Patent Application
20210103688
The present invention relates to systems for modelling drivelines, and in particular for generating driveline efficiency metrics.
US2015/0347670 A1 (JAMES) discloses an approach for calculating driveline efficiency. However an appropriate component efficiency map is not generated as and when required, it is not immediately available for the driveline efficiency processor. This also means that the user may have to spend time importing a map from another source or modelling the efficiency in another software package. The efficiency map would have to be calculated elsewhere and imported. An efficiency map generally has limits in terms of torque and speed for a given gearbox and motor design, and if the operating speed range is changed the motor/gearbox has to be redesigned, which changes the efficiency map. Since the component efficiency maps are not all generated by the same processors, they may not be compatible with each other.
The present invention provides this functionality in a single modelling system.
According to a first aspect of the invention, there is provided a system for modelling a driveline, wherein the driveline comprises a plurality of components, the system comprising:
The driveline-efficiency-processor may be configured to generate one or more additional driveline-metrics based on the received corresponding component model. The one or more additional driveline-metrics may be representative of one or more of the following characteristics: packaging size, power rating, durability, driveability, noise and vibration characteristics.
The component-efficiency-processor may be configured to generate the component-efficiency-map based on a component-detail-level.
The component-efficiency-processor may be configured to:
The component models may comprise component-form-information representative of physical dimensions of the associated component. The driveline-efficiency-processor may be further configured to:
The driveline-efficiency-processor may be further configured to:
The driveline-efficiency-processor may be further configured to generate the driveline-efficiency-metric for the driveline based on a time-varying mass of a vehicle associated with the driveline over the one or more driving-profiles.
The driving-profile may comprise information about the time-varying mass of the vehicle.
The driveline-efficiency-processor may be further configured to generate the driveline-efficiency-metric for the driveline based on a variable gradient associated with the one or more driving-profiles.
The driving-profile may comprise information about the variable gradient.
The driveline-efficiency-processor may be configured to apply a tractive force equation when generating the driveline-efficiency-metric. The tractive force equation may be configured to determine tractive force required for a time-varying mass and/or a variable gradient associated with the one or more driving-profiles.
The driveline-efficiency-processor may be further configured to:
The plurality of drivelines may have variations in one or more or the following inputs:
wherein each combination of inputs represents a different driveline.
The driveline-efficiency-processor may be further configured to:
According to a further aspect of the invention, there is provided a driveline-efficiency-processor for a driveline comprising a plurality of components, wherein the driveline can be operated according to a plurality of control-states-of-operation, wherein the driveline-efficiency-processor comprises:
an analysis-block configured to:
a control-strategy-application-block configured to:
The matrix for each control-state-of-operation in the set of operational-matrices may comprise information about the efficiency of an associated component for a plurality of vehicle operational requirements (such as speed and acceleration values).
The control-strategy-application-block may be configured to:
Each control-state-map may define one or more switchover-thresholds between different control-states-of-operation. The control-strategy-application-block may be configured to determine a revised latest-control-state-map by modifying the switchover-threshold(s).
The control-strategy-application-block may be configured to determine the latest control-state-map by:
The control-strategy-application-block may be configured to determine the latest-control-state-map at step a) using initial-component-efficiency-values.
The control-strategy-application-block may be configured to calculate the driveline-efficiency-metric and control-state-map using an iterative process:
The control-states-of-operation may comprise one or more of:
According to a further aspect of the invention, there is provided a method of modelling a driveline, wherein the driveline comprises a plurality of components, the method comprising:
According to a further aspect of the invention, there is provided a method of processing for a driveline, wherein the driveline comprises a plurality of components, and wherein the driveline can be operated according to a plurality of control-states-of-operation, wherein the method comprises:
There may be provided a computer program, which when run on a computer, causes the computer to configure any apparatus, including a processor, controller or device disclosed herein or perform any method disclosed herein. The computer program may be a software implementation, and the computer may be considered as any appropriate hardware, including a digital signal processor, a microcontroller, and an implementation in read only memory (ROM), erasable programmable read only memory (EPROM) or electronically erasable programmable read only memory (EEPROM), as non-limiting examples.
The computer program may be provided on a computer readable medium, which may be a physical computer readable medium such as a disc or a memory device, or may be embodied as a transient signal. Such a transient signal may be a network download, including an internet download.
