Study On The Load Optimum Distribution Of Header System Combined Heat And Power Units

Along with the structural reform of electric industry from planned economy to market competition, such reasons as the seperation of power plants and electric grid, price competition, and the continued growing of coal price, forcing the power generation enterprises to improve their operation efficiency. Load distribution optimization can effectively improve the efficiency of power generation enterprises without changing their hardware, just redistributing the load instructions of the units. So it caused a lot of power engineers'attention. However, current research of load distribution mainly are focusd on the units in condensing power station, and research about combined heat and power generating units was very poor.This paper's purpose is to take research on the load optimum distribution of combined heat and power generating unit in header system. The paper takes detail analysis on the thermal property of such combined heat and power generating units as back pressure units, extraction condensing units, extraction backpressure units and straight condensing units, then obtains the linear unit consumption characteristics by the unit performance diagram. The linear model of combined heat and power generating unit cannot describe the real performance of units correctly. Based on multiple linear regression theory and abundance of unit operation data, the paper obtains the non-linear characterisic equation of combined heat and power generating unit with polynomial fitting method. This model can describe the real performance of the units with higher accuracy than the existing model. Then it establishs the mathematical model of the load distribution of combined heat and power generating units in head system.After analysing the merit and demerit of the existing traditional optimizationalgorithm, mathematical programming and intelligent algorithm to solve the load distribution model of combined heat and power generating units,it's found that differential evolution algorithm has such advantages as few control parameters, no spetial requirement about objective function and out-standing global convergence. so algorithm is chosen to solve the load distribution problem of combined heat and power generating units in this paper. The basic differential evolution is improved by adopting time-varying diminishing scaling factor and time-varying increasing crossover probability factor. The example analysis proves that the improved differential evolution algorithm is better than genetic algorithm and basic differential evolution algorithm.The most common load distribution model take the upper and lower bound as fixed value, which is not fit for the real operating mode. This paper proposes the calculation formula of the dynamic bound of heat and power load in combined heat and power generating unit. Then it establishs load distribution model of combined heat and power unit, considering this issue, and analysis the influnce of the changement of model.This research is supported by the 2009 annual Harbin excellent discipline leader capital project,"The automatic control technology of header steam pressure and its open control sofeware development"(2009RFKXG005).