Status report on research and validation programs concerning:

• HILS Model Library
• WHVC Test Cycle
• System Work Concept
• Rated Power Determination for a Hybrid Vehicle
• Family Concept-Validity of Certification
• Validation Test Phase 2 (VTP2) Validation Results−Sub-presentations of VOLVO, MAN and IVECO
• VTP2 conclusions & discussion of further actions

Presentation of the results of the validation test program 1 and status of test program 2 (Chassis dyno test program at JRC, HDH drive cycle investigation, and HILS/SILS model verification) and review of open issues for drafting the GTR text.

Mr. Silberholz gave a brief summary of the work achieved during validation test program 1 (pages 3 to 10 of HDH-15-04). He reported that extensive changes on models and model structure have been made. The new structure is component library based and enabled implementation of a flexible signal data bus, which allows adding more signals on the data bus. There was positive feedback from and good cooperation with the OEMs. The next model release is planned for the end of October, and would need to be tested again by the OEMs.

It was agreed that the additional testing of the models should not be a show-stopper for the drafting work. Mr. Silberholz confirmed that the models are described well enough for the drafting process, and could be inserted in their current stage.

Mr. Six presented a detailed analysis of the HDH drive cycle (WHVC with road gradients) on pages 11 to 33 of HDH-15-04. Parts of the investigations were done in conjunction with the chassis dyno tests at JRC. He concluded that the minicycle approach very well aligns WHVC and WHTC. This is considered a solid basis for the new test cycle. The approach considers added and removed payloads for positive and negative work, thereby matching of positive WHTC work independent of test weight (i.e. slopes will adjust work). A great benefit is that negative work is independent of test weight, so that adequate recuperation energy is always available. This better reflects real world operation for different payloads. A summary of the methods analyzed (minicycle, 30 sec average, individual slope, fixed slope) is shown on page 32. The final test cycle proposal of the institutes is a minicycle approach with positive and negative WHTC work, which is a clear improvement over the flat WHVC approach.

The detailed test results are shown on pages 36 to 48 for the Volvo bus and on pages 50 to 59 for the MAN bus. An analysis of the Iveco results is not available, since HILS modelling is still under construction. Additional chassis dyno tests are planned for early 2014.

Mr. Silberholz presented open issues for the drafting on pages 61 to 75 of HDH-15-04. Those issues were discussed in detail and will be taken into account during the further drafting process. The most prominent ones are listed below:

• Handling of gearboxes and shift algorithms; the Japanese regulation allows both generic and OEM specific model for automatic transmission.
• Definition of the rated power of a hybrid system; institutes were asked to check the US EPA approach and to make a proposal.
• How to deal with vehicles which by design cannot reach the maximum WHVC speed; it was suggested to scale power down, but EPA raised some concerns; Japanese approach should be checked.
• Work (engine or system) to be used for the emissions calculation. Mr. Dekker made clear that NL have concerns about the systems approach. The chairman concluded that there seems to be some diverging views within the group and that this will be an important topic to be discussed at 16th HDH meeting.
• Equivalency between post-transmission powertrain test, HILS with verification on chassis dyno and HILS with verification on system bench (pre-transmission powertrain test).

This report is the final report of the work done by the Universities of Technology in Chalmers, Graz and Vienna and research institute TNO performed within the research program on an emissions and CO2 test procedure for Heavy Duty Hybrids (HDH). This report specifically refers to Validation Test Program 1 (VTP1) under the contract for “Developing the Methodology for Certifying Heavy-Duty Hybrids based on HILS”.

Update on development and validation of the new hybrid heavy-duty vehicle test cycles.

4.1 Presentation by research institutes on validation test program 1

Working paper HDH-14-03e is a joint presentation of the three institutes tasked with conducting validation test program 1.

Prof Fredriksson started with an overview of the new model structure (pages 4 to 12). Rationale for the new model structure was a greater flexibility of the modeling approach that allows various combinations of engines, gear boxes and energy storage systems. The new model structure for serial and parallel hybrids was made available for testing at the end of April. It includes a comprehensive component model library, a new signal naming convention based on AUTOSAR and the restructuring of the vehicle models. Relevant powerpack components were included in the library toolbox, while previously developed models were transferred into the model library. The models will be further developed until the end of June 2013 based on feedback from OICA members and other stakeholders.

