Engine Plant Model Development and Controller Calibration using Powertrain Blockset TM

Transcription

1 Engine Plant Model Development and Controller Calibration using Powertrain Blockset TM Brad Hieb Scott Furry Application Engineering Consulting Services 2017 The MathWorks, Inc. 1

2 Key Take-Away s Engine model parameterization is a very nontrivial task Engine controller calibration is a very non-trivial task MathWorks has tools to help make these two tasks more manageable 2

3 Problem Statement How do I use the Powertrain Blockset engine and controller models for my application so I can: Design engine controls? Perform fuel economy and emissions studies? Create and validate dynamometer test plans? 3

4 What we ll Cover Today Parameterizing a Powertrain Blockset engine model Workflow Example: parameterizing a mapped engine model Calibrating a Powertrain Blockset engine controller Workflow Example: calibrating an engine controller 4

5 What are we Parameterizing and Calibrating? 5

6 What we ll Cover Today Parameterizing a Powertrain Blockset engine model Workflow Example: parameterizing a mapped engine model Calibrating a Powertrain Blockset engine controller Workflow Example: calibrating an engine controller 6

7 Powertrain Blockset Si Mapped Engine Model Contains 2D LUT s for each model output Easy to parameterize Great for system level design and development 7

8 Parameterizing an Engine Model - Workflow Model-Based Calibration Toolbox provides tools for the process: Design of Experiments Creating the Design of Experiments CAE Engine Model Data Modeling Engine Dynamometer Parameter Generation Results Optimal Engine Calibration Parameters Fast, Accurate Engine Model for HIL and System Simulation 8

9 Parameterizing an Engine Model - Workflow Model-Based Calibration Toolbox provides tools for the process: Design of Experiments Creating the Design of Experiments CAE Engine Model Engine Dynamometer Gather the data Data Modeling Parameter Generation Results Optimal Engine Calibration Parameters Fast, Accurate Engine Model for HIL and System Simulation 9

10 Parameterizing an Engine Model - Get the data as calibrated Measurements Air Flow Fuel Flow Exhaust Temp Emissions BSFC Dynamometer Control (Steady State) Speed Cmd Measurements Torque Cmd. Speed Measured Speed Measured Data Logger Engine System (Change operating points, fixed calibration) Crankshaft Dynamometer Torque Measured 10

11 Parameterizing an Engine Model - Workflow Model-Based Calibration Toolbox provides tools for the process: Design of Experiments Creating the Design of Experiments Gather the data CAE Engine Model Data Modeling Engine Dynamometer Fitting response surface models (RSM, statistical) to the data Parameter Generation Results Optimal Engine Calibration Parameters Fast, Accurate Engine Model for HIL and System Simulation 11

12 Parameterizing an Engine Model - Workflow Model-Based Calibration Toolbox provides tools for the process: Design of Experiments Creating the Design of Experiments Gather the data Fitting response surface models Developing engine performance maps from RSM s CAE Engine Model Engine Dynamometer Data Modeling Parameter Generation Results Optimal Engine Calibration Parameters Fast, Accurate Engine Model for HIL and System Simulation 12

13 Parameterizing an Engine Model - Workflow Model-Based Calibration Toolbox provides tools for the process: Design of Experiments Creating the Design of Experiments Gather the data Fitting response surface models Developing engine performance maps CAE Engine Model Engine Dynamometer Data Modeling Parameter Generation Validate the result Results Optimal Engine Calibration Parameters Fast, Accurate Engine Model for HIL and System Simulation 13

14 Launch MBC Toolbox From Apps tab From command line >> mbcmodel 14

15 Launch MBC Toolbox 15

16 Parameterizing a Mapped Engine Model - Importing existing data Mapped engine model workflow: Design of Experiments Importing existing data CAE Engine Model Engine Dynamometer Data Modeling Parameter Generation Results Optimal Engine Calibration Parameters Fast, Accurate Engine Model for HIL and System Simulation 16

17 Import Data - Inspect the data Look for anomalies or gaps Filter data to remove anomalies Add derived quantities and unit conversions Graphical views speed inspection 17

18 Parameterizing a Mapped Engine Model - Fitting response surface models Mapped engine model workflow: Design of Experiments Importing existing data CAE Engine Model Engine Dynamometer Fitting response surface models (RSM, statistical) to the data Data Modeling Parameter Generation Results Optimal Engine Calibration Parameters Fast, Accurate Engine Model for HIL and System Simulation 18

