andes.core.discrete.Derivative

class andes.core.discrete.Derivative(u, name=None, tex_name=None, info=None)

Compute the derivative of a variable using numerical differentiation.

Derivative is based on the storage implemented in the Delay class. The delay is set to 1 step so that the current and the previous step are used.

A simple first order derivative is computed using u(t) - u(t-1) / tstep, where tstep is the current step size.

Derivative is intended to be used for algebraic variables because of discontinuity. It can be applied to a state variable, but one should instead implement the right-hand side equation of the state variable in an algebraic equation to obtain the accurate derivative.

Alternatively, the washout filter (andes.core.block.Washout) can be used to implement a numerically stable derivative.

The output of the Derivative class is named <INSTANCE_NAME>_v just like Delay.

__init__(u, name=None, tex_name=None, info=None)

Methods

check_eq(**kwargs)	This function is called in l_check_eq after updating equations.
check_iter_err([niter, err])	Check if the minimum iteration or maximum error is reached so that this discrete block should be enabled.
check_var(dae_t, *args, **kwargs)	Calculate the numerical differentiation.
get_limit_report()	Return a list of dicts describing variables clamped at limits.
get_names()	Available symbols from this class
get_tex_names()	Return tex_names of exported flags.
get_values()
list2array(n)	Allocate memory for storage arrays.
save_unconstrained()	Save the input variable's unconstrained value (call after init pass 1).

Attributes

class_name