Vehicle Simulation Test Platform: Edge Controller Solution

Modern automotive development relies heavily on vehicle simulation test platforms—such as road simulators, multi-axis shaker tables, and component durability test rigs—to replicate real-world load conditions in a controlled laboratory environment. The performance of the control system directly impacts test data validity, development cycles, and costs. With the push toward lightweight electric vehicles and the demand for multi-axis coupled, high-dynamic testing, traditional control architectures face significant engineering challenges.

Multi-Axis Synchronization and Real-Time Load Spectrum Reproduction: Modern test rigs often include 4 to 8 or more independently controlled actuators (hydraulic or electric) to simulate vertical, lateral, and longitudinal loads on wheels, body, or key components. Road load spectra contain frequency components from 0.1 Hz to tens of Hz, requiring command updates and closed-loop control within milliseconds or even sub-millisecond time scales. Traditional “industrial PC + multi-axis motion control card” architectures suffer from independent clocks per card and inherent communication jitter over PCI/PCIe buses. In multi-axis coupled loading, this leads to phase errors between actuators, distorting the reproduced road conditions and reducing test confidence.

High-Fidelity Force and Displacement Signal Acquisition: Actuator force sensors (e.g., strain-gauge load cells) and displacement sensors (e.g., LVDT, magnetostrictive) output low-level differential analog signals, often in the millivolt range. These are susceptible to electromagnetic interference from the test rig and hydraulic power units. Long single-ended cable runs to a centralized data acquisition card cause signal degradation and common-mode noise, leading to static errors or oscillations in force/displacement control loops. Research shows that feedback signal quality is critical for control algorithm performance.

Complex Test Profile Management and Data Synchronization: A complete durability test may involve dozens of different road profiles (e.g., Belgian block, washboard, high-speed circuit), each with specific load spectrum files and parameters. In traditional setups, switching between profiles requires manual file loading and parameter adjustments, which is error-prone and time-consuming. Additionally, force/displacement data, actuator positions, and cycle counts are often stored separately with unsynchronized timestamps, making it difficult to correlate anomalies with specific test conditions for root cause analysis.

Remote Collaboration and Test Monitoring Needs: Test laboratories are often geographically separated from design and analysis teams. Engineers typically need to be on-site to monitor tests, adjust parameters, and download data, leading to delayed responses to unexpected events. For long-duration durability tests (lasting weeks), on-site staffing costs are high.

Unified Control Core: A model like BL372B serves as the main controller. Its heterogeneous computing architecture separates tasks: a quad-core ARM Cortex-A53 processor runs Linux for non-real-time applications such as profile management, data logging, interfacing with simulation software like Simulink Real-Time, and remote communication. A dedicated ARM Cortex-M0 core, under the Linux-RT-5.10.198 real-time operating system, handles time-critical tasks: multi-axis closed-loop control algorithms, high-speed analog acquisition and output, and EtherCAT communication management. This separation ensures stable, low-latency control cycles.

EtherCAT-Based Hard Real-Time Synchronized Drive Network: Using the built-in IgH EtherCAT master stack, all actuator servo drives (or servo-valve control units) are connected on a single real-time network. EtherCAT’s distributed clock mechanism enables sub-microsecond synchronization of command cycles across axes, ensuring that during multi-axis coupled loading, each actuator follows the target load spectrum with precise phase relationships. Communication cycle times can be configured down to 250 µs or lower, meeting the demands of high-frequency load spectrum reproduction.

High-Precision Analog Acquisition and Output: Modular I/O boards are placed close to the actuators to acquire force/displacement signals and output control signals, minimizing analog signal transmission distance and improving noise immunity.

Software-Defined Test Workflow and Remote Collaboration: Upper-level software tools enable centralized test profile management, seamless data exchange with simulation software, and remote test monitoring and data download.

Core Control Unit: BL372B (3× EtherCAT ports, 1× X-slot, 2× Y-slots). Port 1 connects to the actuator servo drive/servo-valve network; Port 2 connects to a data acquisition expansion station or local operator interface; Port 3 connects to the lab Ethernet for communication with simulation hosts and remote access terminals. Processing core: SOM372 (RK3562J, 32 GB eMMC, 4 GB LPDDR4X) provides ample storage for load spectrum files, test data records, and event logs. Operating system: Linux-RT-5.10.198 kernel ensures real-time performance for multi-axis control and high-speed analog I/O.

Seamless Integration with Simulink Real-Time and Similar Tools: The edge controller acts as a real-time target machine, communicating via Ethernet with a host PC running Simulink Real-Time. The host handles high-level tasks like load spectrum model computation and test sequence scheduling, sending target force/displacement values for each axis per control cycle via UDP or shared memory. The edge controller receives these targets, executes high-frequency closed-loop control using feedback from Y34/Y36 boards, and outputs control signals via Y46. It also sends actual force/displacement data back to the host for recording and display. This “host model computation + target hard real-time control” architecture leverages the strengths of both systems.

Test Profile Management with QuickConfig: This tool provides a structured interface for managing test configuration files. Key features include:

• Profile Library: Package different road load spectrum files (time-domain force or displacement spectra) and associated parameters (actuator stroke limits, safety thresholds, sampling rates) as reusable profile templates.
• Sequence Programming: Combine multiple profiles into a complete test sequence with configurable cycle counts and transition times.
• One-Click Loading: At test start, the operator selects a target sequence, and the system automatically loads the corresponding spectrum files and control parameters into the real-time control task, simplifying operation.

Remote Test Monitoring and Data Download with BLRAT: Through a secure remote access channel, engineers in offices or remote labs can connect to the on-site edge controller. Capabilities include:

• Real-Time Monitoring: View target vs. actual force/displacement curves and error traces for each actuator to assess test health.
• Parameter Tuning: Remotely adjust PID gains or feedforward coefficients during the test based on observed errors, optimizing control without interruption.
• Data Download: After or during the test, remotely download recorded force/displacement data and event logs for analysis.
• Fault Handling: When alarms occur (e.g., sensor over-range, excessive following error), view details remotely and stop the test if necessary to prevent equipment damage.