Experimental Development of Fuzzy Controllers for Thermal and Pneumatic Processes

Abstract

In this project, a Fuzzy control system is proposed in an industrial process training module with two independent systems between them, one thermal and the other pneumatic. The control algorithm is developed in Python language v3.6 executed by a Raspberry Pi B+, both controllers depend on the error and change in error that are updated in times of 2 s and 1 s, for temperature and pressure respectively, communication with the plants uses A/D and D/A converters, the thermal Fuzzy was analyzed with three temperature references [50,100 and 150]°C, with a rise time of 191 s, 360 s and 505 s; steady state error of 5.5%, 0.7% y 0.7%, in the pneumatic system the speed of change between references is evaluated from 10 psi to 15 psi varying the activation of the compressor at the beginning of the experiments, the settling times obtained are 111 s and 106 s, with the compressor off the result is 116 s and 88 s, besides a maximum excess of 13% with inherent oscillations to the type system that are in an acceptable range.

Current Trends in Incubator Control for Premature Infants with Artificial Intelligence Based on Fuzzy Logic Control: Systematic Literature Review

“…The fuzzy design model algorithm is divided into four stages, namely Fuzzyfication, Inference Engine, Rule selector and Defuzzyfication. Each input variable and output variable are represented by MF and regions graphically [90]. Some uses of the Fuzzy Tsukamoto method, namely, research on the Fuzzy Tsukamoto method with the system used as a diagnostic tool for liver disease with the use of Android, the conclusions of the research conducted are as follows: (1) Expert system as a diagnosis of liver disease can be designed and implemented for use with well, (2) an expert system for diagnosing liver disease can be implemented using the Fuzzy Tsukamoto method by going through five main processes, namely fuzzyfication, determining the alphapredicate for each Rule, calculating the z value of each Rule, multiplying the alpha predicate by z in each Rule, then defuzzification by dividing the number of alpha-predicate times z by the number of alpha-predicate, (3) this expert system can help diagnose liver disease or recognize common symptoms of liver disease, a system that is built can be used well by the user [91]; [92].…”

“…Traditional logic, which can only be expressed as 0 and 1 or "yes" and "no," was developed to deal with difficulties that could not be resolved by traditional reasoning (crisp logic). Fuzzy logic accepts values that fall in the range of "yes" and "no" [89][90][91][92]. Because each baby using the plant's incubator has individual needs.…”