Study On The Reservoir Operation Chart Of Hydropower Station And Short-term Optimal Operation

The routine operation chart of hydropower station is the basic dependence to make certain the operation of the reservoir. The upper and lower basic operation lines are the border terms for the power station running to deliver guaranteed output. When water level is on the upper basic operation line and the station delivers guaranteed output, it reflects conservation of the time-course reservoir operation; When water level is on the lower basic operation line and the station delivers guaranteed output, it reflects risk of the time-course reservoir operation. Because of the limitation of the historical hydrological data and the randomness of the surface flow, the existing surface flow data can't totally reflect the objective law of the hydrology characteristic in the valley. So the operation chart made with limited hydrological data has its localization shortcoming. In actual operation, we find we can improve the operation benefit of the reservoir through properly modifying the reservoir routine operation chart according to experience. Just based on what have been stated on the above, this text probe into setting up corresponding mathematics model to optimize the routine operation chart in order to improve the average generation and economic benefits of the reservoir under the premise that guaranteeing rate does not change. Obvious, this study has important praxis value and wider perspective for apply.There are many methods to solving interior economical operation of the hydroelectric power such as equal incremental equations method, dynamic programming and non-linear programming. Because the water consumption curve doesn't take on concave shape in the bound of the unit ratings, this method can't be used directly. Non-linear programming need to make equation by curves . It is difficult to solve the problem and hard to get the full optimized answer. Dynamic programming needs to disperse the water consumption curve and the more the dispersed dots is, the higher the precision is. But too many dispersed dots will take up too much computer memory, which leads to "the dimension calamity " especially to the multi-dimensional question.This text adopts improved genetic algorithm to solve hydroelectric power plant's economical operation This method can realize the overall optimum situation when started from several initial values and searched along many routes. Except that, it has a good adaptability and it is a kind of effective self-adaptation random searching algorithms. To the calculating time needed, when the number of power plant or machine group is few, the two methods almost have no difference, but when the number of power plant or machine group increases, genetic algorithm is better than dynamic programming. In order to verify the validity of the improved GA, the author has worked out GA and dynamic programming algorithm procedure and compared the two solved results. The results show that the improve GA has better capacity of self-adaptation searching optimal solution which can get the overall optimum solution.To the cascade hydropower station group, for the existence of time lag of the same volume between the upper station and the downstream station, while making the cascade hydropower station operation plan for one day, the influence of the time lag between the sequential days should be considered. It could be well understood, if there isn't the lag time, all the volume of water in the same period and take part in the equation calculating, then the optimum answer is real answer. But for the existence of time lag, we can't pursue the most benefit of the current day only simply.The text regards the Ge-he-yan and Go-ba-zhou cascade hydropower station system as aIVexample, and set up the lag electric power of the upper station max-mode under the restriction of the system load and the water quantity of the upper reservoir to reflect the inflection of the power between the two days. For the double limitation of the system load and water quantity , the question can't get the solution probably. For this, the text...