Application Research On Multivariable Advanced Control In Complex Thermal System

All the typical complex thermal systems are multivariable and strongly coupled. Taking traditional system control scheme combine with feedforward compensation control, cascade control and other control strategies, decentralized control structure which is composed by multiple SISO feedback control loops and based on PID linear control algorithm , is still used widely. However, this method cannot ensure long-time normal operation, and the plant cannot run at optimum state. At the same time, a lot of energy is wasted and equipment fault is increased. Economy and security are affected seriously. So it is of very important theoretical significance and practical value to study multivariable advanced control methods which are suitable for the complex thermal systems, and take it into automatic control and optimal operation effectively. This paper reviews the advanced control methods and application research situation of thermal systems such as power plant ball mill storage pulverized coal system, and then studies the application research of the multivariable advanced control methods for the complex thermal systems.Chapter 1 is the introduction of the paper. Firstly, it describes the background and the significance of the research, and then reviews the situation and development of advanced control methods for the thermal systems such as the power plant ball mill storage pulverized coal system at present. Lastly, it puts forward the primary work based on the problems existing at practical controls of complex thermal systems.Chapter 2 studies the multivariable system decoupling control method for the ball mill storage pulverized coal system. Based on the chief introduction of production flow for the ball mill storage pulverized coal system, we make the coupling analysis and give the loop matching method of multivariable system. Loop matching should not only consult relative gain array (RGA), but also have to consider the technology and control requirement of practical process to choose more appropriate input and output matching. After choosing the suitable matching, we give the decoupling control scheme of the ball mill storage pulverized coal system based on the dynamic characteristics of each loop. This scheme combines feedforward compensation decoupling with partial decoupling well, and we make the system design and simulation of it. From the simulation results, we can see that the partial feedforward compensation decoupling control can preferably separate the negative pressure loop, and the controlled object is simplified to 3×3 from 2×2.Chapter 3 studies a decentralized relay feedback auto-tuning strategy of multivariable PID controllers for complex thermal systems. Firstly, this part studies the existing auto-tuning methods of multivariable PID controllers from decentralized relay feedback. Then combining the coordinated control object characteristics of coal-fired unit, the author proposes an improved design method for the coordinated control system with feedforward decoupling. The simulation verification results show that the new method can make systems obtain preferable control effects.Chapter 4 combines the iterative feedback theory and the relay auto-tuning method organically and proposes an auto-tuning for multivariable decentralized PID controllers—iterative relay feedback method. Compared with the existing design methods, this one has the advantages of low demands to prior knowledge and model of the controlled object and it has clear physical conceptions, so it is convenient for engineering application. According to the simplified object of the power plant ball mill intermediate coal pulverizing system, this part does control system design and simulation verification by the proposed method and the results show that the controller parameters could be tuned preferably and the system could get a satisfied effect.Chapter 5 discusses multivariable DMC algorithm and proposes an auto-tuning method of DMC parameters. For a typical TITO system, we make some simulations and analyze the influence of each parameter. In the context of simulation, the proposed method is applied to the simplified ball mill storage pulverized coal system, and the result shows that this tuning strategy could control outputs effectively and satisfy industrial control requirement.Chapter 6 presents the application software development. In this part, a predict control engineering software package is designed with VC++ based on PC platform, containing a step response module, a control matrix module, a parameter input module, a system simulation module and other modules. It can be used in WINDOWS systems on PC platform, and also, it has low demands for hardware, so it has a good flexibility. Using this software, we could decrease the tuning time of control parameters and increase the performance of process control.At last, the author summarizes all the works which have been done in the thesis, and look forward further explorations and research questions in the next work.