Augmenting Extended Kalman Filters with Dynamic Temperature Derivative State Models
Augmenting Extended Kalman Filters with dynamic temperature derivative states eliminates dynamic thermal bias drift during rapid ramp conditions.
Observability
Estimation Feasibility
A measure of how well the internal states of a dynamical system can be inferred from knowledge of its external outputs provides the foundation for control system design. When a system possesses the property of observability, a state estimator can uniquely determine the complete internal condition of the system from a finite sequence of sensor measurements.
Mathematical Verification
Control equations and system dynamics are used to construct a specialized matrix to test this system property. If this test matrix has full rank, the system is mathematically proven to be observable. This binary check ensures that the selected sensor configuration provides sufficient information to reconstruct the state vector of the machine.
Sensor Placement
The selection and physical location of measurement devices directly influence whether a system achieves observability. In complex chemical reactors or power grids, placing sensors at strategic points allows the monitoring of critical variables that cannot be measured directly. This indirect monitoring relies on the mathematical relationships between the measured outputs and the unmeasured states.
Optimizing the sensor distribution across a large network ensures that all state variables can be estimated with minimal noise propagation and high reliability.
Practical Limitation
Noise in the measurements and inaccuracies in the system model can degrade the practical reconstructability of the states even when the system is theoretically observable. In such scenarios, the estimation filter requires a longer sequence of data to converge on the true state values. This sensitivity to noise dictates the required precision of the sensing hardware.