Analysis and Implementation of Rapid Computation of an Embedded Predictive Controller in a SoC FPGA

Abstract

With the advancements of Industry 4.0, it becomes necessary to apply advanced control techniques at the lower levels of the automation pyramid, using controllers that operate in the order of microseconds or milliseconds. FPGAs are chosen as the implementation platform due to their processing speed, parallelization, and determinism. However, the exclusive use of FPGA may pose challenges in implementing complex algorithms. Therefore, the aim of this work was to develop a prototype of a fast computing predictive controller using the GPC model, one of the most widely used models in the industry, embedded in an FPGA-based SoC device, employing a mixed architecture. Parts of the code were implemented using parallel computing, while others used sequential computing. The obtained results were compared with other applications, revealing significant improvements in the control signal processing time.