Embodiments of the present invention will now be described by way of example and with reference to the accompanying drawings in which:
The driveline includes a plurality of components, non-limiting examples of which include, for example, internal combustion engines, fuel cells, gas turbines, gearboxes, electric machines, shafts, housings, bearings, clutches, terrestrial/aerospace/marine vehicle chassis, wheels, flywheels, batteries, capacitors, and power electronics. The components can be sub-assemblies of components, or individual components.
The system of
An efficiency map for a gearbox can be generated by calculating the gearbox power losses over a range of speeds and torques. The main sources of power loss in a gearbox can include gear mesh losses due to sliding friction between the gear teeth, gear churning losses due to splashing of the lubricant, and bearing losses. These power losses can be calculated using, for example, the methods defined in ISO standard 14179.
An efficiency map for an electric machine can be generated by calculating the electric machine power losses over a range of speeds and torques. The main sources of power loss can include copper losses due to electrical resistance in the machine windings, iron losses due to hysteresis and eddy currents, and mechanical losses due to bearing friction and windage.
An efficiency map may be a flat map, in that the efficiency values may be substantially constant for a range of operating conditions.
A component-efficiency-processor 104a, 104b can be configured to generate one or more component-metrics in addition to the component-efficiency-map, by optimising the component for one or more other design targets in addition to efficiency, including but not limited to packaging space, vibration characteristics, and durability. For example, a gearbox component-efficiency-processor could optimise the gearbox layout (number, position, and dimensions of shafts, bearings, gears), macro-geometry, and micro-geometry. In this way, one or more component-metrics can be generated by the component-efficiency-processor 104a, 104b.
The system of
There are several advantages of the system of
The potential design space for hybrid vehicles is very large, with a wide range of possible driveline-layouts across the entire spectrum of electrification from pure conventional to pure electric powered vehicles, with many options in between. There are many degrees of freedom in the design and control of the drivelines and components. A rapid whole-system simulation, such as the system of
Traditional vehicle simulations can be computationally intensive, often preventing a full exploration of all of the candidate drivelines in the design space.
The system of
and the resulting driveline-metrics for each candidate driveline can be used to select a driveline from the set of candidates, considering one or more of a range of performance targets.
In addition, the same driveline can be processed with a plurality of different control parameters, which also increases the number of simulations which must be run in order to investigate all options.
Rapid simulation is therefore advantageous in that it enables wider exploration of the design space, considering all of the options for driveline and component design and control.
The driveline-layout 110 can define how the components in the driveline engage/interact with each other. In some examples, the driveline-layout 110 received by the driveline-efficiency-processor 106 can provide information that is used in combination with information stored within the driveline-efficiency-processor 106 to determine how the components in the driveline engage/interact with each other. For example, the driveline-layout 110 may provide an identifier for each component in the driveline, which the driveline-efficiency-processor 106 can use to retrieve an associated component-efficiency-map from a plurality of maps received from component-efficiency-processors 104a, b. The order in which the plurality of components is connected in the driveline can be defined in the driveline layout.
In some examples, the component models 102a, 102b include component-form-information, which can be representative of physical dimensions of the associated component and/or a material from which the component is made. That is, the component models 102a, 102b may include more than just functional definitions of the component.
As shown in
Advantageously, the system of
In some examples, the component-efficiency-processors 104a, 104b can generate the component-efficiency-map based on a component-detail-level 114. Use of such a component-detail-level 114 can be used to set an appropriate balance between (i) speed of processing, and (ii) precision of the component-efficiency-map that is generated by the component-efficiency-processors 104a, 104b. The component-efficiency-processors 104a, 104b can make different physical or mathematical assumptions about the component efficiency model when generating the component-efficiency-map based on the component-detail-level 114, or change the number of points used for generating the component-efficiency-map based on the component-detail-level 114, for example the number of points in the component-efficiency-map (number of speed points multiplied by number of torque points) at which the efficiency is calculated.
The driveline-efficiency-processor 106 can also receive a driveline-efficiency-processor-detail-level 116. Similarly to the component-detail-level 114 for the component-efficiency-processors, the driveline-efficiency-processor-detail-level 116 can be used to set an appropriate balance between (i) speed of processing, and (ii) precision of the driveline-efficiency-metric that is generated by the driveline-efficiency-processor 106. The driveline-efficiency-processor-detail-level 116 can define the number of vehicle operating points (e.g. speed and acceleration) to be considered, and/or the number of power-split-modes-of-operation to be considered in the control strategy, as will be discussed later.
