融合CNN-BiLSTM-Attention的集成学习价格预测

价格预测对于大宗农产品市场的稳定具有重要意义,但是大宗农产品价格与多种因素有着复杂的相关关系.针对当前价格预测中对数据完整性依赖性强与单一模型难以全面利用多种数据特征等问题,提出了一种将基于注意力机制的卷积双向长短期记忆神经网络(CNN-BiLSTM-Attention)、支持向量机回归(SVR)与LightGBM组合的增强式集成学习方法,并分别在包含历史交易、天气、汇率、油价等多种特征数据的数据集上进行了实验.实验以小麦和棉花价格预测为目标任务,使用互信息法进行特征选择,选择误差较低的CNN-BiLSTM-Attention模型作为基模型,与机器学习模型通过线性回归进行增强式集成学习.实验结果表明该集成学习方法在小麦及棉花数据集上预测结果的均方根误差(RMSE)值分别为12.812, 74.365,较之3个基模型分别降低11.00%, 0.94%、4.44%,1.99%与13.03%, 4.39%,能够有效降低价格预测的误差.     

基于XGBoost和TCN-Attention的棉花价格多影响因素选择及预测

棉花价格受多种因素影响而复杂多变, 通过选择合适的数据特征和预测模型可提高棉花价格预测精度. 本文以棉花日现货价格数据为研究目标, 采集了供需关系、国际市场、宏观经济、产业链这4个方面的9项影响因素作为特征, 使用极限梯度提升(XGBoost)算法对棉花价格影响因素进行特征评估筛选, 选取其中5项特征后, 采用引入注意力机制(Attention)的时间卷积网络(TCN) TCN-Attention、TCN、LSTM、GRU等模型对棉花价格进行预测. 通过消融实验和对比实验, 结果表明: (1)经过XGBoost特征筛选后, TCN-Attention价格预测的平均绝对误差(MAE)和均方根误差(RMSE)为41.47和58.76, 与未筛选相比分别降低了77.57%和76.49%. (2)与TCN、LSTM、GRU相比, 本文提出的TCN-Attention模型预测结果更准确, MAE和RMSE均降低50%以上, 运行时间较LSTM、GRU缩短60%.     

电网能量流-经济流耦合分析

依托于同一电力网络的电气源流间的能流信息和市场主体间的交易信息,存在某种天然耦合关系。若不能正确处理,则易导致相关经济流分布信息掌握失准,从而影响系统运行的效能和市场运作的公正性。为此,基于电力网络电能追踪方法正确提供的实时能流信息,考虑上述耦合特性,分析了沿网络源-流对电能配送路径链分布的经济流信息,建立了核算各项费用的实时电价关系式;针对链首和链末2种能量计费方式,提出了具有广泛指导性的费用核算一致性原理,并指出：若不考虑电网实时运行状态影响而直接以上网价、过网价和政府性基金价之和作为下网电价核算相关费用,则会出现不完全满足一致性原理的现象。最后,针对典型交易模式及与其关联的多个网络源-流对,给出了量化其经济流分布的关系式。算例表明,基于考虑耦合特性的实时电价信号分析的经济流,符合核算一致性原理,可为进一步揭示交叉补贴内因提供理论依据。     

独立储能电站参与电力现货市场机制及试运行分析

随着风光新能源装机容量迅速增长,加快健全储能设施参与电力现货市场交易机制有利于促进新能源消纳。为提高系统调节能力,以山东省独立电储能参与现货市场交易机制为研究对象,提出了独立储能以自调度模式参与现货市场的运行机制。仿真运行数据表明,储能装置参与现货市场对促进新能源消纳和提升电力供应保障能力具有重要作用。     

计及分时电价的5 G基站光储系统容量优化配置方法

随着第5代移动通信技术（5G通信）的迅速发展,5G基站数量不断增加,5G通信耗电量大、用电成本高的问题日益突出。为此,提出了考虑光伏和储能接入的5G基站光储系统优化配置方法,以提高5G基站运行的经济性。首先,考虑5G基站负荷情况和配电网分时电价,建立了5G基站光储系统的经济调度模型;然后,通过量子粒子群优化算法计算典型日内5G基站光储系统的最小综合成本,以此确定光伏和储能的最优接入容量;最后,通过算例证明合理配置光伏和储能容量可以提高5G基站系统的经济性。光伏和储能容量的配置受储能成本和峰谷电价差的影响较大,且光储系统所能带来的经济效益随峰谷电价差的增大和储能成本的降低而增大。     