Mr. Silberholz reported on positive feedback from OEMs on the new model structure. Open issues from the discussions are the actuation of different brake systems in the driver model and time vs. distance based slope pattern. He continued with a detailed analysis of the road slope calculation methods (30 sec moving average method vs. minicycle approach, pages 15-19). The two options are considered not directly comparable due to big deviations in resulting reference power pattern. Reason is that the WHTC denormalization method leads to operation points at lower loads and speeds for hybrid powertrains. Therefore, the WHDHC calculation tool may be not suitable any longer. Further investigations on this item will continue.

As a conclusion of validation test program 1, extensive changes on models and model structure were made, the next model release (including stakeholder remarks) will be available shortly, the comprehensive model and work program documentation will start, and drive cycle modifications need further investigation.

Working paper HDH-13-03e is a joint presentation of the three institutes tasked with conducting validation test program 1.

Mr. Six started with a general overview of the progress achieved (pages 3 to 12). He then presented the structure of the thermal models. The battery model was optimized and can be used for capacitor, as well.

Prof. Fredriksson presented the new model structure (pages 13 to 27). Task was to set up a data bus system in the model that allows various combinations of engines, gear boxes and energy storage systems. With the current Japanese open source model, it is difficult to set up such data bus system. Therefore, the models need to be restructured. Two types of interfaces are proposed, the physical interface that is related to how different components are connected physically, and the signal interface that is related to the control/sensor signals needed for the ECUs. For the gtr, the physical interface should be specified, while for the signal interface only a minimum set of signals can be specified. If other signals or more complex models are needed for the simulation, it will be possible for OEMs to include those without affecting the model structure.

The new model structure will be made available by the end of April for testing. Comments are essential to improve the model.

Mr. Six then went on with the drive cycle investigations (pages 28 to 38). Rationale was to develop a method for the WHVC with slopes (also proposed by Japan in 4.2.1) to produce similar emissions results than on the WHTC for conventional vehicles. The method divides the WHVC into mini-cycles and calculates for each mini-cycle the WHTC/WHVC work difference and transforms it into average mini-cycle slopes (page 32). Emissions results based on simulation look promising. It was decided to check both the 30 sec average slope according to Japanese proposal and mini-cycles approach during validation test program 2. The mini-cycle calculation tool will be uploaded to the HDH website.

Mr. Silberholz presented under test methodology investigations the major items that need to be decided for the gtr (pages 39 to 48). He proposed that OEM specific models should be permitted. He further proposed to use on-road measurements for model verification rather than chassis dyno tests. Another item is the gear shift duration that is not included in the Japanese model. EPA is concerned that the HILS cycle might be less transient than in the vehicle. Since most of the issues are questions/proposals for further evaluation, they are handled under 7.2.

Page 40: it was agreed that the use of OEM specific models should be allowed. However, validation and verification of such models need to be ensured for the gtr.
Page 41: standardized component tests will remain in the gtr, but further discussion is necessary, if OEM specific component tests are needed.
Page 42: model verification by means of on-road tests will be investigated during validation test program 2. Powertrain verification will be added.
Page 43: OEM specific interface model need to be approved by TAA during certification process. Further discussion on what can/cannot be included in interface model is needed.
Page 44: multiple ECUs have been on the screen for quite some time. It was agreed that a master ECU would need to be defined, with supportive ECUs to be integrated via interface or software emulation.
Page 45: boundary conditions for re-certification need to be defined in the gtr; but the issue will have to be finally solved between OEM and TAA.
Page 46: information document is outside the scope of the gtr, will be handled by regional authorities.
Page 47: vehicle independent emissions certification will be discussed after validation test program 2.
Page 48: signal frequency will be discussed later; 1 Hz, 10 Hz, 50 Hz will be investigated during validation test program 2.

Prof. Hausberger presented the proposal of the institutes for validation test program 2 (pages 49 to 54). He indicated that TU Graz was able to do the vehicle testing. It was however agreed to accept the offer from JRC to do the vehicle testing and the overall organization. It was further agreed that the institutes should do the final evaluation of the HILS method, since they had done most of the development work.

