Documentation

Execute Engine Mapping Experiment

Assess engine torque, fuel flow, and emission performance results using an existing engine controller calibration.

Dynamometer controller commands a series of engine speeds and torques to the engine controller. At each quasi-steady-state operating point, the experiment records the engine plant model output and the controller commands for the current calibration parameters.

Execute Model Predictive Control Plant Model Experiment

Generate transient engine datasets for linear plant models useful for model predictive controllers.

Dynamometer controller commands engine speed and torque dynamically as a function of time using a pseudo random binary sequence. Experiment records the transient engine torque, temperature, airflow, and emission responses determined from linear dynamic plant model fitting via system identification.

Recalibrate Controller

Match measured engine torque to commanded engine torque across engine operating range.

Dynamometer controller generates a feedforward throttle table by matching the measured engine torque to the commanded engine torque across the engine operating range.

Resize Engine and Recalibrate Controller

Match engine torque to desired engine power and number of cylinders.

Dynamometer resizes the dynamic engine and engine calibration parameters. Also, the dynamometer recalibrates the controller and mapped engine model to match the resized dynamic engine.

For an example, see Resize the SI Engine.

Generate Mapped Engine from Spreadsheet

Generate a mapped engine calibration from a data spreadsheet. Update the mapped engine with the calibrated data.

Dynamometer uses the Model-Based Calibration Toolbox™ to fit data from a spreadsheet, generate calibrated tables, and update the mapped engine parameters.

For an example, see Generate Mapped SI Engine from a Spreadsheet.

Generate Deep Learning Engine Model

Train a deep learning model of dynamic engine behavior from measured laboratory data or a high-fidelity engine model.

Dynamometer uses the Deep Learning Toolbox™ and Statistics and Machine Learning Toolbox™ to generate a dynamic deep learning engine model and update the mapped engine parameters.

For an example, see Generate a Deep Learning SI Engine Model.