随着新能源占比逐渐提高，作为辅助新能源并网的储能技术受到广泛关注。氢储能和燃料电池技术已被我国列为战略性能源技术，并积极应用到市场和企业发展中。电解水制氢可以平抑可再生能源并网带来的波动，帮助电网削峰调频；燃料电池作为提高能源转换率的发电装置，同时具备噪音小、无污染等优势，是消纳可再生能源的有效方式之一。对现有储能技术的发展现状和优势进行对比分析，着重介绍了氢储能系统目前关键技术和研究方向，并对氢储能商业化发展进行展望。

With the gradual increase in the proportion of new energy, energy storage technology, as an auxiliary new energy grid, has attracted wide attention. Hydrogen energy storage and fuel cell technology have been listed as strategic energy technologies in China, and have been actively applied in the market and enterprise development. Hydrogen production from electrolytic water can stabilize the fluctuation caused by the connection of renewable energy to the grid and help the power grid to cut peak and frequency modulation. As a power generation device to improve energy conversion rate, fuel cell has the advantages of low noise and no pollution, it is one of the effective ways to absorb renewable energy. This paper made a comparative analysis of the development status and advantages of the existing energy storage technologies, the key technologies and research directions of hydrogen energy storage system were introduced emphatically, and the commercial development of hydrogen energy storage was prospected.

针对微电网内源荷两端匹配性较差的问题，本文以含径流式小水电的水风互补微电网为研究对象，提出了一种考虑需求响应的优化配置方法。在负荷端，构建了一种动态分时电价机制，对峰谷时段进行了动态划分，在引入替代弹性的基础上，以负荷与可再生能源发电差值累计和最小为优化目标，同时考虑用户参与需求响应的满意度，对动态分时电价中的峰谷电价进行寻优，计算需求响应后的负荷；在源端，考虑到水风之间的互补效应，提出了一种互补度评价指标，并在计及需求响应参与的基础上，建立了以年运营收益和水风互补度为目标函数的优化配置模型。采用改进粒子群算法结合目标规划法求解，仿真结果表明，将价格型需求响应应用到微电网优化配置中，能够将源-荷两侧有效结合，提高可再生能源消纳水平，在保证用户利益的同时提高微电网经济性。

Aiming at the problem of poor matching between the internal source and load of the micro-grid, this paper takes the hydro-wind complementary micro-grid with run-off small hydro power as the research object, and proposes an optimized configuration method considering the demand side response. At the load end, a dynamic time-of-use electricity price mechanism is constructed, and the peak and valley periods are dynamically divided. Based on the introduction of substitution flexibility, the optimization goal is to minimize the cumulative difference between the load and the renewable energy power generation, and the users satisfaction on the participation of demand response is also considered, the peak and the valley price of the dynamic time-of-use electricity price is optimized, and the load after demand response is calculated. At the source end, considering the complementary effects between water and wind, an evaluation index of complementary effect is proposed, and on the basis of taking the demand response participation into consideration, a model with optimization of annual operating income and water and wind complementary effect as the objective function is established. The improved particle swarm algorithm combined with the goal programming method is used to solve the problem. The simulation results show that the application of price-based demand response to the optimal configuration of micro-grids can effectively combine the source and load sides, improve the level of renewable energy consumption, and ensure the interests of users. At the same time, the economic value of the micro grid will be improved.

为提高蝗虫优化算法（GOA）求解多目标问题的性能，提出一种基于多策略融合的混合多目标蝗虫优化算法（HMOGOA）。首先，利用Halton序列建立初始种群，保证种群在初始阶段具有均匀分布和较高多样性；然后，通过差分变异算子引导种群变异，促进种群向优势个体移动同时进行更大范围寻优；最后，利用自适应权重因子根据种群优化情况动态调整算法全局搜索和局部寻优能力，提高优化效率及解集质量。选取7个典型函数进行实验测试，并将HMOGOA与多目标蝗虫优化、多目标粒子群（MOPSO）、基于分解的多目标进化（MOEA/D）及非支配排序遗传算法（NSGA Ⅱ）对比分析。实验结果表明，该算法避免了其他四种算法的局部最优问题，明显提高了解集分布均匀性和分布广度，具有更好的收敛精度和稳定性。

蜣螂优化算法（DBO）是近年提出的智能优化算法，与其他优化算法一样，DBO也存在收敛精度低、易陷入局部最优等缺点。针对DBO的这些局限性，提出一种多策略改进的蜣螂优化算法（MIDBO）。首先，改进雏球和偷窃蜣螂对局部最优解和全局最优解的接受程度，使其根据自身搜索能力动态变化，既提升了种群质量又保持了适应度高的个体的良好搜索能力；其次，融合麻雀搜索算法中的追随者位置更新机制对算法进行扰动，并用贪婪策略更新位置，提升了算法的收敛精度；最后，当算法陷入停滞时引入柯西高斯变异策略，提高了算法跳出局部最优解的能力。仿真实验基于20个基准测试函数和CEC2019测试函数，验证了3种改进策略的有效性，将所改进算法和对比算法的优化结果进行收敛性分析和Wilcoxon秩和检验，证明了MIDBO具有良好的寻优性能和鲁棒性。将MIDBO运用在汽车碰撞优化问题的求解上，进一步验证了MIDBO在求解实际工程问题中的有效性和可靠性。