Testing will start with the Volvo parallel hybrid bus in early May, followed by the MAN serial hybrid bus in early June and the Iveco parallel hybrid truck around the end of June. The OEMs are asked to arrange for a meeting among themselves and JRC to set up a test program and timetable. JRC will be responsible for the coordination of the program. Chassis dyno testing will be done with emissions measurement, on-road testing in accordance with the PEMS rules, but w/o emissions measurement. Engine testing on HILS cycle will be done at OEMs premises.

EPA requested to get the vehicle specifications. EPA further requested to receive an ECU for HILS testing. The Chairman encouraged bilateral discussions between EPA and OICA to solve the issue before the 14th HDH meeting.

Working paper HDH-13-03e is a joint presentation of the three institutes tasked with conducting validation test program 1.

Mr. Six started with a general overview of the progress achieved (pages 3 to 12). He then presented the structure of the thermal models. The battery model was optimized and can be used for capacitor, as well.

Prof. Fredriksson presented the new model structure (pages 13 to 27). Task was to set up a data bus system in the model that allows various combinations of engines, gear boxes and energy storage systems. With the current Japanese open source model, it is difficult to set up such data bus system. Therefore, the models need to be restructured. Two types of interfaces are proposed, the physical interface that is related to how different components are connected physically, and the signal interface that is related to the control/sensor signals needed for the ECUs. For the gtr, the physical interface should be specified, while for the signal interface only a minimum set of signals can be specified. If other signals or more complex models are needed for the simulation, it will be possible for OEMs to include those without affecting the model structure.

The new model structure will be made available by the end of April for testing. Comments are essential to improve the model.

Mr. Six then went on with the drive cycle investigations (pages 28 to 38). Rationale was to develop a method for the WHVC with slopes (also proposed by Japan in 4.2.1) to produce similar emissions results than on the WHTC for conventional vehicles. The method divides the WHVC into mini-cycles and calculates for each mini-cycle the WHTC/WHVC work difference and transforms it into average mini-cycle slopes (page 32). Emissions results based on simulation look promising. It was decided to check both the 30 sec average slope according to Japanese proposal and mini-cycles approach during validation test program 2. The mini-cycle calculation tool will be uploaded to the HDH website.

Mr. Silberholz presented under test methodology investigations the major items that need to be decided for the gtr (pages 39 to 48). He proposed that OEM specific models should be permitted. He further proposed to use on-road measurements for model verification rather than chassis dyno tests. Another item is the gear shift duration that is not included in the Japanese model. EPA is concerned that the HILS cycle might be less transient than in the vehicle. Since most of the issues are questions/proposals for further evaluation, they are handled under 7.2.

EV Standardization:

The European Commission representative mentioned that EV standardization was an important component of harmonization. He added that since there is much innovation happening at the moment, this is an ideal time to discuss standardization issues. Examples were given including battery size, e-performance, and connectors. The objective is not to make standardization rules mandatory but to provide a framework for manufacturers. The European Commission representative suggested adding a standardization question to the questionnaire, such as what types of components could be standardized.

The Chair noted the suggestion and commented that this suggestion went beyond vehicles to components of vehicles.

The Chair of GRPE commented that the EVE should be cautious in the wording used, for example regulations, performance requirements and standardization. He also reminded the group of the role of the WP.29 and the GRPE which is to develop regulations in support of vehicle verification activities and that is was not the role of the IWGs to develop standards.

The EVE co-Chair from Japan noted that some of these same discussions were occurring at the EVS.

The European Commission representative noted the comment made by the Chair of GRPE and suggested that the EVE could assist in initiating this work.

The Chair suggested that this question may be more appropriate for vehicle manufacturers, but also noted interest for what components would benefit most from standardization.

ACTION 6: Continue discussion on benefits of standardization at EVE-05.

EV Performance Presentation:

Dr Tober from the University of Vienna presented document EVE-04-05e, entitled ‘The Energy Requirement of Battery Electric Vehicles Under Different Conditions’.

Working paper HDH-12-03e is a joint presentation of the three institutes tasked with conducting validation test program 1.