In the example of
In some examples, advantageously, the driveline-efficiency-processor 106 can convert a speed-time driving-profile 112 into a speed-acceleration profile. It will be appreciated that such processing changes the representation of the driving-profile 112 from the time domain into the statistical domain. The representation in the statistical domain is an example of a representation of the driving cycle in a non-time-varying form.
It is important to consider real-world driving conditions early in the design process. If a vehicle is designed and optimised for one driving-profile, the vehicle will perform less well under different driving conditions. Time domain modelling can be slow; each timestep is analysed in turn, so the simulation time is proportional to driving-profile duration. Adding more driving-profiles to a time-domain simulation proportionally increases the required computation time. A limited consideration of driving-profiles can lead to solutions that are not robust to variations in driving style.
In examples where the driving-profile is converted into the statistical domain, the driveline-efficiency-processor can perform a single calculation over the speed-acceleration operation space, rather than a separate calculation for each timestep. Thus there is no time penalty for processing a greater number of driving-profiles, and the method can be made more representative of real-world driving by including a wider range of driving-profiles.
In some cases, the driving-profile can include a gradient, which may be a variable/time-varying gradient, which can be incorporated into the simulation to include the effects of a vehicle going up- or downhill. This can be achieved by adding a term to the tractive force equation to represent the component of the gravitational force in the direction of the slope
In road vehicles, the tractive force is the force required at the wheels in order to accelerate the vehicle to meet the driving-profile and to overcome drag forces (which can include air resistance, wheel friction, and road gradient),
F
tractive
=F
acceleration
+F
drag
+F
gradient
One formulation of the tractive force equation is:
F
tractive
=m a(t)+k1 v(t)2+k2 m g cos θ(t)+m g sin θ(t)
where m is the vehicle mass, a(t) is acceleration, k1 and k2 are constants, v(t) is speed, g is the gravitational constant, θ(t) is the road gradient (angle from the horizontal), and (t) indicates that the variable is a function of time.
For small values of θ, the approximation cos θ≈1 can be made. The first term in the equation is the force required to meet the acceleration requirements of the driving-profile, the second term represents air resistance, the third term represents rolling resistance, and the fourth term represents driving up- or downhill. A positive value of θ represents driving uphill, a negative value of θ represents driving downhill, and θ=0 represents driving on a horizontal road.
When the simulation is in the statistical domain, the time-varying gravitational force component (m g sin θ(t)) can be included in the acceleration value of the driving-profile operating points by taking the acceleration values a′ as a′=a(t)+g sin θ(t).
The definition of the driving-profile can include time-varying mass (i.e. a driving-profile that defines mass as a function of time, as well as speed as a function of time). In the statistical domain, time-varying mass can be achieved via a similar method to including road gradients in drive cycles. The tractive force equation can be refactored so that time-varying mass is included in the definition of the non-time-varying acceleration matrix, so the method can be implemented with only minor changes to the definition of driving-profile.
The accelerative force Facceleration required to fulfil the acceleration requirements of the driving-profile is:
F
acceleration
=m(t) a(t).
This equation can be refactored to separate out the time dependency:
F
acceleration
=m
0
a m(t)/m0
where m0 is a constant mass, and m(t)/m0 is a factor that represents the deviation from the constant mass through the driving-profile.
The acceleration values a′ of the operating points are given by a′=a(t) m(t)/m0. If the simulation represents a vehicle that has time-varying mass and the driving-profile has a road gradient, the acceleration values a′ of the operating points are given by a′=a(t) m(t)/m0+g sin θ(t). This method removes the time-dependency from the tractive force equation, and places it into the definition of the driving-profile. The time-variance of the road gradient and/or vehicle mass is therefore accounted for in the driving-profile definition, enabling the tractive force equation to be applied to the driving-profile operating points in the statistical domain.
Applications of vehicle simulation involving time-varying mass can include, but are not limited to, the following:
It will be appreciated that a time-varying mass and/or a time-varying gradient may be represented in the statistical domain, and that such a representation can be considered as a non-time-varying representation of a time-varying mass and/or time-varying gradient.
