Public Member Functions
	__init__ (self, int sensor_id, np.ndarray prior_mean, np.ndarray prior_cov)
Tuple[np.ndarray, np.ndarray]	recursive_update (self, np.ndarray design_vector, float observation, float observation_variance)
Tuple[float, float]	predict_with_uncertainty (self, np.ndarray design_vector)
float	compute_quality_score (self)

Public Attributes
	sensor_id = sensor_id
	posterior_mean = prior_mean.copy()
float	posterior_cov = prior_cov.copy()
	chol_factor = cho_factor(prior_cov, lower=True)
int	n_observations = 0
deque	residual_history = deque(maxlen=100)

Detailed Description

Online Bayesian learning with recursive updates.
Continuously improves calibration as new data arrives.

Definition at line 413 of file autonomous_calibration_engine.py.

Constructor & Destructor Documentation

◆ init()

autonomous_calibration_engine.OnlineBayesianLearner.__init__	(		self,
		int	sensor_id,
		np.ndarray	prior_mean,
		np.ndarray	prior_cov )

Definition at line 419 of file autonomous_calibration_engine.py.

Member Function Documentation

◆ compute_quality_score()

float autonomous_calibration_engine.OnlineBayesianLearner.compute_quality_score ( self )

Compute calibration quality score (0-1).
Based on normalized residuals and sample size.

Definition at line 498 of file autonomous_calibration_engine.py.

◆ predict_with_uncertainty()

Tuple[float, float] autonomous_calibration_engine.OnlineBayesianLearner.predict_with_uncertainty	(		self,
		np.ndarray	design_vector )

Fast prediction with uncertainty

Definition at line 487 of file autonomous_calibration_engine.py.

◆ recursive_update()

Tuple[np.ndarray, np.ndarray] autonomous_calibration_engine.OnlineBayesianLearner.recursive_update	(		self,
		np.ndarray	design_vector,
		float	observation,
		float	observation_variance )

Recursive Bayesian update (Kalman filter form).

Much more efficient than recomputing from scratch.

Args:
    design_vector: φ(v, e) - basis functions
    observation: Measured pressure
    observation_variance: σ² - measurement uncertainty

Returns: (updated_mean, updated_covariance)

Definition at line 434 of file autonomous_calibration_engine.py.