APPROACH OF A FUZZY PID CONTROLLER INVOLVED IN RASPBERRY ARCHITECTURE PI AND PYTHON: A CASE STUDY OF A THERMAL PLANT

Resumo

Computational intelligence has several techniques that are executed in several fields for learn-
ing, perception, prediction and control. We can use computational methods allied to fuzzy logic to obtain

control meshes with a minimal degree of error. In this work we have developed a PID controller based
on fuzzy logic written in Python programming language that can be applied to microcomputers such
as Raspberry Pi for control with robustness, reliability and connectivity. The PID, or also known as
proportional integral derivative controller is a method of control of processes that groups computational

executions, generating minimum degree of error. This control was modeled in Python from object ori-
ented programming and with mathematical and fuzzy libraries, executed in a microcomputer, that has

GPIO‘s (General Purpose Input / Output) which are an interface of data input and output directly on
the board and can be connected to sensors / actuators and M2M (Machine to Machine) communication,
which facilitates the final application in industrial or residential plants, crossing concepts of IoT and
Industry 4.0. The PID controller coupled with fuzzy logic produces a fuzzy-PID controller, robust and
dynamic based on IF-THEN rules and pertinence functions, capable of working with linguistic variables
that model physical media based on human cognition. The set of rules used in the code can be extended
to various models and plants, such as in temperature control. The data obtained can be exported and
analyzed in software such as MATLAB, SCILAB, thus obtaining graphs and analyzes in real time, being

able to perform the parallelism with analysis and simultaneous control. We can apply the control algo-
rithm in an electric thermal plant where the desired temperature and the current temperature are input

data of the plant, these values are fuzzified and are processed mathematically based on the system rules,
inference machine and the fuzzy-PID operation, the values where they are transformed into a PWM
(Pulse-Width Modulation) control signal applied to an electric resistance. As one of the goals of control
and automation is to try to minimize human interference in a machine or procedure, making the system
more optimized, safe and efficient. The computational control establishes relations with the automation
area in order to transform processes control as more ”intelligent” and implement decision making so that
their choices are progressively more reliable.