Modelling a HEV & EV for a WOT driving cycle using Powertrain & vehicle dynamics blockset.

Objective:

• To understand mapped and dynamic models in powertrain blockset.
• To understand switching of models from mapped to dynamic.
• understand how the model calculate miles per gallon and which factor are considered.
• Run the HEV ReferenceApplication with WOT drive cycle & compare by Changing the grade and wind velocity in the environment block.
• Also compare the simulated results of hybrid and pure electric powertrains.

Theory:

Before going directly to the question let us know about the background of modeling so that we can have a clear understanding.

HEV component modeling types

1. First principles modeling
2. DATA driven modeling
3. Behaviorial modeling

First principle modeling : It refers physical equations of a system. we use simulink function blocks to actual represent our system related equations. eg. vehicle dynamics equation modeling.

Used in :

• system equation are already known or can be derived.
• need a calculator to solve the equation and process results.

Tool used: [MATLAB +SIMULINK]

Data driven modeling: steady state mapping implemented via 'lookup tabels"acn be added dynamics with transfer function.combines pre-made, detailed customizable modelfor power train and vehicle dynamics components. connects to unreal engine for virtual enviornment and sensing.

Used in :

• 1st principle or physics difficult
• Detail dynamics not necessary
• Need fast simulating results

Tool used: [POWERTRAIN BLOCKSET +VEHICLE DYNAMICS BLOCKSET] SIMULINK FAETURE

Behavioural modeling:It donot model actual physics but model procedures similar dynamics response. eg. simscape models.

Used in :

• Model 1D mechanism(motor, actuator) and multidomain system(mechtronics, fluid thermal)
• Reduse visual complexit and model without deriving equations.

Tool used: [SIMSCAPE +SIMSCAPE MULTIBODY]

Modeling Tools stage wise

• powertrain blockset- (Pre built applications)
• simscape-( Flexible architecture)

POWERTRAIN BLOCKSET

powertrain blockset are contained in the library of simulink which are used for modeling engine subsystem ,batteries, controllers, traction engines, transmission units models. it provides a platform where a standard model Architrcture can be reused throughout the development process.

Features of blockset

• library of building blocks

Figure.1 Powertrain library blocks

• Pre-built reference application

Figure.2 Pre-build reference application examples

The Powertrain Blockset are the set of fully assembled reference applications, including gasoline, diesel, and hybrid and electric vehicle systems, as a starting point for your powertrain model.

what is the Goal?

• To Provide starting point for engineers to build good plant / controller models
• To Provide open and documented models
• To Provide very fast-running models that work with popular HIL systems

Powertrain Blockset provides two types of models of internal combustion engines:

1. Mapped(cartographic) model
2. Dynamic model

Mapped(Cartographic)engines represent motor behavior at a high level in the form of a set of interpolation tables/ 'lookup' tables (maps) - taking into account torque, fuel consumption, mass air flow, exhaust gas temperature, efficiency and emissions depending on the given load and the measured engine speed. Mapped engines represent macro engine behavior as a set of lookup tables

Dynamic engine models decompose engine behaviour into separate component models that take into account engine dynamics, primarily intake air flow and turbocharger dynamics.(see below figure3)

Figure.3 Dynamic block of SI engine with airflow and turbocharger.

You can switch between types of engine models based on your application. Dynamic engine models are suitable for the design of control algorithms, evaluation and diagnostics, depending on the states of the dynamic subsystem, for example, in the development of closed-loop mixture control algorithms. Cartographic models of engines are suitable for analysis and design that do not require dynamic characteristics of the engine subsystem, for example, in the analysis of engine and transmission requirements for fuel economy, emissions and performance compromises.

What is a variant?

It is feature in simulink which allows us to create multiple model varient into a single model. the user can select a varient as their requirments and depth of detailing information.

Figure.4 variant of sub-system

To covert the mapped model to dynamic model.

• click on the model block
• Goto varient manager
• Choose as active choice as dynamic

simulink model calculating miles per gallon as result output for pre-built hybrid models.

For the model for calculating fuel economy is shown below in the figure in liters per 100 km (miles per gallon ) . It takes fuel flow rate, vehicle speed and battery power(for electric system) as input for visualization of the fuel economy.

It takes the speed from the drive cycle, battery power from soc of battery and fuel flow from engine output to calculate the fuel efficiency.

