Utilizing machine learning techniques for data-driven diagnosis of high temperature PEM fuel cells is beneficial and meaningful to the system durability. Nevertheless, ensuring the robustness of diagnosis remains a cr...Utilizing machine learning techniques for data-driven diagnosis of high temperature PEM fuel cells is beneficial and meaningful to the system durability. Nevertheless, ensuring the robustness of diagnosis remains a critical and challenging task in real application. To enhance the robustness of diagnosis and achieve a more thorough evaluation of diagnostic performance, a robust diagnostic procedure based on electrochemical impedance spectroscopy (EIS) and a new method for evaluation of the diagnosis robustness was proposed and investigated in this work. To improve the diagnosis robustness: (1) the degradation mechanism of different faults in the high temperature PEM fuel cell was first analyzed via the distribution of relaxation time of EIS to determine the equivalent circuit model (ECM) with better interpretability, simplicity and accuracy;(2) the feature extraction was implemented on the identified parameters of the ECM and extra attention was paid to distinguishing between the long-term normal degradation and other faults;(3) a Siamese Network was adopted to get features with higher robustness in a new embedding. The diagnosis was conducted using 6 classic classification algorithms—support vector machine (SVM), K-nearest neighbor (KNN), logistic regression (LR), decision tree (DT), random forest (RF), and Naive Bayes employing a dataset comprising a total of 1935 collected EIS. To evaluate the robustness of trained models: (1) different levels of errors were added to the features for performance evaluation;(2) a robustness coefficient (Roubust_C) was defined for a quantified and explicit evaluation of the diagnosis robustness. The diagnostic models employing the proposed feature extraction method can not only achieve the higher performance of around 100% but also higher robustness for diagnosis models. Despite the initial performance being similar, the KNN demonstrated a superior robustness after feature selection and re-embedding by triplet-loss method, which suggests the necessity of robustness evaluation for the machine learning models and the effectiveness of the defined robustness coefficient. This work hopes to give new insights to the robust diagnosis of high temperature PEM fuel cells and more comprehensive performance evaluation of the data-driven method for diagnostic application.展开更多
Increasing the production and utilization of shale gas is of great significance for building a clean and low-carbon energy system.Sharp decline of gas production has been widely observed in shale gas reservoirs.How to...Increasing the production and utilization of shale gas is of great significance for building a clean and low-carbon energy system.Sharp decline of gas production has been widely observed in shale gas reservoirs.How to forecast shale gas production is still challenging due to complex fracture networks,dynamic fracture properties,frac hits,complicated multiphase flow,and multi-scale flow as well as data quality and uncertainty.This work develops an integrated framework for evaluating shale gas well production based on data-driven models.Firstly,a comprehensive dominated-factor system has been established,including geological,drilling,fracturing,and production factors.Data processing and visualization are required to ensure data quality and determine final data set.A shale gas production evaluation model is developed to evaluate shale gas production levels.Finally,the random forest algorithm is used to forecast shale gas production.The prediction accuracy of shale gas production level is higher than 95%based on the shale gas reservoirs in China.Forty-one wells are randomly selected to predict cumulative gas production using the optimal regression model.The proposed shale gas production evaluation frame-work overcomes too many assumptions of analytical or semi-analytical models and avoids huge computation cost and poor generalization for numerical modelling.展开更多
The wind–thermal bundled power system achieves energy complementarity and optimized scheduling, which is an important way to build a new type of energy system. For the safe and stable operation of the wind–thermal b...The wind–thermal bundled power system achieves energy complementarity and optimized scheduling, which is an important way to build a new type of energy system. For the safe and stable operation of the wind–thermal bundled power system, accurate data-driven analysis is necessary to maintain real-time balance between electricity supply and demand. By summarizing the development and characteristics of wind–thermal bundled power system in China and different countries, current research in this field can be clearly defined in two aspects: short-term wind power prediction for wind farms and performance evaluation of automatic generation control (AGC) for thermal power generation units. For short-term wind power prediction, it is recommended to focus on historical data preprocessing and artificial intelligence methods. The technical characteristics of different data-driven wind power prediction methods have been compared in detail. For performance evaluation of AGC units, a comprehensive analysis was conducted on current evaluation methods, including the “permitted-band” and “regulation mileage” methods, as well as the issue of evaluation failure in traditional evaluation methods in practical engineering. Finally, the relative optimal dynamic performance of AGC units was discussed and the future trend of data-driven research in wind–thermal bundled power system was summarized.展开更多
