Due to the widespread use of energy storage systems in new energy grid-connected,lithium-ion batteries are usually used in parallel and series connections to meet the power and energy requirements of system.However,the series-connected single battery are inconsistent,so capacity and state of charge(SOC)inconsistency is an inevitable problem,which can decrease the available capacity and result in accelerated aging and safety issues.In this paper,a consistency diagnosis method based on charging curve transformation was proposed,and capacity and SOC differences within the battery pack was diagnosed.Focusing on the problem that traditional curve transformation method had a high data storage and computational requirement to battery management system,a QV curve transformation model based on curve transformation method was proposed.The consistency diagnosis for the series-connected battery pack is realized in real-time by parameter identification.Finally,through the analysis of a battery pack with six cells connected in series,the accuracy of the proposed method is verified,and the experimental results indicate that the proposed method can realize both capacity inconsistency and SOC inconsistency diagnosis.
Regional integrated energy system (RIES) widely involves many types of energy coupling, such as electric energy, thermal energy, cold energy and so on, and faces many challenges in planning. Realizing the coor-dination and complementarity of supply and demand among various energy sources and taking into account the flexibility of demand is a problem that must be considered in the process of integrated energy planning. A regional integrated energy planning method considering demand response is proposed in this paper. First of all, the energy configuration and load demand of four typical application scenarios of integrated energy in in-dustrial park, school, hospital and port area are analyzed, and the equipment configuration scheme of typical scenario is given; secondly, the demand response models of electricity, heating and cooling loads guided by price are established respectively. Finally, based on the energy hub (EH) model, a integrated energy planning model aiming at minimizing the sum of equipment investment, operation cost and demand response com-pensation cost is established, and the linearization method is used to solve the model. Relevant examples show that the proposed method can effectively carry out the integrated energy system planning in typical scenarios and reduce the planning cost.
针对电负荷和热负荷的不确定性,在考虑网络拓扑约束和多能耦合建模的基础上,推导出基于可信性理论的多重不确定性风险评估方法.在此基础上,采用清晰等价类将表征不确定性风险的模糊机会约束转化为确定性约束,提出一种多重不确定性下电气热耦合系统风险规避调度方法.通过IEEE 33节点配电网络和32节点供暖网络仿真测试,结果验证所提模型在对冲综合能源系统(integrated energy system,IES)多重不确定性风险的有效性,为能源交易决策和风险评估提供了一种新思路.
热电联产经济调度(combined heat and power economic dispatch,CHPED)是一类复杂的非线性约束最优化问题,智能优化算法则是近年来求解CHPED的主流方法.针对目前智能优化算法求解CHPED存在易于陷入局部最优、精度不佳等问题,提出了一种改进的差分进化算法,即高斯柯西差分进化(Gaussian-Cauchy differential evolution,GCDE)算法.GCDE算法对标准差分进化算法进行了两方面的改进:一方面采用自适应参数策略来调节算法的控制参数,提升算法的求解稳定性;另一方面采用高斯柯西变异策略来进行局部搜索,增强算法的搜索精度.将GCDE算法应用于求解7台机组和24台机组的CHPED模型,并与现有差分进化算法结果进行了对比,结果表明GCDE算法在求解CHPED问题时具有较好的求解精度和稳定性.
超声法是局部放电测量的重要方法.为更有效提取局部放电超声信号中的关键信息,提出一种基于广义S变换和二维卷积神经网络的油纸绝缘局部放电超声信号识别方法.首先,在实验室采集3种典型缺陷的局部放电超声信号,对预处理后的数据进行广义S变换,获得不同放电样本的时频分布图像;然后,构建二维卷积神经网络,将时频图像作为输入,提取超声信号的时频特征;最后,输出对局部放电类型的识别结果.结果表明:该方法对不同局部放电类型的超声信号进行识别,准确率可以达到97.8%,能够更有效地提取超声信号的内在信息并进行局部放电识别.
