Simulated Devices¶

Simulated devices are BEC devices whose readback values are produced by a simulation object instead of physical hardware. They are useful for onboarding, GUI exploration, DAP testing, and safe workflow development.

Not every device is simulated. A device exposes simulation controls only when its device class provides a sim object and exposes it to the client. In the BEC IPython client, that appears as:

dev.<device>.sim

For example:

dev.bpm4i.sim

Simulation Interface¶

The common simulation interface exposes these user-facing controls:

API	Purpose
`dev.<device>.sim.get_models()`	List simulation models available for the device.
`dev.<device>.sim.select_model("ModelName")`	Select the active simulation model.
`dev.<device>.sim.params`	Read or update parameters for the active model.
`dev.<device>.sim.show_all()`	Print the active model, parameters, methods, and model list.

This is runtime state. It is quick to change from the IPython client and is the preferred way to explore a simulated setup interactively.

Model Types¶

The available model names depend on the simulated device type.

1D monitor and waveform simulations use LMFit model classes. A monitor such as bpm4i can use names such as GaussianModel, ConstantModel, or other available LMFit built-in model class names. BEC DAP uses the same LMFit model naming style, which is why GaussianModel can be used both for simulated bpm4i data and for a Waveform DAP curve.

1D monitor simulations also add BEC-specific parameters:

Parameter	Purpose
`ref_motor`	Motor whose current position is used as the model input.
`noise`	Noise mode, for example `uniform`, `poisson`, or `none`.
`noise_multiplier`	Scale used by the noise model.

By default, ref_motor is samx. If your scan uses another motor, update ref_motor so the simulated signal follows the motor you are scanning.

Simulated positioners are different. They simulate positioner behavior, but they do not expose LMFit model selection for their readback.

2D simulated cameras are also different. They use camera-specific model names such as constant and gaussian, not LMFit class names such as GaussianModel.

Runtime Setup and Persistent Defaults¶

Runtime setup changes the current device object:

dev.bpm4i.sim.select_model("GaussianModel")
dev.bpm4i.sim.params = {"amplitude": 100, "center": 0, "sigma": 1, "ref_motor": "samx"}

This is the right approach for exploration, onboarding, demos, and temporary test sessions.

For long-term defaults, put the same setup in the device configuration through sim_init. During device initialization, BEC selects the configured model and applies the configured parameters:

bpm4i:
  deviceConfig:
    sim_init:
      model: GaussianModel
      params:
        amplitude: 100
        center: 0
        sigma: 1
        ref_motor: samx

Use sim_init when a simulated device should always start with the same model in a beamline, plugin, or test configuration.