基于强化学习的电动汽车换电站实时调度策略优化

随着电动汽车的应用推广,换电站的调度优化逐渐成为研究热点。传统的基于换电需求预测值的调度策略在实际应用中面临着难以适应动态干扰因素、预测误差累积等问题。为了解决这些问题,提出了一种基于带基线的蒙特卡罗策略梯度法的换电站实时调度策略,用于优化换电站的充放电策略以及响应电池数量。提出了带基线的蒙特卡罗策略梯度强化学习,并为换电站实时调度问题选取合适的状态空间和动作空间;设计了奖励函数对智能体进行离线训练,从电池状态数据、分时电价和排队电动汽车数量中学习得到最优策略网络;在离线训练好的模型基础上进行实时调度策略测试。基于换电站的服务可用率和经济效益验证了所提调度策略的有效性和经济性,算例结果表明所提策略能对电网负荷起到一定的削峰填谷作用。     

基于iDLMP方法的用户侧灵活性资源电能-备用优化运行

充分利用用户侧分布式能源(DER)的灵活性是改善以新能源为主体的新型电力系统灵活性不足的有效途径之一。搭建含有高比例DER的电力系统市场化运行框架;基于通用虚拟电池模型,建立含有分布式储能系统、电动汽车与光伏的聚合商优化运行模型,进一步考虑配电网安全运行约束,建立配电系统运营商优化运行模型,基于此,提出协调优化聚合商能量计划与备用容量计划的迭代型配电网节点边际价格方法。算例结果表明,所提模型和方法可以在考虑用户设备资源信息安全的前提下有效解决DER因电力市场能量价格与备用容量价格过高或过低引起的线路双向过载问题。     

面向精准价格牌识别的多任务循环神经网络

为了促进智能新零售在线下业务场景的发展, 提高作为销售关键信息价格牌的识别精度. 本文对价格牌识别问题进行研究, 有效地提高了价格牌的识别精度, 并解决小数点定位不准确的难题. 通过深度卷积神经网络提取价格牌的深度语义表达特征, 将提取到的特征图送入多任务循环网络层进行编码, 然后根据解码网络设计的注意力机制解码出价格数字, 最后将多个分支的结果整合并输出完整价格. 本文所提出的方法能够非常有效地提高线下零售场景价格牌的识别精度, 并解决了一些领域难题如小数点的定位问题, 此外, 为了验证本文方法的普适性, 在其他场景数据集上进行了对比实验, 相关结果也验证了本文方法的有效性.     

The price of anarchy in closed-loop supply chains

This paper measures the worst-case efficiency of price-only contracts in closed-loop supply chains (CLSCs) with the price of anarchy (PoA). We model a single-period Stackelberg game in which a manufacturer sells new products to a retailer and collects used products with exogenous retail price and collection price via three alternative reverse channels: (a) the manufacturer collects directly from customers, (b) the retailer collects for the manufacturer, and (c) a third party is awarded a collection subcontract from the manufacturer. We carry out a comprehensive investigation under push–pull configurations to observe how reverse channel structures with different gaming sequences of CLSC members influence the worst-case performance when the demand distribution is over the set of increasing generalized failure rate distributions. From our PoA analysis, we find that the pull system does not always outperform the push system, especially when the retailer is the leader, in contrast to the results for forward supply chains. While the PoA of the push system is dramatically sensitive to the quality condition of used products, the pull system has a constant efficiency loss that is independent of the quality condition. Instead, the PoA of the pull system solely changes with the gaming sequence of the manufacturer. We also find that manufacturer's direct collection is a better reverse channel choice compared to retailer's collection. Additional managerial insights are summarized for discussion.     

HSWOA: An ensemble of hunger games search and whale optimization algorithm for global optimization

The search for food stimulated by hunger is a common phenomenon in the animal world. Mimicking the concept, recently, an optimization algorithm Hunger Games Search (HGS) has been proposed for global optimization. On the other side, the Whale Optimization Algorithm (WOA) is a commonly utilized nature-inspired algorithm portrayed by a straightforward construction with easy parameters imitating the hunting behavior of humpback whales. However, due to minimum exploration of the search space, WOA has a high chance of trapping into local solutions, and more exploitation leads it towards premature convergence. The concept of hunger from HGS is merged with the food searching techniques of the whale to lessen the inherent drawbacks of WOA. Two weights of HGS are adaptively designed for every whale using the respective hunger level for balancing search strategies. Performance verification of the proposed hunger search-based whale optimization algorithm (HSWOA) is done by comparing it with 10 state-of-the-art algorithms, including three very recently developed algorithms on 30 classical benchmark functions. Comparison with some basic algorithms, recently modified algorithms, and WOA variants is performed using IEEE CEC 2019 function set. Statistical performance of the proposed algorithm is verified with Friedman's test, boxplot analysis, and Nemenyi multiple comparison test. The operating speed of the algorithm is determined and tested with complexity analysis and convergence analysis. Finally, seven real-world engineering problems are solved and compared with a list of metaheuristic algorithms. Numerical and statistical performance comparison with state-of-the-art algorithms confirms the efficacy of the newly designed algorithm.