摘要
针对传统算法存在的步长调整及耦合点选择问题,该文提出基于子循环自适应串行交错(SCAS)的时间匹配算法。首先,在步长调整方面,通过加入自适应(ATS)-启发式(HTS)混合变步长算法,解决步长无法实时调整的问题;然后,针对耦合点选择问题,该文在传统常规串行交错(CSS)时间匹配算法及现有改进方案子循环常规串行交错(SCSS)的基础上,提出SCAS时间匹配算法,采用自适应方案确定计算时间窗及预定耦合点,并引入指数平滑法对其精度进行保证;最后,该文基于110 kV油浸式电力变压器建立二维瞬态单分区分匝绕组模型,并将所提算法与其他传统方法进行对比分析。结果表明:在精度方面,与传统CSS算法相比SCAS算法的流场及温度场最大平均相对误差均不超过0.45%,在可接受误差范围内;在效率方面,SCAS算法总体计算效率较CSS算法提高了近20倍。同时,为了说明所提算法的工程价值,该文搭建基于110 kV变压器绕组模型的温升实验平台,将SCAS时间匹配算法应用于该模型中,实验结果表明:在精度方面,SCAS算法与实验达到稳态时最大绝对误差为2.7 K;在计算效率上,本文所提算法的计算效率较传统算法提升约46倍,计算步数约为传统的1/47,减少了计算冗余。
For the time-matching problem of two or more multiple physical fields,the traditional idea generally considers that the physical fields are solved by coupling with fixed equal time steps.This direct coupling algorithm is named conventional serial staggered(CSS),but there are two obvious problems in this time matching method,namely,step size adjustment and coupling point selection.Therefore,this paper proposes a time-matching algorithm based on Sub cyclic adaptive staggered(SCAS).Firstly,in step size adjustment,the Adaptive Time Stepping algorithm(ATS)is added,and the difference of approximate solutions with different accuracy is used as the basis for time step adjustment.At the same time,for nonlinear problems,the Heuristic time stepping method(HTS)is combined to set the number of iterations as the basis for nonlinear discrimination,which solves the problem that the step size cannot be adjusted in real time under strong nonlinear problems.It can obtain a more appropriate step size each time according to the changing trend of the physical field under the allowable error.The introduction of the ATS-HTS hybrid variable step size method reduces the number of computing time steps greatly.Secondly,for the problem of coupling point selection,this paper proposes a SCAS time-matching algorithm based on the CSS time-matching algorithm,which uses an adaptive scheme to determine the calculation time window and predetermined coupling points and introduces an exponential smoothing method to ensure its accuracy.Finally,to verify the effectiveness of the proposed method,a two-dimensional transient single-zone split turn winding model is established.Compared with the traditional CSS algorithm,in terms of accuracy,the average relative error of the flow field of the SCAS algorithm is 0.25%,and the average relative error of the temperature field is 0.45%,which is within the acceptable error range.In terms of efficiency,after combining coupling point selection and step size adaptive selection,the SCAS algorithm has further reduced the number of coupling times and computing time steps,and its overall computing efficiency is nearly 20 times higher than the CSS algorithm.At the same time,to further illustrate the value of the proposed algorithm in engineering practice,this paper has built a temperature rise experimental platform based on the 110 kV transformer winding model and applied the SCAS time matching algorithm to the heat transfer model.The calculation and experimental results show that the maximum absolute error between the SCAS algorithm and the experiment when reaching the steady state occurs at the temperature measuring point No.4 of the 30-wire cakes,which is 2.7 K,because the temperature rise experiment is affected by several uncontrollable factors.Therefore,the error is within the acceptable range.The computational efficiency of the algorithm proposed in this paper is about 46 times higher than that of the traditional algorithm,and the number of computational steps is about 1/47 of the traditional algorithm,which greatly reduces computational redundancy.
作者
刘刚
郝世缘
胡万君
刘云鹏
李琳
Liu Gang;Hao Shiyuan;Hu Wanjun;Liu Yunpeng;Li Lin(Hebei Provincial Key Laboratory of Power Transmission Equipment Security Defense North China Electric Power University,Baoding071003China;State Key Laboratory of Alternate Electrical Power System with Renewable Energy SourcesNorth China Electric Power University,Beijing 102206 China)
出处
《电工技术学报》
EI
CSCD
北大核心
2024年第4期1185-1197,共13页
Transactions of China Electrotechnical Society
基金
国家重点研发计划资助项目(2021YFB2401703)。
关键词
子循环自适应串行交错
混合变步长
二维瞬态流热耦合问题
快速计算方法
温升实验
Sub-cyclic adaptive staggered
mixed variable time step
thermal coupling problem of two dimensional transient flow
fast calculation method
temperature rise experiment
