The error compensation is updated online via an error codebook design where an error tolerant and stable solution is obtained.
The underlying correction term depends on the output observation error and is updated in a mixed continuous-discrete fashion.
Error handling is updated as described above.
After the IEKF iterations are completed at l = L, the corrector's error covariance matrix is updated as P k | k = P k | k - 1 - K ( L ) Φ ( L ) P k | k - 1. (20).
Then, a high gain observer that runs in continuous-time with an output error correction time that is updated in a mixed continuous-discrete fashion is designed in order to estimate the liquid levels in the two upper tanks.
If the flow suffers from errors, its virtual time is updated to contend for the next forwarded chance.
To include the effect of Doppler errors, the signal model is updated as follows: R k Δ f d = A b k, l b exp - j 2 πΔ f d k T c - Δφ sinc ( πΔ f d T c ) + n I, Q, k = 1, 2, …, NM (3).
The accuracy of these predictions is monitored by the event comprehension system, and when a transient spike in prediction error occurs the event model is updated.
At each iteration, the training data are reweighted according to the current prediction error (loss), and the function is updated according to an optimization rule.
