ROVI-org / auto-soh / 10496015336

95%

Build Type

push

github

Committed by web-flow

Commit Message

Dual estimation framework done (#71)

* Defined DegradationModelInterface

The wrapper needed to interact with the A-SOH only filter in the dual estimation
framework is prepared. It does not take a degradation model yet, but that is to
come.

* Just creating the place where dual framework lives

* Changing name for consistency

* Finalized orignal code for cell and degradation specific wrappers

Completed the minimum necessary wrappers for dual estimation framework.

* Making it clearer means have to be flattened

Multivariate random variables require flat mean arrays. Here, we just make sure
that is the case.

* Fixing bug due to change in argument name

* Adding a way for multivariate random variables to be combined

Included the requirement that multivariate random variables can be easily
combined if they are of the same type (that is, of the same distribution, but
not necessarily equal PDFs, or even equal dimensions!)

* CellModel does not need to know which parts of A-SOH are updatable

* Finished first pass of full Dual framework implementation

Dual estimation framework fully implemented, but not yet tested. It also provides
a dual UKF to facilitate use.

* Changing order in which arguments are passed

Initialization of UKF should take critical parameters first, important parameters
second, and tuning/hyper- parameters last. That has been fixed.

* Making some attributes cached properties

This way, it is easier for initialization with incomplete set of parameters can
can be re-set post initialization, but prior to estimation steps.

* Facilitating type-hinting for dual UKF initialization

* Fixing bug in A-SOH initialization

A-SOH initialization to multivariate gaussian in the provide UKF helper function
had a bug: it passed the wrong argument to the function. That has been fixed.

* Fixes A-SOH values not being passed as np.ndarray

* Fixes bug related to returning None

Operations being... (continued)

Run Details

170 of 178 new or added lines in 6 files covered. (95.51%)

7 existing lines in 1 file now uncovered.

1067 of 1129 relevant lines covered (94.51%)

0.95 hits per line