19 Fitting Models to the Data - Generate response surface models Default models automatically fitted to all responses Inspect quality of fit Try out alternatives 19

20 Parameterizing a Mapped Engine Model - Developing engine performance maps Mapped engine model workflow: Design of Experiments Importing existing data Fitting response surface models CAE Engine Model Engine Dynamometer Developing engine performance maps from RSM s Data Modeling Parameter Generation Results Optimal Engine Calibration Parameters Fast, Accurate Engine Model for HIL and System Simulation 20

21 Calibration Generation Tool Fill tables Export cal tables 21

22 Calibration Generation Tool - Generating look up tables 22

23 Calibration Generation Tool - Fill tables Inspect surfaces Adjust table values in extrapolation areas Export to MATLAB, Excel or Cal tool 23

24 Parameterizing a Mapped Engine Model - Export and validate result Mapped engine model workflow: Importing existing data Fitting response surface models Developing engine performance maps Design of Experiments CAE Engine Model Engine Dynamometer Data Modeling Export and validate the result Parameter Generation Results Optimal Engine Calibration Parameters Fast, Accurate Engine Model for HIL and System Simulation 24

25 Export Tables to MATLAB 25

26 Validate the Result 26

27 Validate the Result Accuracy for 1200 sec of FTP75 sim: % diff in FE was 0.31% Run time for 1200 sec of FTP75 sim: PTBS Mapped engine model 28.4 sec GT Power FRM engine model 1449 sec Mapped engine model sim ~51x faster 27

28 Parameterizing a Mapped Engine Model - Summary Mapped engine model workflow: Design of Experiments Importing existing data CAE Engine Model Engine Dynamometer Fitting response surface models (RSM, statistical) to the data Developing engine performance maps from RSM s Data Modeling Parameter Generation Results Validate the result Optimal Engine Calibration Parameters Fast, Accurate Engine Model for HIL and System Simulation 28

29 What we ll Cover Today Parameterizing a Powertrain Blockset engine model Workflow Example: parameterizing a mapped engine model Calibrating a Powertrain Blockset engine controller Workflow Example: calibrating an engine controller 29

30 What are we Parameterizing and Calibrating? 30

31 What are we Calibrating? Throttle area percent (TAP) Wastegate Fraction Lambda Spark Advance Intake Cam Advance Exhaust Cam Retard 31

32 Calibrating Optimal Base Engine Control Tables - Workflow Model-Based Calibration Toolbox provides tools for the process: Design of Experiments Creating the Design of Experiments CAE Engine Model Engine Dynamometer Gather the data Fitting response surface models (RSM, statistical) to the data Data Modeling Calibration Generation Results Developing optimal base calibration tables Optimal Engine Calibration Parameters Fast, Accurate Engine Model for HIL and System Simulation Export calibration to controller 32

33 Calibrating Optimal Base Engine Control Tables - Creating the DoE Optimal base engine control calibration workflow: Design of Experiments Creating the Design of Experiments CAE Engine Model Data Modeling Engine Dynamometer Calibration Generation Results Optimal Engine Calibration Parameters Fast, Accurate Engine Model for HIL and System Simulation 33

34 Calibrating Optimal Base Engine Control Tables - Creating the DoE I/O of Turbocharged Direct-Injection 1.5L DOHC Engine Model with Dual-Independent Continuously Variable Cam Phasing Table Breakpoints Optimal Tables RPM Load ICP ECP Minimum BSFC Objective Exhaust Temperature Turbocharger Speed Knock limit Residual Fraction AFR Spark Advance Waste-gate area TAP Intake Manifold Pressure Auxiliary Table Constraints Optimal Tables 34

35 Calibrating Optimal Base Engine Control Tables - Creating the DoE 35

36 Calibrating Optimal Base Engine Control Tables - Gather the data Optimal base engine control calibration workflow: Design of Experiments Creating the Design of Experiments CAE Engine Model Engine Dynamometer Gather the data Data Modeling Calibration Generation Results Optimal Engine Calibration Parameters Fast, Accurate Engine Model for HIL and System Simulation 36

