Subaru saved 2,000 man hours with PXI-based hybrid vehicle testing

Sponsored: Subaru saves 2 000 man hours with PXI-based hybrid vehicle testing

PXI has long been established as the de facto standard for building automated test systems, but its uses stretch beyond solely manufacturing or production test.

With the PXI platform’s capability for running real-time operating systems or deploying code to FPGAs, it is also well suited to use earlier in the development process, particularly for complex devices like electronic control units (ECUs) and full authority digital electronics control (FADEC) systems.

Today, automobiles are equipped with an enormous number of ECUs to manage expanded functionality and advanced controls in the vehicle. In a hybrid vehicle, the motor ECU plays an even more complicated role as it manages the interaction between the conventional engine and the electric motor, along with its power systems.

Fuji Heavy Industries, parent company of Subaru, set out to develop its first hybrid vehicle - the Subaru XV Crosstrek Hybrid. This was their preliminary attempt to deliver a production model hybrid vehicle targeting both domestic Japanese and North American markets.

Tomohiro Morita of FUJI Heavy Industries explained, “Our engineers had developed a motor ECU for an earlier hybrid prototype, but the component did not meet the rigorous requirements to take a vehicle to market. For the production model vehicle, the ECU needed various control functionalities to prevent damage to the vehicle body and to ensure driver and passenger safety under various operating conditions, even scenarios that would be impossible or impractical to test on physical hardware.”

“For example, under icy driving conditions, a wheel can experience a sudden loss of traction. During acceleration this can cause a dramatic increase in motor speed and needs to be handled safely. However, this safety behaviour cannot be physically reproduced on a dynamometer and is time consuming and difficult to reproduce on a test track. Since complex control algorithms for specific safety conditions like this need to be developed and verified, the testing needed to account for outlying operating conditions to satisfy the quality level required for a production model vehicle.”

A new approach

FUJI engineers connected the ECU to a real-time electric motor simulation to test and verify a variety of conditions, including the extreme outliers that might break the system in traditional mechanical testing. They developed a mechanism to sufficiently confirm this software simulation approach with three primary goals for successful testing:

  • Verify ECU functionality in various conditions, including extreme environments not easily created or replicated.
  • Map test cases to requirements to ensure complete test coverage.
  • Perform regression tests with ease to quickly validate design iterations.

To achieve these goals, the engineering team used a V-diagram approach to launch the design and verification process. The diagram describes a phased methodology for embedded software design and deployment validation, including test points at each stage. In multiple stages of the design process, the team needed the hardware-in-the-loop (HIL) system to verify the motor ECU against a real-time motor simulation that accurately represented the actual vehicle motor. Additionally, using the HIL system, they could meet traceability requirements by recording test results automatically and automating regression tests when an ECU change was made.

System success

The new verification system consists of a real motor ECU and the PXI-based HIL system that simulates motor operations. The HIL system can represent any operating condition of the motor by setting physical parameters such as inductances or resistances. It can also set parameters of the power electronics, including fault conditions or test scenarios such as combinations of load torque and desired rotating speed.

By simply changing a parameter in the middle of the test, the HIL system can easily simulate complex test scenarios. This could include the previously mentioned loss of traction or even a power electronics fault in the inverter that would destroy physical hardware. When the operator requests a test pattern, the HIL system responds the way a real motor would, and the overall system response can then be cross-referenced with expectations to validate that the controller safely handles the test case.

NI Subaru diagram  

Diagram: verification environment using the PXI-based HIL system

“Because the computational performance required for this process was so high, we felt National Instruments was the only supplier that could meet these requirements” explained Morita. “We chose core system hardware based on NI FlexRIO, FPGA-based PXI modules. The modules executed a model representing the simulated operation of the motors, with all deployed programs using NI LabVIEW system design software.”

We created the test scenario, including required torque and rotating speed, chronologically in an excel spreadsheet. According to these conditions, the motor ECU operates and sends out signals, such as a pulse-width modulation signal, to the HIL system. The HIL system receives these signals then simulates operation of a real motor. More specifically, the computational process is performed and the result is an output at the same speed as the real motor. The resulting signals representing the torque and the triphase current are returned to the motor ECU. We automated the verification process using LabVIEW to read and execute the Excel spreadsheets for the test scenarios, with obtained results written automatically to the excel spreadsheet for the test report.”

Benefits of choosing the NI platform

In the HIL system, the simulation loop rate, equivalent of the temporal resolution in simulation, was a critical factor. For the motor ECU, the loop rate needed to be 1.2 µs or less for the simulator to work.

Morita commented that, “Most simulation platforms from other suppliers use CPUs for computation, resulting in a loop rate in the range of 5 µs to 50 µs. NI FlexRIO used the FPGA for control and computational purposes to meet the required simulation rate of 1.2 µs. Additionally, because the NI FlexRIO has a high-capacity, built-in dynamic random access memory, we could use the JMAG-RT model provided by JSOL Corp.’s JMAG software tool chain. This made it possible to represent the highly non-linear characteristics closer to the real motor.”

“Moreover, our engineers could program the FPGA on the NI FlexRIO device graphically with the NI LabVIEW FPGA Module, which made it possible to develop a system with FPGA technology in a short time frame without using a hardware description language.”

One of the major benefits of a PXI-based, automated system is the amount of engineering time saved in performing the tests, as well as removing the risk of human error that can occur in manual testing.

Morita continued, “All the test patterns developed can be run automatically in only 118 hours. Performing all the tests manually would take an estimated 2,300 hours. The HIL system delivered additional time-saving advantages that included a significant reduction in the number of setup procedures, such as preparing a motor bench and a test vehicle, and it removed the need for test personnel to be qualified to handle high-voltage equipment.”

The software-defined approach with PXI offered significant advantages to FUJI when performing verification of ECUs for the Subaru XV Crosstrek Hybrid. From increasing throughput to reduce test times, to options for running in real-time or on FPGAs for higher performance and faster loop rates, PXI is well-suited to testing complex electronic devices throughout the design process.

Further information

Recent articles

Info Message

Our sites use cookies to support some functionality, and to collect anonymous user data.

Learn more about IET cookies and how to control them