The large blast furnace is essential equipment in the process of iron and steel manufacturing. Due to the complex operation process and frequent fluctuations of variables, conventional monitoring methods often bring f...The large blast furnace is essential equipment in the process of iron and steel manufacturing. Due to the complex operation process and frequent fluctuations of variables, conventional monitoring methods often bring false alarms. To address the above problem, an ensemble of greedy dynamic principal component analysis-Gaussian mixture model(EGDPCA-GMM) is proposed in this paper. First, PCA-GMM is introduced to deal with the collinearity and the non-Gaussian distribution of blast furnace data.Second, in order to explain the dynamics of data, the greedy algorithm is used to determine the extended variables and their corresponding time lags, so as to avoid introducing unnecessary noise. Then the bagging ensemble is adopted to cooperate with greedy extension to eliminate the randomness brought by the greedy algorithm and further reduce the false alarm rate(FAR) of monitoring results. Finally, the algorithm is applied to the blast furnace of a large iron and steel group in South China to verify performance.Compared with the basic algorithms, the proposed method achieves lowest FAR, while keeping missed alarm rate(MAR) remain stable.展开更多
Due to the problems of few fault samples and large data fluctuations in the blast furnace(BF)ironmaking process,some transfer learning-based fault diagnosis methods are proposed.The vast majority of such methods perfo...Due to the problems of few fault samples and large data fluctuations in the blast furnace(BF)ironmaking process,some transfer learning-based fault diagnosis methods are proposed.The vast majority of such methods perform distribution adaptation by reducing the distance between data distributions and applying a classifier to generate pseudo-labels for self-training.However,since the training data is dominated by labeled source domain data,such classifiers tend to be weak classifiers in the target domain.In addition,the features generated after domain adaptation are likely to be at the decision boundary,resulting in a loss of classification performance.Hence,we propose a novel method called minimax entropy-based co-training(MMEC)that adversarially optimizes a transferable fault diagnosis model for the BF.The structure of MMEC includes a dual-view feature extractor,followed by two classifiers that compute the feature's cosine similarity to representative vector of each class.Knowledge transfer is achieved by alternately increasing and decreasing the entropy of unlabeled target samples with the classifier and the feature extractor,respectively.Transfer BF fault diagnosis experiments show that our method improves accuracy by about 5%over state-of-the-art methods.展开更多
Background:In the face of continued degradation and loss of wetlands in the Yangtze River floodplain(YRF),there is an urgent need to monitor the abundance and distribution of wintering waterbirds.To understand fully o...Background:In the face of continued degradation and loss of wetlands in the Yangtze River floodplain(YRF),there is an urgent need to monitor the abundance and distribution of wintering waterbirds.To understand fully observed annual changes,we need to monitor demographic rates to understand factors affecting global population size.Annual reproduction success contributes to dynamic changes in population size and age structure,so an assessment of the juvenile ratio(i.e.first winter birds as a proportion of total number aged)of overwintering waterbirds can be an important indicator of the reproductive success in the preceding breeding season.Methods:During 2016-2019,we sampled juvenile ratios among 10 key waterbird species from the wetlands in the YRF.Based on these data,we here attempt to establish a simple,efficient,focused and reliable juvenile ratio monitoring scheme,to assess consistently and accurately relative annual breeding success and its contribution to the age structure among these waterbird species.Results:We compared juvenile ratio data collected throughout the winter and found that the optimal time for undertaking these samples was in the early stages of arrival for migratory waterbirds reaching their wintering area(early to mid-December).We recommend counting consistently at key points(i.e.those where>1%biogeographical flyway population were counted)at sites of major flyway importance(Poyang Lake,East Dongting Lake,Shengjin Lake,Caizi Lake,Longgan Lake and Chen Lake).Based on this,the error rate of the programme(155 planned points,the count of 10 waterbird species is 826-8955)is less than 5%.Conclusions:We established a juvenile ratio monitoring programme for 10 key waterbird species in the wetlands of the YRF,and discuss the feasibility and necessity of implementing such a future programme,and how to use these data in our monitoring and understanding of the population dynamics of these waterbird populations.展开更多
基金supported by the National Natural Science Foundation of China (61903326, 61933015)。
文摘The large blast furnace is essential equipment in the process of iron and steel manufacturing. Due to the complex operation process and frequent fluctuations of variables, conventional monitoring methods often bring false alarms. To address the above problem, an ensemble of greedy dynamic principal component analysis-Gaussian mixture model(EGDPCA-GMM) is proposed in this paper. First, PCA-GMM is introduced to deal with the collinearity and the non-Gaussian distribution of blast furnace data.Second, in order to explain the dynamics of data, the greedy algorithm is used to determine the extended variables and their corresponding time lags, so as to avoid introducing unnecessary noise. Then the bagging ensemble is adopted to cooperate with greedy extension to eliminate the randomness brought by the greedy algorithm and further reduce the false alarm rate(FAR) of monitoring results. Finally, the algorithm is applied to the blast furnace of a large iron and steel group in South China to verify performance.Compared with the basic algorithms, the proposed method achieves lowest FAR, while keeping missed alarm rate(MAR) remain stable.
