agentlib.modules.controller package

Submodules

agentlib.modules.controller.bangbang module

class agentlib.modules.controller.bangbang.BangBang(*, config: dict, agent: Agent)

Bases: SISOController

A bang–bang controller (2 step or on–off controller), also known as a hysteresis controller, that switches abruptly between two states.

do_step(inp_var: AgentVariable)

Controller step function. Needs to be a generator function, thus using yield instead of return

property gain

Get the gain of the BangBang controller

property last_out_val

Last output value of the controller

loop_sim()

Loop over the do_step function

pydantic model agentlib.modules.controller.bangbang.BangBangConfig

Bases: SISOControllerConfig

Special config for a BangBang-Controller

Config:

• arbitrary_types_allowed: bool = True

• validate_assignment: bool = True

• extra: str = forbid

• frozen: bool = True

Fields:

• gain (float)

Validators:

field gain: float = 1

model_post_init(context: Any, /) → None

This function is meant to behave like a BaseModel method to initialise private attributes.

It takes context as an argument since that’s what pydantic-core passes when calling it.

Parameters:

• self – The BaseModel instance.

• context – The context.

agentlib.modules.controller.controller module

This modules defines re-use able controller modules, such as the standard Controller and the SISOController

class agentlib.modules.controller.controller.Controller(*, config: dict, agent: Agent)

Bases: BaseModule

Base class for all controller tasks within an agent

abstract do_step(inp_var: AgentVariable)

Controller step function. Needs to be a generator function, thus using yield instead of return

loop_sim()

Loop over the do_step function

process()

“Only called on run() to initialize the step.

property step: Generator

Return the generator for the do_step function

class agentlib.modules.controller.controller.SISOController(*, config: dict, agent: Agent)

Bases: Controller

Base class for all controller having one single input and one single output

property lb

The lb value

register_callbacks()

A SISO controller has only one input and only reacts to this input.

property reverse

The reverse value

property ub

The ub value

pydantic model agentlib.modules.controller.controller.SISOControllerConfig

Bases: BaseModuleConfig

Check all inputs of a SISO-Contoller

Parameters used in all SISO Controllers. ub: Upper bound of controller output lb: Lower bound of controller output reverser: Change of sign.

Config:

• arbitrary_types_allowed: bool = True

• validate_assignment: bool = True

• extra: str = forbid

• frozen: bool = True

Fields:

• input (agentlib.core.datamodels.AgentVariable)

• lb (float)

• output (agentlib.core.datamodels.AgentVariable)

• reverse (bool)

• ub (float)

Validators:

• check_bounds » lb

• check_value_type » input

• check_value_type » output

field input: AgentVariable = AgentVariable(name='u', type='float', timestamp=None, unit='Not defined', description='Not defined', ub=inf, lb=-inf, clip=False, allowed_values=[], value=0, alias='u', source=Source(agent_id=None, module_id=None), shared=None, rdf_class=None)

Validated by:

• check_value_type

field lb: float = -inf

Validated by:

• check_bounds

field output: AgentVariable = AgentVariable(name='y', type='float', timestamp=None, unit='Not defined', description='Not defined', ub=inf, lb=-inf, clip=False, allowed_values=[], value=0, alias='y', source=Source(agent_id=None, module_id=None), shared=None, rdf_class=None)

Validated by:

• check_value_type

field reverse: bool = False

field ub: float = inf

validator check_bounds  »  lb

Check if upper and lower bound values are correct

validator check_value_type  »  output, input

model_post_init(context: Any, /) → None

This function is meant to behave like a BaseModel method to initialise private attributes.

It takes context as an argument since that’s what pydantic-core passes when calling it.

Parameters:

• self – The BaseModel instance.

• context – The context.

agentlib.modules.controller.pid module

class agentlib.modules.controller.pid.PID(*, config, agent)

Bases: SISOController

A proportional–integral–derivative controller (PID controller or three-term controller) with anti-wind up. It continuously calculates an error value e(t) as the difference between a desired set point and a measured process variable and applies a correction based on proportional, integral, and derivative terms (denoted P, I, and D respectively) +——————–+—————+—————+———————+ | Parameter | Kp | Ki=(Kp/Ti) | Kd=(Kp*Td) | +——————–+—————+—————+———————+ | Rise time | Decrease | Decrease | Minor change | | Overshoot | Increase | Increase | Decrease | | Settling time | Small change | Increase | Decrease | | Steady-state error | Decrease | Eliminate | No effect in theory | | Stability | Degrade | Degrade | Improve if Kd small | +——————–+—————+—————+———————+

Configs:

setpoint (float): Set point of the controller Kp (float): proportional gain Ti (float): integration time in s Td (float): derivative time in s ub (float): high control limit lb (float): low control limit reverse(boolean): change of sign

property Kp

Get the current Kp value from data_broker or config

property Td

Get the current Td value from data_broker or config

property Ti

Get the current Ti value from data_broker or config

do_step(inp_var: AgentVariable)

Controller step function. Needs to be a generator function, thus using yield instead of return

loop_sim()

Loop over the do_step function

property setpoint

Get the current setpoint value from data_broker or config

pydantic model agentlib.modules.controller.pid.PIDConfig

Bases: SISOControllerConfig

Pydantic data model for pid controller configuration parser

Config:

• arbitrary_types_allowed: bool = True

• validate_assignment: bool = True

• extra: str = forbid

• frozen: bool = True

Fields:

• Kp (agentlib.core.datamodels.AgentVariable | float)

• Td (agentlib.core.datamodels.AgentVariable | float)

• Ti (agentlib.core.datamodels.AgentVariable | float)

• setpoint (agentlib.core.datamodels.AgentVariable | float)

Validators:

• convert_to_variable » Kp

• convert_to_variable » Td

• convert_to_variable » Ti

• convert_to_variable » setpoint

field Kp: AgentVariable | float = AgentVariable(name='Kp', type='float', timestamp=None, unit='Not defined', description='Proportional gain', ub=inf, lb=-inf, clip=False, allowed_values=[], value=1, alias='Kp', source=Source(agent_id=None, module_id=None), shared=None, rdf_class=None)

Validated by:

• convert_to_variable

field Td: AgentVariable | float = AgentVariable(name='Td', type='float', timestamp=None, unit='seconds', description='Derivative time in s', ub=inf, lb=-inf, clip=False, allowed_values=[], value=0, alias='Td', source=Source(agent_id=None, module_id=None), shared=None, rdf_class=None)

Validated by:

• convert_to_variable

field Ti: AgentVariable | float = AgentVariable(name='Ti', type='float', timestamp=None, unit='Not defined', description='Integration time in s', ub=inf, lb=-inf, clip=False, allowed_values=[], value=inf, alias='Ti', source=Source(agent_id=None, module_id=None), shared=None, rdf_class=None)

Validated by:

• convert_to_variable

field setpoint: AgentVariable | float = AgentVariable(name='setpoint', type='float', timestamp=None, unit='Not defined', description='Pid Setpoint', ub=inf, lb=-inf, clip=False, allowed_values=[], value=None, alias='setpoint', source=Source(agent_id=None, module_id=None), shared=None, rdf_class=None)

Validated by:

• convert_to_variable

validator convert_to_variable  »  Td, Ti, setpoint, Kp

model_post_init(context: Any, /) → None

This function is meant to behave like a BaseModel method to initialise private attributes.

It takes context as an argument since that’s what pydantic-core passes when calling it.

Parameters:

• self – The BaseModel instance.

• context – The context.

Module contents