Research On Fractional-order Model And Control Of Combustion System Of Supercritical Power Plant

With the advantages of supercritical power plant over conventional subcritical drum plant, such as higher efficiency, lower emissions, the supercritical units are becoming the major part of the thermal power plant. Fuel control system of supercritical units is the important partant of whole unit control system. The control effect of the fuel system has a direct relationship with the units’ security and efficiency. However, the characteristics of great inertia, large delay and variability of coal quality make it difficult to operation and control, so it is meaningful and necessary to study the model and control algorithm of the fuel system of a supercritical power plant.Heat storage coefficient of the supercritical power plant is not only the requirement to obtain it’s mechanism model but also an important parameter that is helpful to understand the boiler’s dyanmic characteristics. The heat storage of supercritical boiler main consists of metal’s and working fluid’s, but the metal’s heat storage changes little when the supercritical unit is in variable load operation. So solving the working fluid’s heat storage has the more important and practical meaning. In fact, it is difficult to achieve water wall’s heat storage cofficient, because the phase transformation point of water vary with different operating conditions. So it first needs to get the length of subcooled water section, evaporator section and superheated section. After the lengths of the three sections were solved, we used the mass, volume, and energy balance equations to obtain the water wall’s heat storage cofficient and superheater’s heat storage cofficient respectively. A 660 MW ultra supercritical power plant is taken as an example and used the boiler’s designed data to get the heat storage coefficient at different operating points. It was found that the superheater’s heat storage accounts for about 70%~85% of the whole boiler’s heat storage, and the water wall’s heat storage is about 15%~30%. Above all, the boiler’s heat storage shows little change under the whole operating conditions. These important conclusions not only provide a theoretical basis for its application in the fuel control system, but also is significant to control the fuel system quickly and precisely when the load changs.The fractional-order models of the pressure and intermediate point temperature in supercritical power plant were obtained based on the indentification methods both in frequency-domain and time-domain. In the frequency-domain identification process, LCR method was used to convert the sample data from time domain into frequency domain, finally the fractional order model were acchieved according to the Levy’s technique. In the time-domain identification process, we combined the G-L definition and the partical swarm optimization to get the fractional order model. From the identified results, we can see the acchieved fractional-order modes have good fitting effect to the sample data, so they suggest the indentification methods both in frequency-domain and time-domain are available.Using the simple mode of the supercritical power plant, a PIλDμ controller was designed to compare with the traditional PID controller in the combustion control system during the load changing process. The result showed the PIλDμ controller has advantage over PID. In addition, a fractional-order internal model controller(IMC) is designed on the basis of identified model. From disturbance tests, we knew the fractional-order IMC is better than integer-order IMC. At last, we can conclude that the control algorith based on fractional-order theory is promising.