Automatización del sistema de control de temperatura del refrigerante, usando controladores lógico difusos, para el motor de una locomotora diésel-eléctrica de Ferrocarriles del Ecuador Empresa Pública.

Abstract

The design of a fuzzy logic controller for the control of the coolant temperature of the diesel engine of one of the electric-diesel locomotives was implemented in the company Ferrocarriles del Ecuador (FEEP). The Takagi Sugeno Kang (TKS) inference method was used in a first-order system that reduced the system establishment time. The method previously used in Electric Diesel Locomotives to control the temperature of the refrigerant (TR) was an on-off control through the use of contractors, which activate the fan at a specific temperature. The main problem with this type of controller is that it does not keep the system stable, and, in the case of FEEP, this variation causes the TR of the diesel engine (DE) to tend to go outside the optimal operating range. This causes increased wear on the moving parts in the long term. A database was acquired that was collected by a 32-bit embedded system and stored on a PC, to obtain an approximate mathematical model, through the installation of sensors. The derived variables are Coolant Temperature, Oil Temperature, DE Revolutions, and Fan Current. Furthermore, a PID controller was designed using the Ziegler and Nichols technique to adjust the Kp, Ki, Kd gains. From which the expert knowledge was taken for the design of the fuzzy logic controller. When validating the results of the implementation, the fuzzy logic controller presented a more stable TR control (80.17ºC), and a standard deviation of 0.6ºC. The use of artificial intelligence techniques for finer tuning of singletons was recommended for future analysis.