Architecture
As shown in the diagram, Dyshlo contains the core library (red) and extensive on-target verification scaffolding (blue). Only the items shown in red are used in the production application. POSIX SITL builds are supported for development and experimentation purposes.
SITL/HITL simulation framework
The SITL/HITL electromechanical simulation engine enables extensive V&V and automatic parameter optimization. The framework is compatible with the industry-standard data science stack based on Python+SciPy+Pandas+Plotly+Jupyter.
Applications
- Propulsion systems for electric aviation (manned and unmanned)
- Traction drives for electric vehicles and robotics
- Micromobility systems
- High-performance actuators and servo drives
- Research
Capabilities
- Efficient field-oriented control of PMSM and BLDC motors utilizing bleeding edge technologies from Zubax Robotics.
- Automatic characterization (system identification) of the connected motor and its mechanical load.
- Sensorless and sensored operating modes with various state estimators.
- Various optimal control strategies: torque, velocity (incremental nonlinear dynamics inversion), position, voltage.
- MTPA and field weakening.
- High-frequency injection (HFI) for robust sensorless operation at standstill (soon!).
- Dynamic modulation frequency foldback (soon!).
- Compatibility with modular multilevel converters (MMC).
- Built-in electromechanical simulation engine for R&D and V&V.
Features
- Completely platform-agnostic, header-only.
- Compatible with any conventional MCU platform (FPU required).
- Built-in SITL/HITL verification framework with 94% branch coverage.
- Self-explanatory, well-designed OOP codebase in modern C++.
- AUTOSAR C++14 compliance.
- Codebase is continuously analyzed by several distinct static analysis tools.
- Usable in bare metal environments or preemptable RTOS.
- Does not require exceptions, heap, RTTI.
- Compatible with interrupt-free environments for safety-critical applications.