Mr. Six started the presentation with a general overview of the progress achieved. Development of the driver models for the different approaches (WHVC and WHDHC) has been finished. Inclusion of non-electric components has been largely finalized, and the development of the thermal models will be completed within the next two weeks. From meetings with OEMs (see page 7), it was reported that a fully vehicle independent approach does not seem to be feasible, and that some OEMs considered that a SILS approach would be preferred over HILS. Most commonly used powerpack components are already included in the component library with a planetary gearbox model added recently. The component list will be checked, if vehicles for validation test program 2 are identified.

Mr. Silberholz continued with the description of the thermal models (task 1.6). Details of task 1.6 are shown on pages 9 to 17. The thermal model for the aftertreatment system (ATS) has been finished, the models for the coolant and lube oil circuit will be finished by the end of January, and the models for battery and electric motor will be finished within the next week and validated by the end of January.

Investigation into the different drive cycle options (WHVC vs. WHDHC) is an important part of validation test program 1. The results are shown on pages 18 to 34 for the serial hybrid. The two options only lead to comparable positive traction work, if the WHVC is accompanied with slopes. Even then, the power distribution over rotational speed at the wheel hub is different between the two options, as shown on page 27. Using the WHTC as basis leads to highly fluctuating torque signals that may cause ECU errors, as already indicated by Japan at the 11th HDH meeting. In order to solve the problem, curve smoothing is necessary. Curve smoothing was demonstrated to have no significant impact on the positive work (page 31) nor on the power distribution (page 32).

The major focus of Task 2 is currently the restructuring of the models, which is a prerequisite for setting up the data bus system needed for the component library. The results are shown on pages 37 to 44. The restructuring requires two types of interfaces (page 39). The physical interface is related to how different components are connected together, physically. The signal interface is related to control/sensor signals needed for the ECU. The proposed solution is the port based approach shown on pages 40 to 42. This approach allows easy interchanging of components including integration of OEM subsystems.

Status of the work towards developing a methodology for certifying heavy-duty hybrids based on “hardware in the loop simulation” (HILS).

Mr. Six started with the presentation of working paper HDH-11-04-Rev1. First, he indicated that the paper is a joint presentation of the three institutes involved. The institutes have a regular exchange of information.

The basis for the development of the serial hybrid simulator is the Japanese open source model for a serial hybrid provided by JARI. As a first step, an ECU control strategy with different ICE operation points was added to the model. Further, a driver model for the WHVC approach was developed. Details of task 1.1 are shown on pages 5 to 10.

Task 1.2 covers the development of a driver model appropriate for the WHDHC approach. Rationale for the WHDHC approach is to be in closer agreement to the test procedure for conventional ICEs. It is therefore essential that the driver model allows running the simulator with test cycles consisting of power and rpm at the wheel hub and at the power pack shaft.

This requires two PID controllers, cycle reference tables and a watchdog system that prevents too aggressive control parameters, such as gradients. Task 1.3 covers the extension of the simulator with a library of non-electric components. Details of tasks 1.2 and 1.3 are shown on pages 13 to 18.

In order to develop the software ECU in a manner as realistic as possible, input of the OEMs is needed. An interface list and a component list were submitted to the OEMs for review and have been uploaded to the HDH webpage. Meetings with OEMs are planned to start in November 2012. The OEM input is specifically needed for task 1.5 (additional powerpack components) and task 1.6 (thermal models).

Mr. Silberholz continued the presentation with an overview of the thermal models. Thermal models are considered for the aftertreatment system (ATS), for the coolant and lube oil circuit, and for battery and electric motor. The ATS thermal models have been implemented in Simulink and are currently being validated with existing measurement data. Measurements for the parameterization of the thermal model for coolant and lube oil are currently conducted on an engine test bed. The thermal models for battery and electric motor are under preparation in cooperation with the Institute Electrical Measurement and Measurement Signal Processing at TU Graz. OEM input and validation is needed after the implementation. Details of task 1.6 are shown on pages 22 to 34.

After implementation of the models, simulation runs and validation of basic functions will be conducted. While relatively simple control strategies will be used for the model validation, the simulation runs are suggested to be based on largely realistic control strategies. Generic values, measurement data of the institutes and measurement data from OEMs (if available) will be used as input data. Task 1 is planned to be completed by the end of January 2013.