Complex engineered systems are often difficult to analyze and design due to the tangled interdependencies among their subsystems and components. Conventional design methods often need exact modeling or accurate struct...Complex engineered systems are often difficult to analyze and design due to the tangled interdependencies among their subsystems and components. Conventional design methods often need exact modeling or accurate structure decomposition, which limits their practical application. The rapid expansion of data makes utilizing data to guide and improve system design indispensable in practical engineering. In this paper, a data driven uncertainty evaluation approach is proposed to support the design of complex engineered systems. The core of the approach is a data-mining based uncertainty evaluation method that predicts the uncertainty level of a specific system design by means of analyzing association relations along different system attributes and synthesizing the information entropy of the covered attribute areas, and a quantitative measure of system uncertainty can be obtained accordingly. Monte Carlo simulation is introduced to get the uncertainty extrema, and the possible data distributions under different situations is discussed in detail The uncertainty values can be normalized using the simulation results and the values can be used to evaluate different system designs. A prototype system is established, and two case studies have been carded out. The case of an inverted pendulum system validates the effectiveness of the proposed method, and the case of an oil sump design shows the practicability when two or more design plans need to be compared. This research can be used to evaluate the uncertainty of complex engineered systems completely relying on data, and is ideally suited for plan selection and performance analysis in system design.展开更多
Based on the traditional numerical simulation and optimization algorithms,in combination with the layered injection and production"hard data"monitored at real time by automatic control technology,a systemati...Based on the traditional numerical simulation and optimization algorithms,in combination with the layered injection and production"hard data"monitored at real time by automatic control technology,a systematic approach for detailed water injection design using data-driven algorithms is proposed.First the data assimilation technology is used to match geological model parameters under the constraint of observed well dynamics;the flow relationships between injectors and producers in the block are calculated based on automatic identification method for layered injection-production flow relationship;multi-layer and multi-direction production splitting technique is used to calculate the liquid and oil production of producers in different layers and directions and obtain quantified indexes of water injection effect.Then,machine learning algorithms are applied to evaluate the effectiveness of water injection in different layers of wells and to perform the water injection direction adjustment.Finally,the particle swarm algorithm is used to optimize the detailed water injection plan and to make production predictions.This method and procedure make full use of the automation and intelligence of data-driven and machine learning algorithms.This method was used to match the data of a complex faulted reservoir in eastern China,achieving a fitting level of 85%.The cumulative oil production in the example block for 12 months after optimization is 8.2%higher than before.This method can help design detailed water injection program for mature oilfields.展开更多
With more commercialized automated vehicles(AVs)shortly entering the market,evaluating their fuel economy has become an important topic.The traditional fixed-profile test methods only indicate the effect of powertrain...With more commercialized automated vehicles(AVs)shortly entering the market,evaluating their fuel economy has become an important topic.The traditional fixed-profile test methods only indicate the effect of powertrain efficiency on fuel economy and cannot reflect the influences of control algorithms or driving behaviors on fuel consumption under real-world traffic conditions.Therefore,a data-driven simulation method for evaluating the real driving fuel economy of automated vehicles is developed.It utilizes naturalistic driving data to reconstruct test cycles.This method can inspect the performance of automated vehicle control algorithms under realistic traffic conditions in terms of fuel economy.The naturalistic driving data collected on urban expressways and freeways were used to model the longitudinal driving scenarios.Then,the fuel consumption of automated and human-driven vehicle was evaluated under the simulated scenarios via the Monte Carlo integration approach.It was found that the fuel consumption rate of the automated vehicle was 11.944 L/100 km under the car-following scenarios.The human-driven vehicle had a fuel consumption rate of 10.124 L/100 km under the same traffic conditions.The tested automated vehicle control algorithm is tuned to achieve better performance in terms of safety and travel efficiency and hence it tends to maintain a relatively steady time headway to the leading vehicle.It applies more frequent accelerating/braking cycles and higher average speed compared to typical human drivers.These characteristics lead to a higher fuel consumption rate of the automated vehicle.The presented method provides an accurate and efficient way to analyze the fuel economy performance of automated vehicles under practical conditions.This method can easily be scaled for large-scale traffic flow analyses.It can also be used to study the effects of human driving styles on fuel economy.展开更多
基金supported by the Chinese Scholarship Council(Nos.202208320055 and 202108320111)the support from the energy department of Aalborg University was acknowledged.
