rcognita.simulator.Simulator

class rcognita.simulator.Simulator(system, state_init, sys_type='diff_eqn', disturb_init=None, action_init=None, time_start=0, time_final=1, sampling_time=0.01, max_step=0.005, first_step=1e-06, atol=1e-05, rtol=0.001, is_disturb=0, is_dynamic_controller=0, ode_backend='SciPy')

Class for simulating closed loops (system-controllers).

sys_type

Type of system by description:

diff_eqn : differential equation \(\mathcal D state = f(state, u, q)\)
discr_fnc : difference equation \(state^+ = f(state, u, q)\)
discr_prob : by probability distribution \(X^+ \sim P_X(state^+| state, u, q)\)
Type

: string

where
\(state\) : state
\(u\) : input
\(q\) : disturbance
compute_closed_loop_rhs

Right-hand side description of the closed-loop system. Say, if you instantiated a concrete system (i.e., as an instance of a subclass of system class with concrete compute_closed_loop_rhs method) as system, this could be just system.compute_closed_loop_rhs.

Type

: function

sys_out

System output function. Same as above, this could be, say, system.out.

Type

: function

is_dynamic_controller

If 1, the controller (a.k.a. agent) is considered as a part of the full state vector.

Type

: 0 or 1

state_init, disturb_init, action_init

Initial values of the (open-loop) system state, disturbance and input.

Type

: vectors

time_start, time_final, sampling_time

Initial, final times and time step size

Type

: numbers

max_step, first_step, atol, rtol

Parameters for an ODE solver (used if sys_type is diff_eqn).

Type

: numbers

See also

systems module

__init__(system, state_init, sys_type='diff_eqn', disturb_init=None, action_init=None, time_start=0, time_final=1, sampling_time=0.01, max_step=0.005, first_step=1e-06, atol=1e-05, rtol=0.001, is_disturb=0, is_dynamic_controller=0, ode_backend='SciPy')
Parameters

  • sys_type (: string) –

    Type of system by description:

    diff_eqn : differential equation \(\mathcal D state = f(state, u, q)\)
    discr_fnc : difference equation \(state^+ = f(state, u, q)\)
    discr_prob : by probability distribution \(X^+ \sim P_X(state^+| state, u, q)\)

  • where –

    \(state\) : state
    \(u\) : input
    \(q\) : disturbance

  • compute_closed_loop_rhs (: function) – Right-hand side description of the closed-loop system. Say, if you instantiated a concrete system (i.e., as an instance of a subclass of System class with concrete compute_closed_loop_rhs method) as system, this could be just system.compute_closed_loop_rhs.

  • sys_out (: function) – System output function. Same as above, this could be, say, system.out.

  • is_dynamic_controller (: 0 or 1) – If 1, the controller (a.k.a. agent) is considered as a part of the full state vector.

  • state_init (: vectors) – Initial values of the (open-loop) system state, disturbance and input.

  • disturb_init (: vectors) – Initial values of the (open-loop) system state, disturbance and input.

  • action_init (: vectors) – Initial values of the (open-loop) system state, disturbance and input.

  • time_start (: numbers) – Initial, final times and time step size

  • time_final (: numbers) – Initial, final times and time step size

  • sampling_time (: numbers) – Initial, final times and time step size

  • max_step (: numbers) – Parameters for an ODE solver (used if sys_type is diff_eqn).

  • first_step (: numbers) – Parameters for an ODE solver (used if sys_type is diff_eqn).

  • atol (: numbers) – Parameters for an ODE solver (used if sys_type is diff_eqn).

  • rtol (: numbers) – Parameters for an ODE solver (used if sys_type is diff_eqn).

Methods

__init__(system, state_init[, sys_type, …])

param sys_type

Type of system by description:

do_sim_step(*args, **kwargs)

get_sim_step_data()

Collect current simulation data: time, system state and output, and, for completeness, full closed-loop state.

initialize_ODE_solver()

reset()