Random step maneuver algorithm with normally distributed starting times

The random step maneuver with uniformly distributed starting times has the disadvantage that it cannot focus the starting time on the more efficiency time. It decreases the penetration probability. To resolve this problem, a random step penetration algorithm with normal distribution starting time is proposed. Using the shaping filters and adjoint system method, the miss distance with different starting times can be acquired. According to the penetration standard, the time window ensuring successful penetration can be calculated and it is used as the 3σ bound of the normally distributed random maneuver. Simulation results indicate that the normally distributed random maneuver has higher penetration probability than the uniformly distributed random maneuver.

考虑风险和碳交易机制的微电网分布鲁棒优化调度

提高微电网风电的消纳能力是实现双碳目标的重要手段。该文构建了一种考虑极限场景和净负荷条件风险价值的微电网低碳分布鲁棒优化模型。首先,针对源-荷不确定性构造基于Wasserstein距离的概率分布模糊集,并考虑极限场景修正该模糊集,以提高模糊集的鲁棒性,减少场景数目,提高求解效率。其次,以包括供能成本、阶梯式碳交易成本和风险成本的微电网日运行成本最小为优化目标,建立分布鲁棒优化调度模型,并利用分段线性近似法和对偶理论将模型转换为线性规划问题求解。最后,通过算例仿真进行验证,结果表明,考虑极限场景的分布鲁棒优化方法具有较强的鲁棒性;引入阶梯式碳交易成本可以有效减少微电网的碳排放量,并提高风电的消纳能力;净负荷条件风险价值可以有效平衡微电网运行的经济性与风险性。