A Fu zzy PID Co n tr ol l er fo r Ex hau s t Gas Rec i r cu lat io n Sys tem Zhao Wenrui 1 ,Pan Haibing 2 1 Department of Mechanical and Electronic Engineerin g ,Beijing Modern Vocational and Technical College 2 College of Engineering ,Huazhong Agricultural University Keywords: Fuzzy PID ; Optimization; Gasoline Engine Abstract. As environment problem is one of the world issues, the emission limits for vehicle are becoming stricter and s tricter. To fulfill the requirements, many new technologies are adopted, such as the technology of combustion controlling, EGR, the three way catalytic converter for gasoline engine and the particulate trap for the diesel engines. This paper focus on develop a suitable controller for the EGR system of a 1.5L gasoline engine. At first, a GT-power model is built and verified, and then a PID controller and a fuzzy PID controller are established by Simulink and coupled with GT-power model. Comparisons between the two strategies are investigated. 1.Introduction There are two difficulties in the technology of EGR for gasoline engine: first, the HC and CO emission increases contrarily if the EGR rate controls unsteadily and improperly; second, the EGR affects the stability of the engine when it is in its low power output, and also decreases the torque when the engine in its high power output. The work conditions for gasoline engine changes frequently. In order to satisfy the demand of emission and power, an excellent and steady EGR rate control system is demanded. Xie Hui studied a control system that coupled the EGR and the internal residual exhaust gas for a gasoline HCCI engine. Peter Shih dev eloped a nov el reinforcement learn ing-based du al- control methodolog y adaptive NN (neural network) controller to control the EGR operation of a SI (spark ignition) engine. The controller is stable under higher EGR conditions [8]. In this paper, a fuzzy PID EGR control strategy for a gasoline EGR system is developed. First, a GT-power model was built and verified for the gasoline engine. Then a PID and a fuzzy PID control model are constructed based on the GT-power model. The step signal, the real time signal and interference signal are used to test the response and anti-interference performance of the two models. 2.The Engine Model The engine is a four-cylinder, 1.5L gasoline engine. In order to simplify the research, people tend to use the mathematics functions [9-11] or some approximate models to represent the engine performance. For example, to construct an automatic NO control system, Yoshio Yoshioka built a neural net to learn the relation among the engine output power, EGR and NO concentration which represents the engine system [12]. Studies reveal that such an approximate model has a simpler structure, but a mass of experiment data is needed to train and adjust it. Otherwise it could not avoid errors. A GT -power model is used here to represent the engine system The GT-power mode was built based on th e engine parameters, as shown in Fig. 1. The performance of the GT -power model is verified by the experiment data. It is shown in Fig. 2, the errors between the simulation and the experiment results are less than 5%. 32
8
Embed
A Fuzzy PID Controller for Exhaust Gas Recirculation System (2)
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
8/12/2019 A Fuzzy PID Controller for Exhaust Gas Recirculation System (2)
Abstract. As environment problem is one of the world issues, the emission limits for vehicle are
becoming stricter and stricter. To fulfill the requirements, many new technologies are adopted, suchas the technology of combustion controlling, EGR, the three way catalytic converter for gasoline
engine and the particulate trap for the diesel engines. This paper focus on develop a suitable
controller for the EGR system of a 1.5L gasoline engine. At first, a GT-power model is built and
verified, and then a PID controller and a fuzzy PID controller are established by Simulink and
coupled with GT-power model. Comparisons between the two strategies are investigated.
1. Introduction
There are two difficulties in the technology of EGR for gasoline engine: first, the HC and CO
emission increases contrarily if the EGR rate controls unsteadily and improperly; second, the EGR
affects the stability of the engine when it is in its low power output, and also decreases the torquewhen the engine in its high power output.
The work conditions for gasoline engine changes frequently. In order to satisfy the demand of
emission and power, an excellent and steady EGR rate control system is demanded. Xie Hui studied
a control system that coupled the EGR and the internal residual exhaust gas for a gasoline HCCI
engine. Peter Shih developed a novel reinforcement learning-based dual- control methodology
adaptive NN (neural network) controller to control the EGR operation of a SI (spark ignition) engine.
The controller is stable under higher EGR conditions [8].
In this paper, a fuzzy PID EGR control strategy for a gasoline EGR system is developed. First, a
GT-power model was built and verified for the gasoline engine. Then a PID and a fuzzy PID control
model are constructed based on the GT-power model. The step signal, the real time signal and
interference signal are used to test the response and anti-interference performance of the two models.
2. The Engine Model
The engine is a four-cylinder, 1.5L gasoline engine. In order to simplify the research, people tend to
use the mathematics functions [9-11] or some approximate models to represent the engine
performance. For example, to construct an automatic NO control system, Yoshio Yoshioka built a
neural net to learn the relation among the engine output power, EGR and NO concentration which
represents the engine system [12]. Studies reveal that such an approximate model has a simpler
structure, but a mass of experiment data is needed to train and adjust it. Otherwise it could not avoid
errors.
A GT -power model is used here to represent the engine system The GT-power mode was built based on the engine parameters, as shown in Fig. 1.
The performance of the GT -power model is verified by the experiment data. It is shown in Fig. 2,
the errors between the simulation and the experiment results are less than 5%.
32
8/12/2019 A Fuzzy PID Controller for Exhaust Gas Recirculation System (2)