Validación experimental del modelo de cálculo de consumo de combustible desarrollado por el Toolbox QSS de Simulink

Abstract

The objective of this research was to validate the use of a vehicle model made in the Simulink QSS environment from experimental tests that allow knowing an equation with the possibility of reaching an approximation of the actual fuel consumption in selected vehicles in Riobamba city to know the reliability in the use of the simulation method. Based on two experimental test methods, the OBD II reading method under the Torque Pro interface through PID and the gravimetric method, methods that through the Pearson coefficient are related to the fuel consumption in l/100km obtained in the QSS environment to validate the enunciated vehicle model. For the correlation of the experimental method and QSS method, it focuses on knowing the R2 of said comparisons where the Gravimetric method vs. Simulink QSS method has a high tendency to a straight line since in Chevrolet Luv D-max, Hyundai i10, Mazda BT- 50, an R2 value of 0.817 obtained; 0.865; 0.974; respectively, therefore, the level of confidence that using their equations is analyzed to know the approximate value of the actual fuel consumption utilizing RMSE and the maximum deviation of each equation concerning to the QSS model, indicating an acceptable level of confidence of up to 90% since that the RMSE of the proposed vehicles is between 0.10. In this way, it shows that the use of two experimental methods allowed to have a comparative diagram of results and the selection of the one closest to reality, to validate the method of the Simulink QSS environment and, as a result, achieve a consumption closest to the reality from the simulator vehicle model in a more straightforward way without the need to do experimental tests again.