types.core.ControlPolicy

types.core.ControlPolicy

Control policy/controller π(x).

Maps state to control action: u = π(x).

Parameters

Name	Type	Description	Default
x	StateVector	Current state (nx,)	required

Returns

Name	Type	Description
	ControlVector	Control action (nu,)

Examples

>>> # LQR policy
>>> def lqr_policy(x: StateVector) -> ControlVector:
...     return -K @ x
>>> 
>>> # Nonlinear policy
>>> def nonlinear_policy(x: StateVector) -> ControlVector:
...     return -np.tanh(K @ x)
>>> 
>>> # Neural network policy
>>> def nn_policy(x: StateVector) -> ControlVector:
...     return model.forward(torch.tensor(x)).numpy()

# types.core.ControlPolicy { #cdesym.types.core.ControlPolicy }

`types.core.ControlPolicy`

Control policy/controller π(x).

Maps state to control action: u = π(x).

## Parameters {.doc-section .doc-section-parameters}

| Name   | Type        | Description         | Default    |
|--------|-------------|---------------------|------------|
| x      | StateVector | Current state (nx,) | _required_ |

## Returns {.doc-section .doc-section-returns}

| Name   | Type          | Description          |
|--------|---------------|----------------------|
|        | ControlVector | Control action (nu,) |

## Examples {.doc-section .doc-section-examples}

```python
>>> # LQR policy
>>> def lqr_policy(x: StateVector) -> ControlVector:
...     return -K @ x
>>> 
>>> # Nonlinear policy
>>> def nonlinear_policy(x: StateVector) -> ControlVector:
...     return -np.tanh(K @ x)
>>> 
>>> # Neural network policy
>>> def nn_policy(x: StateVector) -> ControlVector:
...     return model.forward(torch.tensor(x)).numpy()
```