Study On The Efficient Parameter Identification Method And Platform For The Steam Turbine And Its Governing System

As one of the most important parts, the steam turbine speed governor system plays an important role in the safe and economical operation of the whole plant, through the speed and load adjusting. The models of the steam turbine and its governing system with good accuracy and rapidity are necessary to build a powerful controlling strategy for the governor system. Parameter identification not only provides accurate models for governor system, but also could be used in status inspection, fault diagnosis and performance prediction. Since the traditional parameter identification methods used in the steam turbine and its governing system have several shortages like great work load and long period by hand, a more efficient identification method is essential. Three powerful identification algorithms, an improved identification scheme and an identification platform are presented in the thesis. The main work and achievements of this thesis are listed as follows:(1) A gravitational search algorithm with variable gravitational coefficient(VGSA) and an improved particle swarm optimization- genetic algorithm(IPSO-GA) are proposed in this thesis based on the existing heuristic optimization algorithms. Test results from parameter identification of the steam turbine and its governing system demonstrate their excellent identification performance. Meanwhile, a novel identification method, called direct identification method(DIM), is introduced according to Least Square theory in this thesis. The parameter identification tests validate that DIM is an effective method with high identification performance and accurate results.(2) A novel identification scheme with several stages based on the parameter sensitivity analysis is proposed. This identification scheme not only guarantees the whole system accuracy, but also makes the subsystems parameter reasonable. Simulation experiments show that the proposed identification method performs well in identification process with high accuracy, stability and fast optimizing speed, compared with the conventional identification method.(3) An identification algorithm library is developed, including particle swarm optimizaitin(PSO), genetic algorithm(GA), gravitational search algorithm(GSA), gravitational search algorithm with variable gravitational coefficient(VGSA), improved particle swarm optimization- genetic algorithm(IPSO-GA), Recursive Least-squares method(RLS) and direct identification method(DIM). Based on the identification algorithms and mathematic model library, an integration platform is developed for the parameter identification of steam turbine and its governing system.In order to test the performance of the identification algorithms and the identification scheme, the platform is applied to the parameter identification of a 300 MW steam turbine system and a 600 MW steam turbine system. The results show that the proposed methods in this thesis perform well and have the performance with high identification efficiency and accuracy through the comparison with traditional methods. Hence, it has provided new effective ways for parameter identification of the steam turbine and its governing system.