The Research Of Complex Thermal Power Engineering System Based On Chaotic Optimization Theory

Strong coupling, non-linearity and large delay are the characteristics of complex thermal power plants, which are difficult to control. New control strategies have been worked on by domestic and foreign scholars, in order to achieve better control qualities. It is well known that control and decision-making problems can be regarded as the optimizations of structural and parameter. Therefore, developing a new optimization technology is the key step to solve the existing problems in thermal power systems.Chaos occurs in many nonlinear systems. Ergodic, stochastic and regular properties are the characteristics of chaos, which can track any state in a certain scope without repetition. Due to its significant characteristics, chaotic optimization, in recent years, has become one of the forefront projects and important issues. In this dissertation, chaotic optimization is shown to have great future in engineering applications by the detailed analysis of chaotic optimization method, and the thorough study and experiments of its applications in some typical heat power engineering systems. The main contributions of this thesis are listed as follows:Firstly, chaotic dynamics and chaos optimization theories were expatiated fully and systematically. Two typical chaotic systems and two common chaos optimization algorithms were quantitatively analyzed. By analysis and conclusion, generally existed problems of chaotic optimization were discussed. That was helpful for further studies on chaotic optimization.Secondly, aiming at the non-linear characteristics of dead-time and saturation in the turbine regulating system, chaotic optimization strategy integrated with cerebellar model articulation controller and PID combined control was proposed. This algorithm was used in turning-parameters design of the turbine regulating system. Compared with the results of the combined control method and PID control, this control method improved the system's control precision and enhanced the system's response speed.Thirdly, aiming at the characteristics such as multivariable, strong coupling, nonlinear and time-varying parameters for coordinated control system of unit power plant, diagonal recurrent neural network combined with PID control theory was researched. Meanwhile, an improved mutative scale chaotic optimization algorithm was proposed for tuning weight parameters of neural network and PID parameters to realize multivariable optimal control. Simulation results showed that this control method has many advantages such as fast response, strong robustness and good self-adaptability under different load conditions of 100% and 70%.Finally, considering the large inertial time-delay characteristic of the fresh steam temperature variations in thermal power plants, a novel PID control strategy with radial basis function network tuning based on chaotic and genetic algorithm was proposed. The intelligent PID controller was formed by cinheriting not only the advantages of conventional PID cascade controlling, but also gaining the self-study ability of intelligent controllers. Simulation results evidently showed that the control system performance is better than the conventional cascade control.