statistics – Is there a way to limit the level / trend variance to a certain value in the component model not observed in python?

I build an unobserved component model using the & lt; trend & # 39; method.

from statsmodels.tsa.statespace.structural import UnobservedComponents

ucm_model = UnobservedComponents (endog, exog = exog, level = lltrend, mle_regression = True)
result = ucm_model.fit ()
print (result.summary ())

Now, I want to limit the level and variance of the trend to a certain range to make the series smoother and avoid oversampling. The 'strend & # 39; performs a similar job by setting the trend variance to zero. Is there a way to specify and set the level / trend variance on a different (non-zero) value?