⭐⭐⭐⭐⭐ Sensor networks for agriculture based on #FPGA

✅ Objectives: Acquire signals from humidity sensors located on a porous surface, sunlight, air temperature and humidity. Compare the performance of at least two neural networks for time series. Measure the processing times,% of CPU used and% of memory used, in the processor #NIOSII, #ARM or both, as the case may be (radar chart).

✅ Goals: Implement an Embedded System based on #FPGA that acquires signals from humidity, temperature, incident light sensors and is capable of storing the readings of these sensors in memory Double Data Rate 3 Synchronous Dynamic Random-Access Memory (# DDR3_SDRAM), access is It will do it through the Hard Processor System #HPS. The training process of the neural network for time series can be carried out on a conventional computer, then the model trained in C ++ language must be executed by the #NIOSII processor in the #FPGA. The system should output a VGA or HDMI monitor that clearly displays the rating results. The FPGA-based system is the one who executes the filtering blocks and neural network, but the groups that have telematics students will not use the #VGA or #HDMI output but must access the visual interface through a web application deployed in a web application server running on a linux operating system with #ARM processor.

✅ Abstract: The main task is to analyze the signal that represents the behavior of humidity on the porous surface as a function of parameters such as incident light, humidity and air temperature; that is, soil moisture over time will depend on the conditions of the environment in which it is found 1. The neural network for time series will have the ability to predict the behavior of soil moisture over time just by analyzing the incident light, humidity and air temperature data. The present work does include the data acquisition stage for subsequent processing with a neural network for prediction of time series (soil moisture). After the acquisition of the sensor signals, a processing will be carried out that includes: Characterization and forecasting of time series. In the acquisition stage, the communication between the different sensors with their respective protocols (LDR-ADC, soil humidity -ADC, humidity and air temperature - onewire) will be carried out with the FPGA 1. Characterization is the process of extracting relevant information during time windows determined by the students (1 second is recommended). The time windows are defined in the experiment during the acquisition of the signals from the different sensors 1. In the forecast of the time series, the neural network (NN) uses the most representative characteristics with which the algorithms have a better performance in predicting the outputs. For proper performance, the NN algorithms require that the signals be processed in a previous preprocessing stage.

✅ Reference : 1 Romero, G., Salazar, C., & Asanza, V. (2015). Desarrollo de un Prototipo de Sistema Hidrometeorológico. Revista Tecnológica-ESPOL, 28(5). ✅ Hardware: (1) #DE10_Standard o #DE10_Nano (1) Monitor VGA / HMDI (1) #LDR, (1) #DHT11 y (1) #FC_28

Read related topics: ⭐ Projects Digital Systems Design #FPGA ✅ Introduction ✅ Programming ✅ #Proteus #PCB Design ✅ Paper: Monitoring of system memory usage embedded in #FPGA ✅ Embedded System Projects ➡️ End Device #Arduino #FreeRTOS ➡️ End Device + Coordinator #Raspberry Pi #Python ➡️ #Proteus #PCB Design ✅ TinyOS for sensor networks #XBEE ✅ Electronic Prototype Development using #ALTIUM #CircuitMaker ✅ Instalación de #ALTIUM #CircuitMaker y especificaciones del módulo #ESP32 ✅ Microcontrollers Application using #Labview ✅ Práctica 1: Salidas Digitales #Arduino

Posted Mar 20, 2021 by Asanza, Victor