FPGA-Based Adaptive Modeling Method for Modular Robots

With the development of modular robot technology, various forms of modular robots have emerged, such as new unmanned ground vehicles (NUGV) and reconfigurable manipulators. Modular robots can change their structure by replacing modules to meet different environments and tasks. The modules of modular robots are functionally independent and can be flexibly replaced without affecting the entire system. However, due to the different structures of the modules and the different devices they carry, new modules often require redesigning the control program and even the circuit board. In order to improve the design and verification of models, this paper proposes a new design and implementation method that revolves around the Simulation Description Format (SDFormat) model. First, the SDFormat model describes a new modular robot. This work parses the model and generates control logic described in a hardware description language (HDL). Finally, the control logic runs on a field-programmable gate array (FPGA) control board. The approach follows the design philosophy of Model-Based Systems Engineering (MBSE): SDFormat models adapt control logic to a set of identical modular devices; FPGA programmable logic makes the extracted control logic easy to implement. In addition, this module runs and maintains the original acceleration capabilities of the FPGA. To validate this approach, a simple modular robot model is built. The results show that the control logic is correct and the performance of the modular robot is as expected.