Research On Practical Method Of Short-term Scheduling Of Cascaded Hydropower Reservoir Systems

With the development of hydropower systems,more and more complex demands are encountered by hydropower scheduling.Short term hydropower scheduling is a complex nonlinear optimization.Normally,a hydropower system schedule is formulated as dynamic programming,progressive optimality,network flow method,linear programming,mixed integer programming,genetic algorithm,particle swarm optimization etc.,to optimize a preset objective function such as profit maximization,energy maximization,or maximum peak shaving.The solution for large scale systems is complex and time consuming.Unfortunately,the short term schedules are often required to be made in very short time,which confine the use of mathematic programming methods.This problem needs to be solved.Taking Case study of hydropower system in Hongshui River operated by China Southern Power Grid(CSG)as example to study how to formulate the method of short-term power schedules for cascaded hydropower systems.In the paper,the main work as follows:(1)The short-term hydropower scheduling problem in hand is very complex.The hydropower system has quite complex topologies with many cascaded reservoirs in the same river system.The reservoirs may have very different storage capacity with significant water travel time that makes the decisions coupled between several time steps.In other words,the decisions in one time period have strong impact on what's possible to do in later time steps.Beside which,one of the most difficult aspects of studied problem is a great deal of period coupling constrains including power ramping constraints,minimum continuous generating periods,minimum periods of a power up/down.So,the problem belongs to the multi-objective min-max mixed integer problem with complex constraints.We intend to resolve the model by a successive local search approach to obtain near optimal solutions.To the local search approach,an improving solution is found by searching the “neighborhood” of current solution.For this problem,the neighboring solutions are transformed from “current solution”,and it improves computational efficiency.(2)To obtain near optimal schedules of hydropower systems effectively without solving complex mathematic programming,a conceptual scheduling method is proposed.Without an objective function,the basic scheduling concepts of saving water,servicing for power system balance,coordination of up and down stream reservoirs,together with most of the constrains of optimization models are used to direct the scheduling of cascaded hydropower systems.The method is composed of a series of operations for whole time horizon requirements such as final reservoir level,total energy or total turbine water of a plant,total energy of several cascaded reservoirs,and reducing spill water.To achieve one target,a successive power scheduling method is proposed in which temporal order of power rise/fall based on load characters,and a local scheduling algorithm can satisfy period coupled constraints are combined.Case study of hydropower system in Hongshui River operated by China Southern Power Grid(CSG)shows that the proposed method is effective in obtaining power schedules and complex power distribution rules can be achieved.(3)For a cascaded system or a group of reservoirs,sometimes,the total energy is given as a control target with no constrains on reservoir final storage,total energy or turbine water.There are two basic rules to distribute the total energy to each reservoir and each period: sorting rules for period and advantage for store energy spatially.Under the basic rules,a coordination mode is proposed to distribute total energy of a cascaded system or a group of reservoirs.The study case shows that the method can directly reflect the intention of the opertor in the short-term power schedule.