf()

wfn = WindFarmNetwork.from_pbf(filepath='data/DTU_letters.osm.pbf', cables=[(7, 2000.0)])

wfn

Optimize with EWRouter

Optimize with default settings.

router = EWRouter()
res = wfn.optimize(router=router)
print(wfn.length())
wfn

1647.108905210097

Optimize with user-defined settings.

router = EWRouter(
        maxiter=20000,
        feeder_route='straight',
        verbose=True)
res = wfn.optimize(router=router)
print(wfn.length())
wfn

1707.0505869707563