基金supported in part by the National Natural Science Foundation of China(61933015)in part by the Central University Basic Research Fund of China under Grant K20200002(for NGICS Platform,Zhejiang University)。
文摘Due to the problems of few fault samples and large data fluctuations in the blast furnace(BF)ironmaking process,some transfer learning-based fault diagnosis methods are proposed.The vast majority of such methods perform distribution adaptation by reducing the distance between data distributions and applying a classifier to generate pseudo-labels for self-training.However,since the training data is dominated by labeled source domain data,such classifiers tend to be weak classifiers in the target domain.In addition,the features generated after domain adaptation are likely to be at the decision boundary,resulting in a loss of classification performance.Hence,we propose a novel method called minimax entropy-based co-training(MMEC)that adversarially optimizes a transferable fault diagnosis model for the BF.The structure of MMEC includes a dual-view feature extractor,followed by two classifiers that compute the feature's cosine similarity to representative vector of each class.Knowledge transfer is achieved by alternately increasing and decreasing the entropy of unlabeled target samples with the classifier and the feature extractor,respectively.Transfer BF fault diagnosis experiments show that our method improves accuracy by about 5%over state-of-the-art methods.
基金supported by the National Natural Science Foundation of China(Grant Nos.31870369,31970433)China Biodiversity Observation Networks(Sino BON)+1 种基金Innovative Research Group Project of the National Natural Science Foundation of China(CN)No.31670424。
文摘Background:In the face of continued degradation and loss of wetlands in the Yangtze River floodplain(YRF),there is an urgent need to monitor the abundance and distribution of wintering waterbirds.To understand fully observed annual changes,we need to monitor demographic rates to understand factors affecting global population size.Annual reproduction success contributes to dynamic changes in population size and age structure,so an assessment of the juvenile ratio(i.e.first winter birds as a proportion of total number aged)of overwintering waterbirds can be an important indicator of the reproductive success in the preceding breeding season.Methods:During 2016-2019,we sampled juvenile ratios among 10 key waterbird species from the wetlands in the YRF.Based on these data,we here attempt to establish a simple,efficient,focused and reliable juvenile ratio monitoring scheme,to assess consistently and accurately relative annual breeding success and its contribution to the age structure among these waterbird species.Results:We compared juvenile ratio data collected throughout the winter and found that the optimal time for undertaking these samples was in the early stages of arrival for migratory waterbirds reaching their wintering area(early to mid-December).We recommend counting consistently at key points(i.e.those where>1%biogeographical flyway population were counted)at sites of major flyway importance(Poyang Lake,East Dongting Lake,Shengjin Lake,Caizi Lake,Longgan Lake and Chen Lake).Based on this,the error rate of the programme(155 planned points,the count of 10 waterbird species is 826-8955)is less than 5%.Conclusions:We established a juvenile ratio monitoring programme for 10 key waterbird species in the wetlands of the YRF,and discuss the feasibility and necessity of implementing such a future programme,and how to use these data in our monitoring and understanding of the population dynamics of these waterbird populations.
