*As Featured on NI.com
Original Authors: Marc Scherer, ITK Engineering AG
Edited by Cyth Systems
The requirements for testing steering systems have increased enormously. Along with mechanical tests, highly dynamic tests of electrical steering systems on test benches are now common and are increasingly being performed under realistic conditions. Additional requirements result from the use of active test objects with their own actuators, whose behavior is strongly influenced by the contained ECU.
ITK Engineering AG (ITK) has delivered a highly dynamic steering test bench for realistic testing to an Asian automobile manufacturer. It features full automation based on National Instruments (NI) PXI, VeriStand and TACware®, ITK’s software for test bench automation, developed with LabVIEW and LabVIEW Real-Time.
Left: Overview of the utilized software functions of VeriStand, Right: Data from high dynamics test.
A Highly Dynamic Test Bench for Testing Electric Power Steering Systems
To minimize costly road tests with test vehicles, test benches must enable realistic testing. For one of their customers, ITK therefore developed a highly dynamic test bench for electric power steering systems that enables haptic tests in addition to automated and simulation-based test sequences.
To create realistic test conditions, the same physical quantities are applied to the steering system on the test bench as would occur during test runs with the system installed in a test vehicle. A hydraulic load actuator generates forces or assumes positions that act on the steering system and are equivalent to real street loads. Target force values, for example recorded during previous test drives, can be reproduced on the test bench and controlled dynamically over a range up to 25 kN. To enable realistic testing and fulfill stringent requirements for dynamics and precision, highly effective technologies such as synchronized hydraulic cylinders and mechanical structures with optimized vibration characteristics are used.
Furthermore, automated steering wheel angles and torques are regulated by a steering machine. If necessary, steering motions can also be performed manually with a steering wheel (“Driver-in-theLoop”). This allows haptics and subjective driving feel to be evaluated directly on the test bench. (Figure 1)
Test Bench Automation
Due to the high requirements for performance and real-time capability, ITK chose the powerful NI PXIe-8135 system as the run-time platform for test bench automation. Communication with the actuators and sensors of the test bench was implemented with the versatile multifunction NI X Series PXIe-6363 data acquisition modules and NI Industrial Communications for EtherCAT. NI VeriStand provides the basic software environment for real-time based tests in the test bench automation. With its integrated functions for test sequence definition and execution using the NI Stimulus Profile Editor, integration of simulation models and capability for user customization of the GUI during operation, VeriStand hit the right buttons with the developers. VeriStand’s open architecture was a key factor in the selection process. This allowed additional LabVIEW elements and functions to be integrated and extended in the VeriStand workspace (user interface). Also, functions needed in the real-time system could be added with LabVIEW Real-Time in the form of an asynchronous custom device.
Precise Control for Realistic Testing
The quality of test bench feedback control is crucial for realistic testing. In addition to the highest possible control accuracy, target values must be regulated quickly and efficiently. Furthermore, it must be possible to adjust test bench feedback control design, for example for different steering system variants, flexibly and with low user effort. The main sources of the high control system requirements for the steering test bench are the interaction between angle feedback control (steering machine) and force feedback control (load actuator) and various non-linear effects, such as stiction and mechanical play. The negative influence of angle feedback control on force feedback control is amplified by the active power assist of the steering system under test. Combined with increased dynamic characteristics of the control loop, this can easily lead to instability. Suitable feedback control algorithms with optimal parameters, minimal signal latencies and 20 kHz control sampling rates provide effective compensation for disturbances, cross-coupling and non-linearities, as well as extremely high stability.
Figure 3 shows a test run with angle and force control under highly dynamic conditions. In this case the control target values came from measurements in a real vehicle, but they can also be calculated online using a simulation model. In that test mode the test bench would be “in-the-Loop”, which means in the same control loop as the steering system and a vehicle model.
To ensure test bench feedback control performance even with new steering variants or with altered control loop transfer characteristics, test bench operators can automatically redesign the feedback controller, including controller parameters, on the test bench without any need for expert knowledge. This significantly reduces setup times. This methodology is provided by ITK’s in-house developed automation solution TACware®.
After just nine months of development time, ITK delivered a turnkey custom test bench for highly dynamic, realistic testing of electric power steerings. It is built on the combination of NI PXI, VeriStand and the TACware® software, which is based on LabVIEW and LabVIEW Real-Time. Along with basic tests such as manual target value setting and automated test sequences, steering system testing “in-the-Loop” and “Driver in-the-Loop” are equally possible. In addition, the integrated methodology for automated feedback controller design and parametrization significantly reduces setup times for changing test object variants or modification of constraints.
Marc Scherer, ITK Engineering AG
Edited by Cyth Systems