The main goal of the thesis is to correctly identify the type of temperature sensor and then perform temperature measurements. After the measurements are taken, the objective is for the algorithm to recognize the sensor. Upon recognition, the goal is to begin outputting the actual temperature.
The thesis includes the development of a circuit on a protoboard, establishing communication, creating a graphical interface, and the core problem, the recognition algorithm. The algorithm is implemented using the method of least squares. To execute this, it is necessary to reliably obtain data from the sensors and learn the algorithm. Using the acquired data, we calculate the coefficients of curves. Measurements are then carried out, and different curves with their respective coefficients are adjusted to the measured data. As a result of the adjustment, we obtain temperatures. If these temperatures are very close to the actual temperatures, correct recognition is achieved.
Initially, we analyze each sensor and derive equations for calculating resistance and temperature. We then choose a suitable wireless protocol, as the system's functionality depends on communication. To ensure the accuracy of data readings, we update the circuit with a more powerful analog-to-digital converter. To simplify the system's use, we create a graphical interface designed to accept essential data for program operation. The source code is written and documented in a way that facilitates possible future work. Due to space limitations in the thesis, the parts of code and the code documentation is included, and the entire code is accessible at
https://github.com/selimdino/Recognition-of-temperature-sensors/tree/main/src.
The thesis's outcome is a complete system where water temperatures in containers are entered, measurements are taken with the sensor, and the recognized sensor is displayed, providing the option to measure temperature with it.
