Font Size: a A A

Developing And Simulation Research Of The Control Model And Control Strategy Of Diesel Engine

Posted on:2003-04-09Degree:DoctorType:Dissertation
Country:ChinaCandidate:Y HangFull Text:PDF
GTID:1102360062995216Subject:Power Machinery and Engineering
Abstract/Summary:PDF Full Text Request
Along with the even more strict requirements for the emissions of engines, how to realize the electronic controls of diesel engines has been the major task of the current research in this field. The electronic controlled diesel engine has such characteristics as more adjustable parameters, accuracy of regulation, great flexibility and freedom, etc. all of these can ensure the engine run at the best tradeoff point of economy, power and emission, consequently can make the engine get the best performances. But the developing process of the electronic control system is so complicate, need great financial and labor supports. Especially to our nation's situation, being in the process of the beginning, lacking of experiences, we surely will confront with the problems of huge lab tests, low developing efficiency and high costs. Using of the numerical simulation technology, with advanced software, to develop the so-called "Virtual Laboratory" and to model and simulate the engine and its control methods, then to compile the model and finally generate practical control code. These will obviously help us to understand and master the design steps of the control, to validate the control strategy, to optimize the control parameters and to shorten the developing period.Based on the software of MATLAB/SIMULINK, this thesis has developed a system to design and simulate the speed control strategies of diesel engine. First a quasi-linear model of 6110 turbocharged diesel engine was built, then traditional control theory such as PID were applied. By analyzing the problems that the traditional PID algorithm faced, the thought of combining fuzzy logic control with PID control was presented, thus can adopt the manual control experiences into the system. So an intelligent fuzzy self-tuning PID controller was developed, and the simulation results show the great improvement of the engine control performance, especially to overcome the disturbance (such as the change of load).Tracing the developing trends of the modern control technology, the intelligent control methods with self-learning ability, such as single neural cell control, neural network control, were also been applied on the control of diesel engine. Took advantage of the self-learning and self-adaptation abilities, the single neural cell control method was combined with the PID method, and furthermore the single neural PSD method was also brought out. Good performances were validated through simulation, thus provided an easy and feasible way for the control of diesel engine.The engine controller is in fact a multi-dimensional, complicated non-linear mapping device of the inputs and outputs. Therefore with the ability of mapping non-linear and complex functions, the multi-layer forward neural network can be used to establish these relationships based on the experimental data. This thesis first trained a BP net offline with great deal of experimental data, to establish the primary relationships of the inputs and outputs, then with the genetic algorithm the net was trained online, and eventually a neural network controller with quite good performances was built up.It's important to develop the Rapid Prototypes of the engine and the controller in the whole process; these real-time simulation prototypes can realize hardware-in-loop (HIL) test and strategy quick validation. In this paper, with the toolbox of Real-Time Workshop and xPC Target, the real-time simulation was made, thus support the HIL greatly.Research made in this thesis is the primary but important step in developing the actual controller, but the key point is the bringing forward and the experience on a new methodology, thus can support the future work with theory and software. By doing this, we can understand and master the design steps, preselect and optimize the controller's parameters, shorten the developing time, so will absolutely has great positive effects on the diesel engine electronic control engineering.
Keywords/Search Tags:Diesel Engine, Numerical Simulation, Control Strategy, PID Control, Fuzzy Control, Neural Network Control, Genetic Algorithm
PDF Full Text Request
Related items