Monte Carlo Predictor
meganlim edited this page Jul 17, 2018
·
8 revisions
The Monte Carlo Predictor implements a Monte Carlo sampling-based prediction algorithm. Given a system model and specified parameters, the algorithm simulates several realizations of the system model for the given inputs and computes predicted event times and the trajectories of certain system variables during the prediction horizon.
The Monte Carlo Predictor is configured by specifying the system model and algorithm parameters including process noise variance values and input uncertainty values, a number of samples, and a prediction horizon. Then, given the time of prediction and an estimate of the current system state, the predict member function can execute the prediction algorithm.
| Parameter | Description |
|---|---|
| Predictor.numSamples [Required] | The number of samples used by a predictor. |
| Predictor.horizon [Required] | The time horizon for prediction (in seconds), which is relative to the current time. |
| Predictor.loadEstimator [Optional] | The load estimator to use, the default is Floating Average Load Estimator |