Fuzzy infrared sensor for liquid level measurement: A multi-model approach
| Autor: | Julio C. Ramos-Fernández, Francisco Rafael Trejo-Macotela, Marco Antonio Márquez-Vera, Joel Suárez-Cansino, O. Alonso-Hernández, Jorge A. Ruiz-Vanoye, Virgilio López-Morales |
|---|---|
| Rok vydání: | 2020 |
| Předmět: |
Computer science
business.industry Fuzzy set 0207 environmental engineering Automotive industry 02 engineering and technology 01 natural sciences Fuzzy logic Automotive engineering Computer Science Applications Liquid fuel 010309 optics Transducer Engine efficiency Modeling and Simulation Control system 0103 physical sciences Calibration Electrical and Electronic Engineering 020701 environmental engineering business Instrumentation |
| Zdroj: | Flow Measurement and Instrumentation. 72:101696 |
| ISSN: | 0955-5986 |
| Popis: | The measurement of liquid fuel in automotive vehicles, is critical to prevent malfunctions or mayor failures, predict travel autonomy, ignition engine efficiency and fuel economy. In this paper we introduce a novel system to measure the level of fuel in automotive tanks. The design of a sensor for measuring the volume of fuel level based on infra-red light emitter-receiver is illustrated. The dynamics volume on fuel in cars tank, can be represented with the variation of the height level in the tank and the surface, of due to the design with mechanical tannks asymmetry of the tanks. Other trouble is the response of the transducer, normally it has non linear operations i.e., dead zone, saturation, hysteresis; among others. In order to improve the approximation fuel volume to be measured, the fuzzy modelling Takagi-Sugeno (T-S) methodology was applied to the analogical signal obtained from the hardware that compose the sensor. In order to obtain reliable data from the transducer, a closed-loop control system to characterize and calibrate the sensor of fuel flow was implemented. The calibration of fuel volume sensor contained in the tank, was also validated with laboratory flasks and a cross validation with an electronic feedback control. A model with three fuzzy sets having cubic equations as consequent, has a highest precision than by using linear consequents. For the case of study, the RMS error of the fuzzy sensor was 128 mL in a tank of 7500 mL ( 1.7 % ) . |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|
Full Text from ScienceDirect