The Madgwick Quaternion Filter Unit (MQFU) is a high-performance FPGA-based hardware accelerator designed for real-time inertial attitude estimation. This project implements the Madgwick Quaternion Filter using fixed-point arithmetic, significantly improving throughput and reducing processor workload compared to software-based implementations.
- Configurable Fixed-Point Design: Python-based configuration tool for dynamic adjustment of range and precision.
- Efficient Hardware Integration: Packaged as a Xilinx IP block with AXI interface for seamless SoC integration.
- Wishbone Interface Support: Includes a Wishbone-compatible interface for easy integration with open-source and custom SoC architectures.
- VeeRWolfX SoC Integration: Demonstrated integration into the VeeRWolfX SoC, showcasing compatibility with a RISC-V-based processor and peripheral system.
- Improved Performance: Offloads computationally intensive attitude estimation from the CPU, enhancing real-time capabilities.
The MQFU is synthesizable on Xilinx FPGAs, such as the Artix-7, and can be adapted for various inertial measurement systems, making it ideal for applications in drones, satellites, and robotics.