Task 2 (adaptation of the HILS simulator for parallel hybrids) will start in October 2012. The institutes propose that the model validation and simulation should be done with vehicles intended for validation test program 2. Task 3 includes the description of the test procedure and the user manual for the software. It was indicated that the Task 3 can only be on schedule, if the HILS structure is retained. Establishment of a component library, as favored by the HDH IWG, does not fit into the current HILS model.

Status of the work towards developing a methodology for certifying heavy-duty hybrids based on “hardware in the loop simulation” (HILS).

Proposal from the research institutions involved in assessing and adapting hardware-in-the-loop simulation (HILS) methodologies to the evaluation of heavy-duty hybrids for the anticipated second phase work to produce a HILS tool for use in GTR 4 hybrid vehicle provisions.

Investigation into various methods in use for evaluating heavy-duty hybrid emissions and CO2 performance.

Mr. Planer presented working paper HDH-09-13. He briefly summarized the results of the TU Vienna work package and indicated that the final report would be available, soon.

Meanwhile, the final report has been published and circulated to the HDH group as working paper HDH-09-15. He then went on with the suggested validation phase 1 (pages 17 to 19).

As already introduced at the 8th HDH meeting, validation phase 1 is intended to verify the suggested changes of the research institutes to the Japanese HILS method. The proposal is based on the ECU as software in the loop as basis for further programming and software development. As task 1, a serial hybrid simulator would be developed, since the ECU is considered easier to implement. The simulator would cover both the WHVC and the WHDHC as input cycles for the simulation, and would be extended with a library for non-electric components, a library of new components, such as planetary gear box, power split, and with thermal models (exhaust gas aftertreatment components, coolant, lube oil, battery and electric motor). Task 2 would adapt the simulator to a parallel hybrid. As a result of tasks 1 and 2, task 3 would consist of amendments to the Japanese procedure with respect to the methods for component testing, test cycle definition and the simulation method. In task 4, an interface system would be defined to be used for real ECUs.

The participants principally agreed to the proposal. It was common understanding that starting with the serial hybrid would be most beneficial. JARI indicated that they had completed a serial hybrid model, which would be made available to TU Vienna for the validation study. JARI further informed that they already have a simple serial hybrid ECU and would support TU Vienna in developing the software ECU. OICA agreed to deliver input to task 1.4.

OICA offered to sponsor at least part of validation phase 1 depending on the total budget needed. The secretary asked the participants to consider contributing to the validation. In order to get a clearer picture, Mr. Planer was asked to submit a quote to the secretary by mid-April. Task 6 on page 19 (real vehicle testing) of working paper HDH-09-13 is not considered part of validation phase 1, but validation phase 2 and should therefore not be part of the quote. In order to meet the revised project plan the program should start no later than June 2012. Final approval of validation phase 1 will be on the agenda for the 10th HDH meeting.

Mr. Martinez indicated that EU COM would most likely take over the budget for validation phase 2. Intention would be to run this validation at JRC. OICA will check by the 11th HDH meeting (October 2012) which HDVs could be offered for this validation. It is expected to run at least one parallel and one serial hybrid.

TU-Wien presentation on its work with IFA on the assessment of HILS methodology and suggestions for the GTR validation test program.

Mr. Planer presented working paper HDH-09-13. He briefly summarized the results of the TU Vienna work package and indicated that the final report would be available, soon.

Meanwhile, the final report has been published and circulated to the HDH group as working paper HDH-09-15. He then went on with the suggested validation phase 1 (pages 17 to 19).

As already introduced at the 8th HDH meeting, validation phase 1 is intended to verify the suggested changes of the research institutes to the Japanese HILS method. The proposal is based on the ECU as software in the loop as basis for further programming and software development. As task 1, a serial hybrid simulator would be developed, since the ECU is considered easier to implement. The simulator would cover both the WHVC and the WHDHC as input cycles for the simulation, and would be extended with a library for non-electric components, a library of new components, such as planetary gear box, power split, and with thermal models (exhaust gas aftertreatment components, coolant, lube oil, battery and electric motor). Task 2 would adapt the simulator to a parallel hybrid. As a result of tasks 1 and 2, task 3 would consist of amendments to the Japanese procedure with respect to the methods for component testing, test cycle definition and the simulation method. In task 4, an interface system would be defined to be used for real ECUs.

