Optimization and State Estimation Fundamentals

Explore state and parameter estimation in engineering, focusing on estimating battery state of charge from voltage, current, and temperature measurements using estimators like the Kalman filter.

Discover state space representation as a matrix form for dynamic systems with inputs and outputs, deriving xdot equals Ax plus Bu and y equals Cx plus Du for Kalman filters.

Visualize state space systems with block diagrams, derive x dot = Ax + Bu and y = Cx + Du, and relate observability to state estimation via the observability matrix.

Explore state-space representation with a practical mass–spring–damper model, derive A, B, C, D, and simulate in Matlab/Simulink to connect theory with real-world dynamic systems.

Converts an open-circuit voltage based battery model into a discrete-time state-space form, with state of charge and capacitor voltage as states, enabling Kalman filtering for state-of-charge estimation.

Present a Matlab-based open-circuit voltage-rc battery model in state space, using current input to predict terminal voltage and state of charge, with state-space matrices and parameters.

Explore Kalman filters to estimate states in dynamic systems by fusing model predictions with measurements. Learn how the method handles model uncertainty and measurement noise to produce optimal state estimates.

Explore the conceptual Kalman filter, fusing a priori model predictions and sensor measurements to yield an optimal corrected state estimate with quantified uncertainty.

Explain Kalman filter basics in a state-space framework, detailing covariance matrices P and Q and measurement noise R, and show how the Kalman gain fuses predictions with measurements.

Explore Kalman filter implementation as a predictor-corrector method for linear systems, with extended Kalman filter for nonlinear cases, updating a priori to posteriori estimates via Kalman gain and covariance.

Learn how the extended Kalman filter handles nonlinear systems by linearizing with a jacobian to form the H matrix, enabling prediction and correction, with tuning challenges noted.

Apply an extended Kalman filter in MATLAB to estimate the battery state of charge with a combined model. Generate data, linearize the model, and implement the EKF to track charge.

Learn how to implement the extended Kalman filter in MATLAB for estimating the state of charge, using prediction and correction steps, covariance matrices, and process and measurement uncertainty.