具有可变爬坡率约束的电力系统优化与精确调度

The project is aimed to realize the optimization based accurate scheduling for power system considering variable ramping constraints, and systematic study is carried out from three aspects including refining of the system operation costs, formulating an optimization based accurate scheduling models considering variable ramping constraints, and designing optimization algorithm with high quality and efficiency. Firstly, a double integrator system dynamics model is formulated to reveal the change law of generation costs with generation level, and a functional relationship between the optimal climbing strategy and the instantaneous power at the start and end of that period is established accordingly, so that the system operation costs are refined by taking into account the ramping costs which can guide the rational allocation of climbing parameters, and then lead to the establish the variable ramping constraints. The difficulties in calculation of the ramping costs which caused by the complex functional relationship between the ramping costs and climbing strategy is overcome. Secondly, under the conditions that the variation of generation level is bounded and continuous, an optimization based accurate scheduling model which considering variable ramping constraints is formulated based on the functional relationship built above. Hence solve the problem that the schedule obtained with the traditional scheduling model may be unrealizable. Finally, a systematic approach is developed, so the optimization based accurate scheduling model considering variable ramping constraints, i.e. a continuous-time optimal control problem which has a complex and dynamic operational constraints, is successfully converted into a nonlinear programming problem. Therefore, the model can be solved exactly and efficiently. This project is of great significance for improving the scheduling levels of power system, and promoting sustained and rapid development of new energy utilization in China.

本项目以具有可变爬坡率约束的电力系统优化与精确调度为研究目标，从生产成本精细化，调度模型精确化，求解算法优质、高效化三个方面展开系统性研究。首先，通过建立双积分系统动力学模型，揭示生产成本随负荷的变动规律，建立最优爬坡策略与调度时段首、末时刻输出功率的函数关系，从而将爬坡成本计入生产成本，进而引导爬坡参数合理配置来建立可变爬坡率约束，有效解决爬坡成本与爬坡策略之间存在复杂泛函关系对计算爬坡成本造成的困难，实现生产成的本精细化；其次，在机组输出功率连续有界变化的条件下，基于上述函数关系建立具有可变爬坡率约束的电力系统精确调度模型，解决传统模型调度结果可能无法实现的问题；最后，研究将该模型-一个具有复杂动态运行约束的连续时间最优控制问题，转化为非线性规划问题的系统性方法，进而实现对模型优质、高效求解。本项目对提高现有电力系统调度水平，促进我国新能源高效利用具有重要意义。

为了规模化开发利用新能源，应对高不确定性新能源的大规模接入，针对我国贫油、少气、富煤的能源结构布局，需要探索充分发挥火电与新能源发电的互补机制与理论方法。本项目旨在从提高电力系统对不确定性新能源的经济规模化消纳能力出发，通过研究火电机组对新能源电力消纳支撑能力与成本分析，1、建立了机组出力方式与机组爬坡成本之间的关系；2、机组出力方式与机组的快速深度变负荷容量之间的关系；3、在考虑能量与备用容量可精确实现性的基础上，将传统的电力系统优化与调度过程中的机组的能量约束与备用容量约束转化为调度时段首末时刻机组的瞬时输出功率。基于这3个研究结果，实现在电力生产调度过程中，考虑到机组的快速深度变负荷能力，能够在电力系统生产调度过程中，同时实现对机组的快速深度变负荷能力进行优化，从而实现挖掘火电系统备用贡献潜力，在不启停任何发电机组的条件下，提高系统对清洁能源消纳能力。基于动态规划求解框架下，实现调度模型中离散变量与连续变量的解耦，并对纯连续优化问题，建立基于多种群粒子群和布谷鸟的联合寻优算法，从而提高了动态多种群粒子群算法的全局搜索能力和算法效率。