The participants principally agreed to the proposal. It was common understanding that starting with the serial hybrid would be most beneficial. JARI indicated that they had completed a serial hybrid model, which would be made available to TU Vienna for the validation study. JARI further informed that they already have a simple serial hybrid ECU and would support TU Vienna in developing the software ECU. OICA agreed to deliver input to task 1.4.

OICA offered to sponsor at least part of validation phase 1 depending on the total budget needed. The secretary asked the participants to consider contributing to the validation. In order to get a clearer picture, Mr. Planer was asked to submit a quote to the secretary by mid-April. Task 6 on page 19 (real vehicle testing) of working paper HDH-09-13 is not considered part of validation phase 1, but validation phase 2 and should therefore not be part of the quote. In order to meet the revised project plan the program should start no later than June 2012. Final approval of validation phase 1 will be on the agenda for the 10th HDH meeting.

Mr. Martinez indicated that EU COM would most likely take over the budget for validation phase 2. Intention would be to run this validation at JRC. OICA will check by the 11th HDH meeting (October 2012) which HDVs could be offered for this validation. It is expected to run at least one parallel and one serial hybrid.

Mr. Schneeweiss presented the work program of the Institute for Powertrains & Automotive Engineering (IFA) at the TU Vienna. After a general introduction, he first focused on the interface model. The major task of the interface is connecting the simulation model and the hardware ECU (see slide 10). The interface model is manufacturer specific and remains confidential, but must be disclosed to the approval authority. In case of multiple ECUs, the most important functions of less important control units could be implemented via the interface model of the HILS system, which is a kind of simplified software-in-the-loop.

The Japanese simulation model is realized with Simulink, a well established programming language, which is based on physical models and lookup tables (see slide 15). The most important model is the powertrain model (see slide 16). In Japan, five different powertrain models exist, but for a worldwide regulation more powertrain models would likely have to be developed. In order to solve this problem, IFA is proposing a component library. Components are physical entities (e.g. engine, motor-generator, battery, clutch, gearbox etc.) that are combined to result in a special powertrain topology. IFA also suggests to develop temperature models to take into account hybrid control strategies for optimized engine warmup. Prof. Hausberger added that this would also be needed for the warm-up of any aftertreatment system. The Chairman recalled that the WHDC procedure includes a cold start and a hot start test, and this issue must be dealt with in a global regulation. It was agreed that this could finally result in two different engine cycles for cold start and hot start.

The Chairman then raised the question where the models would be defined. It was general understanding that the gtr could only include some base models, like in Japan. Furthermore, it will not be possible to amend the gtr every time a new model is proposed by a manufacturer. The Secretary proposed an approach of defining in the gtr general guidelines of model construction on the basis of the component library proposed by IFA. With this approach, verification of the models becomes more important.

IFA then introduced an alternative to the Japanese vehicle based approach for consideration of the group. This alternative uses the premise that engine work of a conventional powertrain (on the WHTC) and of a hybrid powertrain (from HILS) should be comparable. The simulation would be done using the engine WHTC torque from WHTC speed input signal at the gearbox, as illustrated in slide 26. The approach is elaborated in more detail in working paper HDH-07-05rev by TU Graz.

IFA is proposing to do the model verification by comparing simulation results to actual measured data from test track driving using a random driving cycle. This is due to problems, which might occur in chassis dyno testing (correct recuperation, high cost).

IFA concluded that the Japanese component testing provisions could basically be used for a global regulation. Component testing strongly depends on the modeling depth and on the desired accuracy. Under this aspect, Mr. Jackson asked how the durability of the hybrid systems would be taken into account. This is currently not included in the Japanese regulation, but the Chairman emphasized that at least some aspects like durability of the storage system should be covered by the gtr.

Finally, Mr. Schneeweiss indicated that the final report would be issued by the end of 2011.

Mr. Schneeweiss presented the work program of the Institute for Powertrains & Automotive Engineering (IFA) at the TU Vienna. After a general overview of the institute and its core competences, he emphasized IFA’s experience with simulation models and hybrid drivetrains. He then explained the project approach in detail.