Implementation of a model for isolated microgrids with photovoltaic energy sources through emulation in an application-specific specific instruction-set processor (ASIP)

Abstract

The main objective of the present research was to develop a mathematical model for microgrid simulation and implement it in an application-specific processor to measure the performance gains obtained with this approach against the performance of a general use processor. Given that this project was a first approach the topic of microgrid simulation, its scope was limited to stand-alone microgrids with photovoltaic generators as the main source of energy. A mathematical model was presented and implemented in the simulation program. An Intel Core i7-8750H was utilized as the general-purpose processor. The model was programmed in Python 3.6 language with the use of C optimized libraries: Numpy for arithmetic operations and Pandas for data handling tasks. Differential equations were solved with Euler’s explicit integration method that was selected for its simplicity which could be leveraged during the hardware implementation stage of the project. The validation of the simulation program was done through the comparison of results against PowerSim which was selected as a benchmark simulation software. A specific microgrid architecture was selected and simulated with open-loop control, for which the approximate absolute error was 1% for the steady state regime. The same simulation program was then ported to C and implemented in a Xilinx Zedboard FPGA development board. High-Level Synthesis (HLS) was used to convert the program to its hardware counterpart. The previous microgrid architecture was programmed in the board for comparison against the general-purpose CPU. The performance gains were measured based on execution time. For the selected architecture, and with varying number of elements in the microgrid, the FPGA based program was 14.2 to 87.1 times faster. Real-time simulation was also achieved with up to 2.7 times greater magnitude of signal oscillations during the system transient.