A component-efficiency-map is shown associated with some of the main components in the driveline. For example, an engine-efficiency-map 214 is shown associated with the engine 202. The component-efficiency-maps associate component efficiency values with a plurality of component-operating-conditions of the component. The specific component-operating-conditions can vary from component to component. For example, for the engine 202 the component-operating-conditions can be speed and torque, and for the battery the component-operating-conditions can be power and state of charge.
A control strategy must be defined in order to determine how the driveline operates for specific vehicle operational requirement (such as speed and acceleration). Control parameters are associated with the control strategy. The main aspects of the control strategy are listed in the table below, along with associated control parameters. The values of one or more control parameters can together be referred to as a control-state-map. These are discussed in more detail later in this document.
Power-threshold-lines and gear-threshold-lines are examples of switchover-thresholds.
In the example in
When the vehicle is in the hybrid-mode-of-operation 234, the power demand at the wheels is provided by more than one power source. In a hybrid electric vehicle, the power sources can be an engine and one or more electric motors. Different methods may be used to determine the power-split-mode-of-operation, i.e. how the power demand is divided between the different power sources.
For example, consider a driveline with two power sources, an engine and an electric motor. The table below enumerates some different control strategies that may be used to determine the engine output power, along with the associated control parameters. The required power output of the electric motor can then be calculated by subtracting the engine power from the total power demand. Note that the power output of the electric machine can be negative (therefore generating electricity) if the power output of the engine exceeds the power demand.
Returning to
The driveline 200 of
In some examples, for each preceding component in the driveline, the simulation can determine one or more of the following for each component:
In this example, the simulation tracks back from the wheels 404 to a gearbox 406 to an internal combustion engine 408 T (torque) and n (speed) values are identified in
The backwards simulation method can be simpler and faster than the forwards simulation of
The analysis-block 704 can implement the process described in
The analysis-block 704 generates a set of operational-matrices 712 as an output. The set of operational-matrices 712 provides information about the efficiency of the components in the driveline for each control-state-of-operation, for a plurality of vehicle operational requirements (such as speed and acceleration values).
The control-strategy-application-block 706 processes the operational-matrices 712, initial-control-parameter-values 716, a received driving-profile 718, and initial-component-efficiency-values 717 in order to generate a suitable control-state-map 720 for the driving-profile 718. This can involve iteratively calculating the efficiency of the driveline for a plurality of different control-state-maps.
At step 802, the overall system efficiency is calculated for each control-state-of-operation (for example, this could be for each gear ratio and for each propulsion mode).
One method of calculating system efficiency is illustrated in
Some powers can be either inputs or outputs depending on the direction of the powerflow.
Calculating equivalent-propulsion-power requires knowing the component efficiency values over the driving-profile, in order to “track” power from one place in the drivetrain to another (see
Returning to
At step 804, the gear ratio for each speed-acceleration operating point is chosen. In some examples, this choice can be determined by which ratio gives the best overall system efficiency. For example, for each and every speed-acceleration operating point, the system efficiency for each gear-mode-of-operation is compared, and the gear-mode-of-operation with the best system efficiency is selected as a preferred gear-mode-of-operation for the speed-acceleration operating point. Once all of the preferred gear-modes-of-operation have been selected, the values for the values for the gear-threshold-lines between consecutive gear ratios can be determined. In this way, the control-strategy-application-block effectively determines a gear-shift-map similar to that illustrated in
At step 806, a propulsion-mode-map is set, similar to that illustrated in
The value of the power-threshold-line is then changed in order to bring the net-battery-charge-increase of the next iteration of sub-loop 808 closer to the target value. After several iterations, the net-battery-charge-increase value will converge. In this way, the sub-loop 808 determines a net-battery-charge-increase value over the driving-profile for a plurality of control-state-maps; compares the net-battery-charge-increase values with each other or a predetermined threshold; and based on the comparison, selects one of the plurality of control-state-maps as for further processing.
In some examples, it can be advantageous for the battery charge level to be balanced over a driving-profile. In such an example, the processing at step 808 can result in a propulsion-mode-map for which the net-battery-charge-increase value is close to zero.