Figure.5 performance calculation block inside visualization block

if we see inside the mask the simulink calculator model look like (see below)

Figure.6 logic for calculation of fuel economy of HEV

the three main inputs to calculate MPG are:

1) Vehicle Speed : Root mean square of the speed is calculated and integrated over the time span to get the avg. distance travelled then the unit conversion from m to km by dividing it by 1000 and furthur in the same block divided by 100 for calculating per 100km distance. (see block 1)

2) Fuel Flow Rate: Fuel flow rate(m^3/sec) integrated over the range of time to get volume of fuel flow(m^3). then a gain of 1000 to convert to liters.(see block 2)

3) Battery Power: the w power of battery is converted to kw by dividing by 1000. As 1 us gallon contain 33.6 Kwh which will convert kw unit equivalent to gallon of fuel furthur dividing it by 3600 for per hr conversion and then gain of 0.003785 to get the unit equivalent battery power in rate of fuel flow i.e. m^3/sec(see block 3)

this is given as an input to the fuel flow calculation therefore the Total fuel flow rate = (fuel flow rate from engine + battery energy equivalent flow rate)

if we divide the total distance per 100 km by liters of fuel flow(liters) then we get fuel economy as L/100km (see block 5)

if we want to get fuel economy in US MPG convert distance to miles by gain of 0.0006213 and volume liters to us gallon by gain of 264.172 (see block 6).

Run the HEV ReferenceApplication with WOT drive cycle. Change the grade and wind velocity in the environment block.

These built in reference models are available in the simulink help section. A quick and good starting point for an engineer to analyze the performance and results.

we are looking at the powertrain + vehicle dynamic blockset for a HEV multomode reference model .

This the how the model looks like when you open the reference model a forward approaching model starting from driving cycle to vehicle dynamic calculation( Figure.7) .

Figure.7 HEV block model in simulink using powertrains and vehicle dynamics

The driving cycle is change to WOT with changes parameters also you can see the drive cycle plot (speed vs time )

figure 8. Drive cycle parameter block and WOT drive cycle plot

If we run the simulation for 100 sec and see the results (below figure. 9) with no change in the enviornment block.

Figure.9 Result plots for HEV with normal condition i.e no change in enviornment condition

below you can see figure. 10 the inside the enviornment block with changed grade value from 0 to 3 and wind velocity from 0 to 7 m/s.

figure. 10 Logic Model inside the enviornment block with change parameters.

Results for the HEV for changed enviornment parameters.(figure.11)

figure. 11 Results for the HEV for changed enviornment parameters.

A we can compare the graphs between the unchanged(C1) and changed(C2) enviornment condition and observe that :

1) With C1 condition the vehicle dynamic speed traces and reaches the nominal speed ( drive cycle) at nearly 25 sec while in the C2 condition the speed reaches with a delay nearly at 40sec . As it shows more time required to reach in the C2 case of increased wind velocity and gradeability.

2) SOC in the case C1 the charge drops from 80 % to nearly 68 % within initial 20 sec as vehicle using motor drivetrain then after that engine drives the vehicle and state regeneration starts during decceleration period which regains the soc till 77%. but in the case C2 the motor only assist the engine as tractive torque required is more in the C2 case the main torque is provided by engine and motor only assist, which leads to gradual fall of SOC to 55 % till 70sec and then regeneration at decceleration regaining the soc to 66%

3) The fuel economy for the case C1 on an avg remains 10-11 MPGE while in the case C2 it drops over a period with avg 8 MPGE.

Compare the simulated results of hybrid and pure electric powertrains.

keeping the drive cycle for EV same as for HEV i.e WOT with same parameters.

figure. 12 change in drive cycle (WOT) for pure EV

Results

figure.13 Results of pure EV plots for no change in enviornment parameters

figure.14 Results of pure EV plots for no change in enviornment parameters figure15. Change in enviornment block parameters

We can see in the graphs that iy takes much more time for a EV to approach the desired speed drops to 74-76 % before reaching the deccleration stage that is high as compared to HEV soc and the fuel economy is much better in the case of EV. The max rpm value is also lower in the case of EV. the time req. to reach the nominal speed(40m/s) in HEV in normal condition is 23 to 24 sec whereas in case of EV it took 28-29 sec .

conclusion:

It is a very fast useful approach of a pre-built model which gives a good starting point for an engineer to analyze the behaviour of EV, HEV,split-HEV ,CI, SI mapped, dynamic varients.It is well suited in situations where understanding the principle is difficult and dynamic behavior is not that necessay needed detail. It provides a platform where a standard model Architrcture can be reused throughout the development process.