Wenhang Jiang, Jiu-fu Liu, Anbang Peng, Guodong Liu, Rong Zhang
{"title":"利用方差分析量化水库发电人工神经网络模型不确定性来源的贡献","authors":"Wenhang Jiang, Jiu-fu Liu, Anbang Peng, Guodong Liu, Rong Zhang","doi":"10.2166/nh.2022.052","DOIUrl":null,"url":null,"abstract":"\n There are many sources of uncertainty in reservoir operation. The presence of these uncertainties might lead to operation risks, which directly affect the comprehensive benefit of reservoirs. This study developed a simple framework to quantify the uncertainty contribution arising from the inputs, model structures, model parameters, and their interaction in the reservoirs. We established a deterministic reservoir operations model with the intention of maximizing power generation, and the scheduling results with the inputs and optimal output datasets were used for data-driven models – artificial neural networks (ANNs). The time period, inflow, storage, and inflow in the last period were chosen as input, integrating with ANN models of different structures and parameters, to produce an ensemble of 10-day forecasts of power generation. The analysis of variance (ANOVA) method was applied to quantify the contribution of the uncertainty sources. The results demonstrated that the inputs were the predominating source of uncertainty in the reservoir operation, especially from May to October. In addition, the uncertainty caused by the interactions between the three sources of uncertainty was more considerable than that of the model structure or parameter in November–April, and the uncertainty contributions of the model structure or parameter were relatively marginal.","PeriodicalId":55040,"journal":{"name":"Hydrology Research","volume":" ","pages":""},"PeriodicalIF":2.7000,"publicationDate":"2022-12-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"1","resultStr":"{\"title\":\"Quantifying the contribution of uncertainty sources of artificial neural network models using ANOVA for reservoir power generation\",\"authors\":\"Wenhang Jiang, Jiu-fu Liu, Anbang Peng, Guodong Liu, Rong Zhang\",\"doi\":\"10.2166/nh.2022.052\",\"DOIUrl\":null,\"url\":null,\"abstract\":\"\\n There are many sources of uncertainty in reservoir operation. The presence of these uncertainties might lead to operation risks, which directly affect the comprehensive benefit of reservoirs. This study developed a simple framework to quantify the uncertainty contribution arising from the inputs, model structures, model parameters, and their interaction in the reservoirs. We established a deterministic reservoir operations model with the intention of maximizing power generation, and the scheduling results with the inputs and optimal output datasets were used for data-driven models – artificial neural networks (ANNs). The time period, inflow, storage, and inflow in the last period were chosen as input, integrating with ANN models of different structures and parameters, to produce an ensemble of 10-day forecasts of power generation. The analysis of variance (ANOVA) method was applied to quantify the contribution of the uncertainty sources. The results demonstrated that the inputs were the predominating source of uncertainty in the reservoir operation, especially from May to October. In addition, the uncertainty caused by the interactions between the three sources of uncertainty was more considerable than that of the model structure or parameter in November–April, and the uncertainty contributions of the model structure or parameter were relatively marginal.\",\"PeriodicalId\":55040,\"journal\":{\"name\":\"Hydrology Research\",\"volume\":\" \",\"pages\":\"\"},\"PeriodicalIF\":2.7000,\"publicationDate\":\"2022-12-10\",\"publicationTypes\":\"Journal Article\",\"fieldsOfStudy\":null,\"isOpenAccess\":false,\"openAccessPdf\":\"\",\"citationCount\":\"1\",\"resultStr\":null,\"platform\":\"Semanticscholar\",\"paperid\":null,\"PeriodicalName\":\"Hydrology Research\",\"FirstCategoryId\":\"93\",\"ListUrlMain\":\"https://doi.org/10.2166/nh.2022.052\",\"RegionNum\":4,\"RegionCategory\":\"环境科学与生态学\",\"ArticlePicture\":[],\"TitleCN\":null,\"AbstractTextCN\":null,\"PMCID\":null,\"EPubDate\":\"\",\"PubModel\":\"\",\"JCR\":\"Q2\",\"JCRName\":\"Environmental Science\",\"Score\":null,\"Total\":0}","platform":"Semanticscholar","paperid":null,"PeriodicalName":"Hydrology Research","FirstCategoryId":"93","ListUrlMain":"https://doi.org/10.2166/nh.2022.052","RegionNum":4,"RegionCategory":"环境科学与生态学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q2","JCRName":"Environmental Science","Score":null,"Total":0}
Quantifying the contribution of uncertainty sources of artificial neural network models using ANOVA for reservoir power generation
There are many sources of uncertainty in reservoir operation. The presence of these uncertainties might lead to operation risks, which directly affect the comprehensive benefit of reservoirs. This study developed a simple framework to quantify the uncertainty contribution arising from the inputs, model structures, model parameters, and their interaction in the reservoirs. We established a deterministic reservoir operations model with the intention of maximizing power generation, and the scheduling results with the inputs and optimal output datasets were used for data-driven models – artificial neural networks (ANNs). The time period, inflow, storage, and inflow in the last period were chosen as input, integrating with ANN models of different structures and parameters, to produce an ensemble of 10-day forecasts of power generation. The analysis of variance (ANOVA) method was applied to quantify the contribution of the uncertainty sources. The results demonstrated that the inputs were the predominating source of uncertainty in the reservoir operation, especially from May to October. In addition, the uncertainty caused by the interactions between the three sources of uncertainty was more considerable than that of the model structure or parameter in November–April, and the uncertainty contributions of the model structure or parameter were relatively marginal.
期刊介绍:
Hydrology Research provides international coverage on all aspects of hydrology in its widest sense, and welcomes the submission of papers from across the subject. While emphasis is placed on studies of the hydrological cycle, the Journal also covers the physics and chemistry of water. Hydrology Research is intended to be a link between basic hydrological research and the practical application of scientific results within the broad field of water management.