At step 810, the gear-shift-map identified at step 804 and the propulsion-mode-map selected at step 806 are combined to provide a single control-state-map
At step 812, the component efficiency values are updated. The new component efficiencies are calculated based on the driving-profile, the control-state-map determined at step 810, and on the operational-matrices calculated by the analysis-block 704 in
After step 812, for each iteration of the loop after the first, the method determines whether or not each of the latest component efficiency values are acceptable and satisfy a predetermined criterion, for example whether or not they have converged to an acceptable extent. An acceptable-tolerance-value can be used to define whether or not the component efficiency values have converged to an acceptable extent. Such an acceptable-tolerance-value is also an example of a driveline-efficiency-processor-detail-level 116. That is, a high acceptable-tolerance-value gives fewer iterations of the loop and therefore finds the answer faster, a smaller acceptable-tolerance-value will result in more iterations but a more accurate result over all.
The result of the determination at step 812 is shown schematically by the split in the arrows 813 in
If the efficiency values for each component have not converged following the processing at step 812, then the method returns to step 802 where the overall system efficiency is calculated for each control-state-of-operation, but this time using the component-efficiency-values calculated at step 812 instead of the initial-component-efficiency-values 818.
The method of
At step 502, a model is initialised. Such initialisation can include defining the components to be used in the drivetrain, how they are connected, the appropriate values of component parameters and component efficiency maps, and appropriate values of control parameters. At step 504, a simulation of a driving-profile is run for the model that was initialised at step 502. In a first iteration, the simulation can be run for an initial control-state-map, which defines how the driveline is controlled for specific vehicle-operational-requirements (such as speed and acceleration values as defined by the driving-profile).
At step 506, the results of the simulation are calculated in order to generate data signals, which are output at step 508. The data signals are indicative of the efficiency of the driveline, when controlled in accordance with the control-state-map that was used at step 504. The data signals can also be indicative of the control-state-map that was used in the previous iteration of the simulation.
At step, 510, a user manually adjusts the control-state-map with a view to improving/optimising performance. This can be with the intention of improving efficiency. Then, with the adjusted control-state-map, the method returns to step 502 to initialise the new model, before a new simulation is run at step 504.
The method of
At step 602, a model is initialised, which can be similar to step 502.
At step 604, the method analyses the drivetrain for all control states. This step is carried out by the analysis-block 704, following the process defined in
The output of step 604 can be a set of operational-matrices for each component for each control state of the driveline. The operational-matrices contain component values (for example, speeds, torques, and/or power values) as a function of vehicle speed and acceleration. Because step 604 involves the calculation of matrices for every drivetrain control state, the optimisation loop (606, 608, 612) simply selects the optimum combination of the operational-matrices which have already been calculated, in order to attain the best overall system efficiency (as described in
At step 606, the method applies an initial control-state-map. The initial propulsion-mode-map is determined by choosing an initial threshold power value. This defines which of a plurality of propulsion modes to use as a function of the vehicle speed and acceleration. In some examples, the selection of which gear ratio (corresponding to one of a plurality of available gear ratios) should be used, is determined based on which ratio gives the best overall system efficiency. At step 608, the method involves calculating the overall-equivalent-system-efficiency for the control-state-map that was applied at step 606. For the first iteration of the loop this generates an initial driveline efficiency value. For subsequent iterations that apply additional control-state-maps, this generates additional driveline efficiency values. Step 608 also includes comparing the results to one or more targets. As discussed above, such targets can include convergent component efficiency values, and optionally a target net-battery-charge-increase over the driving-profile.
At step 612, the method then includes applying an automatic optimisation loop for retuning to step 606 to recalculate and apply a different control-state-map. The 606-608-612 loop in
This automatic optimisation loop then continues until the targets at step 608 are satisfied, at which time output data signals are provided at step 610. The output data signals can be representative of a control-state-map that defines which of the control-states-of-operation should be applied for a range of vehicle speed-acceleration values, and a system-efficiency value associated with vehicle operation using the control-state-map for the drive cycle.
Notably, step 604 of calculating operational-matrices for all components and all control-states-of-operation, which can be relatively processing-intensive, is outside of the optimisation loop 612. This enables the method of generating a control-state-map to be very efficient in terms of the amount of processing that is required, and also the time taken to optimise the control-state-map.
| Number | Date | Country | Kind |
|---|---|---|---|
| 1601849.1 | Feb 2016 | GB | national |
| 1606813.2 | Apr 2016 | GB | national |
| Filing Document | Filing Date | Country | Kind |
|---|---|---|---|
| PCT/IB2017/050541 | 2/1/2017 | WO | 00 |
