考虑阀点效应的基于神经网络算法的负荷经济调度  

Economic Load Dispatching Based on Neural Network Algorithm Considering Valve Point Effect

作　　者：徐艺玮 李光熹 刘晗 XU Yiwei;LI Guangxi;LIU Han(School of Electrical and Automation Engineering,Nanjing Normal University,Nanjing 210000,China;Lianyungang Power Supply Company,State Grid Jiangsu Electric Power Company,Lianyungang Jiangsu 222100,China)

机构地区：[1]南京师范大学电气与自动化工程学院,南京210000 [2]国网江苏省电力有限公司连云港供电分公司,江苏连云港222100

出　　处：《自动化与仪器仪表》2024年第6期10-14,共5页Automation & Instrumentation

摘　　要：经济负荷调度(Economic load dispatching, ELD)问题具有非凸、非光滑特性的特点,在电力系统运行及元启发式算法研究中,均是重要课题。提出了一种基于神经网络算法(Neural network algorithm, NNA)的ELD求解模型,考虑发电机组阀点效应因素,将阀点效应所引起的煤耗成本计入目标函数,建立了考虑阀点效应约束的经济调度模型。以调度总成本为目标函数,在满足负荷需求及系统运行约束的条件下,最大程度地降低运行总成本。通过对IEEE-30、IEEE-39及IEEE-118标准测试系统的仿真,分析了算法的搜索能力及迭代收敛情况,并同其他方法的结果进行对比,验证了所提方法的有效性和适用性。The economic load dispatching(ELD)problem has non-convex and non-smooth characteristics.It is an important topic in power system operation and meta-heuristic algorithm research.The paper proposes an ELD solving model based on neural network algorithm(NNA),which takes into account the valve point effect of generator units,and includes the coal consumption cost caused by the valve point effect.An economic dispatch model considering the valve point effect constraint is established.Taking the total scheduling cost as the objective function,minimizing the total operating cost while meeting the load demand and operation constraints.By simulating IEEE-30,IEEE-39,and IEEE-118 systems,the search ability and iterative convergence were analyzed.The results are compared with those of other methods to verify the effectiveness and applicability of the proposed method.

关 键 词：经济负荷调度 神经网络 阀点效应 运行约束 

分 类 号：TM732[电气工程—电力系统及自动化] TP29[自动化与计算机技术—检测技术与自动化装置]