37 Calibrating an Optimal Base Cal Table - Get the data from CAE engine models 37

38 Calibrating Optimal Base Engine Control Tables - Get the data from calibration sweeps Dynamometer Control (Steady State) Speed Cmd Actuator Cmds. Torque Cmd. Speed Measured Measurements Air Flow Fuel Flow Exhaust Temp Emissions (EO/TP) MAP Measurements MAT A/F Turbo Speed Turbine press ratio Compressor press ratio Speed Measured Turbine temp out Compressor temp out EGR pct. EGR cooler temp out Intercooler temp out Data Logger Actuator Commands Throttle Intake cam Wastegate Exhaust cam Injector EGR valve Spark Turbo Engine (Change operating points and sweep Actuator Cmds) Crankshaft Dynamometer Torque Measured 38

39 Calibrating Optimal Base Engine Control Tables - Fitting response surface models Optimal base engine control calibration workflow: Design of Experiments Creating the Design of Experiments Gather the data Fitting response surface models (RSM, statistical) to the data CAE Engine Model Engine Dynamometer Data Modeling Calibration Generation Results Optimal Engine Calibration Parameters Fast, Accurate Engine Model for HIL and System Simulation 39

40 Calibrating Optimal Base Engine Control Tables - Generate response surface models from data Default models automatically fitted to all responses Inspect quality of fit Try out alternatives 40

41 Calibrating Optimal Base Engine Control Tables - Develop optimal base calibration tables Optimal base engine control calibration workflow: Design of Experiments Creating the Design of Experiments Gather the data Fitting response surface models Developing optimal base calibration tables from RSMs CAE Engine Model Engine Dynamometer Data Modeling Calibration Generation Results Optimal Engine Calibration Parameters Fast, Accurate Engine Model for HIL and System Simulation 41

42 Calibrating Optimal Base Engine Control Tables - Developing calibration tables Import response surface models Run optimizations Analyze tradeoffs and sensitivity Fill tables Export cal tables 42

43 Calibrating Optimal Base Engine Control Tables - Developing calibrations from response surface models Import response surface models Run optimizations Analyze tradeoffs and sensitivity Fill tables Export cal tables 43

44 Calibrating Optimal Base Engine Control Tables - Run optimizations Define objective Define constraints Determine operating point weights 44

45 How to calculate the weights for a sum optimization Use MATLAB to calculate weights for a drive cycle Drive cycle data Weights as % from total time MATLAB program reads measurements from Excel measurement file and calculate weights automatically Histogram 45

46 Calibrating Optimal Base Engine Control Tables - Run optimizations Evaluate optimization results Diagnose optimization convergence issues 46

47 Calibrating Optimal Base Engine Control Tables - Analyze tradeoffs and sensitivity Evaluate local sensitivity Determine if tradeoffs are needed 47

48 Calibration Generation Tool - Fill tables Inspect surfaces Export to MATLAB, Excel or Cal tool 48

49 Optimal Base Calibrations Completed Throttle area percent (TAP) Wastegate Fraction Lambda Spark Advance Intake Cam Advance Exhaust Cam Retard 49

50 Calibrating Optimal Base Engine Control Tables - Export and validate the result Optimal base engine control calibration workflow: Design of Experiments Creating the Design of Experiments Gather the data Fitting response surface models Developing optimal base calibrations CAE Engine Model Engine Dynamometer Data Modeling Calibration Generation Export calibration to controller Results Optimal Engine Calibration Parameters Fast, Accurate Engine Model for HIL and System Simulation 50

51 Export Tables to Calibration Tool 51

52 Calibrating Optimal Base Engine Control Tables - Summary Optimal base engine control calibration workflow: Design of Experiments Creating the Design of Experiments CAE Engine Model Engine Dynamometer Gather the data Fitting response surface models Developing optimal base calibrations Export calibration to controller Optimal Engine Calibration Parameters Data Modeling Calibration Generation Results Fast, Accurate Engine Model for HIL and System Simulation 52

53 Key Take-Away s Engine model parameterization is a very nontrivial task Engine controller calibration is a very non-trivial task MathWorks has tools to help make these two tasks more manageable 53

54 Contact us to Learn More Scott Furry Brad Hieb Design of Experiments CAE Engine Model Engine Dynamometer Data Modeling Calibration Generation Results Optimal Engine Calibration Parameters Fast, Accurate Engine Model for HIL and System Simulation 54

55 Q & A?? 55