Study On The Optimum Initial Operation Pressure Of Steam Turbine Unit Based On The Improved Krill Herd Algorithm

In recent years,with the change of the power consumption structure and the rapid development of the national economy,more and more large capacity units of thermal power plant are involved in startup-halt for adjusting the peak output.The steam turbines run under the condition of long-term variable load,which makes the thermal economic performance significantly reduced.In order to maintain the best running state of steam turbine under the long-term variable conditions,the initial pressure value of steam turbine must be optimized.The optimal initial pressure will be taken as the main steam pressure.The results show that the method can reduce the heat rate of steam turbine effectively and can guide steam turbine much better towards safe and economic operation.To optimize the initial steam pressure,two major problems need to be solved: 1)building accurate turbine heat rate model;2)determining the optimization program.Thus,this paper focuses on two issues to study,its main contents are described as follows:Firstly,for the shortcomings of the Krill Herd algorithm which it is easy to fall into local optimum and low convergence accuracy,this paper proposes two improved method:an Opposition Adaptive Krill Herd algorithm(OAKH)and an Ameliorated Opposition Adaptive Krill Herd algorithm(A-OAKH).In order to test the validity of above two improved algorithm,they are applied to 10 classical benchmark functions.Compared with Biogeography-Based Optimization(BBO)and KH,experiments show that OAKH and A-OAKH are validity,and then A-OAKH owns better search performance than OAKH.Secondly,the A-OAKH algorithm is adopted to optimize the parameter of Fast Learning Network(FLN),which is used to set up the prediction model of heat rate.The main function of the A-OAKH algorithm is to optimize the parameters of FLN to find out the optimal model parameters.And the optimal model parameters are brought into the FLN to build the prediction model of heat rate.Finally,in the basis of A-OAKH-FLN model,the A-OAKH algorithm is applied again to optimize the main steam pressure value that the unit is at the lowest heat consumption rate in the feasible pressure range.In other words,the purpose is to seek theoptimal initial pressure value of each load of steam turbine.Experiments show that the heat rate is reducing.The optimized turbine sliding pressure operation curve was compared with the design cuve,and the result shows that the optimized sliding pressure operation curve could be reduced the heat rate effectively and it could guide steam turbine much better to safe and economic operation.