Improved Genetic Algorithm Based Innerplant Economical Operation For Hydroelectric Power Station

Researching the innerplant economical operation of hydroelectric power station is of great practical significance for the further development of hydropower,solving the isues such as the power shortage in China,the obvious contradiction between supply and demand etc. According to the actual operation and the key questions of the hydroelectric power station which needs to take into account,the thesis improves the model of Genetic Algorithm mathematics,the methods of optimization calculation and programming of the innerplant economical operation in the hydropower station.And integrated with the example of the Liujiaxia Hydropower Station,it clarifies a concrete realization of process on the use of Genetic Algorithm for the innerplant economical operation in hydropower station.The major research works are as follows:1.Improving the traditional Genetic Algorithm's mathematical model for the innerplant economical operation of hydroelectric power station.The model is composed of two sub-models,namely the optimal unit commitment(UC)model and the optimal load distribution(LD)model according to characteristics.As for the LD,it was generally considered only the units minimum and maximum output operating limits in the past.On the basis of this,the thesis increases by the restrictions of avoiding cavitation and vibration zones which are non-safe range bounds.And as for the UC,based on considering only the fixed operating output of the model,it adds to take account of the cost of start-up/shut-down of hydraulic units,as well as minimal uptime and downtime constraints to make unit commitment plan.So the mathematical model has been greatly improved than ever before.2.Improving the optimal methods of genetic algorithm for the model.For the LD,it was generally generated randomly initial solution in the past.In this thesis,the formation of initial populations is joined with cavitation and vibration bounds of units to reduce the search space and time of them.And for the UC,the initial populations are designed minimal uptime/downtime,minimal output power,and other constraints to guide the initial populations generated,substantially saving the time that randomly generated feasible solution.At the same time,the enactment of the optimal hydropower Station Operation of the program is a dual decision-making process,namely the UC and the LD.According to the theory that an arbitrary sub-strategy of the optimal strategy is optimal,the decision-making of the UC is transferred the LD tables to solve,which reduces the complexity of the calculation,and simplifies the calculation procedures.3.On the calculation software,the mixed programming of Visual C++ and MATLAB is deeply studied in the economical operation.Meanwhile,the interface programming of Visual C++ and Microsoft Access database based on ADO(ActiveX Data Object)is studied also. And the procedures of genetic algorithm are made for the innerplant economical operation of the Liujiaxia Hydropower Station.4.To verify the model and the optimal methods proposed in the paper to be feasible and effective,the thesis makes optimal calculation with them for the operation of the Liujiaxia Hydropower Station.The calculation is accordence with the scheduling requirements of the Hydropower Station on March 18,2008.Then the calculation result is compared with the actual operation of scheduling.It showed that:under the 109.96m head in the day,it saves 106.645×10~4m~3 than the actual operation,and improves the efficiency of 4.12 percent.In conclusion,the improved genetic algorithm model and the algorithm are feasible and effective in the economical operation of hydropower station.And a new idea based on Genetic Algorithm is provided for the innerplant economical operation in hydroelectric power station.