文摘Utilizing machine learning techniques for data-driven diagnosis of high temperature PEM fuel cells is beneficial and meaningful to the system durability. Nevertheless, ensuring the robustness of diagnosis remains a critical and challenging task in real application. To enhance the robustness of diagnosis and achieve a more thorough evaluation of diagnostic performance, a robust diagnostic procedure based on electrochemical impedance spectroscopy (EIS) and a new method for evaluation of the diagnosis robustness was proposed and investigated in this work. To improve the diagnosis robustness: (1) the degradation mechanism of different faults in the high temperature PEM fuel cell was first analyzed via the distribution of relaxation time of EIS to determine the equivalent circuit model (ECM) with better interpretability, simplicity and accuracy;(2) the feature extraction was implemented on the identified parameters of the ECM and extra attention was paid to distinguishing between the long-term normal degradation and other faults;(3) a Siamese Network was adopted to get features with higher robustness in a new embedding. The diagnosis was conducted using 6 classic classification algorithms—support vector machine (SVM), K-nearest neighbor (KNN), logistic regression (LR), decision tree (DT), random forest (RF), and Naive Bayes employing a dataset comprising a total of 1935 collected EIS. To evaluate the robustness of trained models: (1) different levels of errors were added to the features for performance evaluation;(2) a robustness coefficient (Roubust_C) was defined for a quantified and explicit evaluation of the diagnosis robustness. The diagnostic models employing the proposed feature extraction method can not only achieve the higher performance of around 100% but also higher robustness for diagnosis models. Despite the initial performance being similar, the KNN demonstrated a superior robustness after feature selection and re-embedding by triplet-loss method, which suggests the necessity of robustness evaluation for the machine learning models and the effectiveness of the defined robustness coefficient. This work hopes to give new insights to the robust diagnosis of high temperature PEM fuel cells and more comprehensive performance evaluation of the data-driven method for diagnostic application.
基金funded by National Natural Science Foundation of China(52004238)China Postdoctoral Science Foundation(2019M663561).
文摘Increasing the production and utilization of shale gas is of great significance for building a clean and low-carbon energy system.Sharp decline of gas production has been widely observed in shale gas reservoirs.How to forecast shale gas production is still challenging due to complex fracture networks,dynamic fracture properties,frac hits,complicated multiphase flow,and multi-scale flow as well as data quality and uncertainty.This work develops an integrated framework for evaluating shale gas well production based on data-driven models.Firstly,a comprehensive dominated-factor system has been established,including geological,drilling,fracturing,and production factors.Data processing and visualization are required to ensure data quality and determine final data set.A shale gas production evaluation model is developed to evaluate shale gas production levels.Finally,the random forest algorithm is used to forecast shale gas production.The prediction accuracy of shale gas production level is higher than 95%based on the shale gas reservoirs in China.Forty-one wells are randomly selected to predict cumulative gas production using the optimal regression model.The proposed shale gas production evaluation frame-work overcomes too many assumptions of analytical or semi-analytical models and avoids huge computation cost and poor generalization for numerical modelling.
文摘The wind–thermal bundled power system achieves energy complementarity and optimized scheduling, which is an important way to build a new type of energy system. For the safe and stable operation of the wind–thermal bundled power system, accurate data-driven analysis is necessary to maintain real-time balance between electricity supply and demand. By summarizing the development and characteristics of wind–thermal bundled power system in China and different countries, current research in this field can be clearly defined in two aspects: short-term wind power prediction for wind farms and performance evaluation of automatic generation control (AGC) for thermal power generation units. For short-term wind power prediction, it is recommended to focus on historical data preprocessing and artificial intelligence methods. The technical characteristics of different data-driven wind power prediction methods have been compared in detail. For performance evaluation of AGC units, a comprehensive analysis was conducted on current evaluation methods, including the “permitted-band” and “regulation mileage” methods, as well as the issue of evaluation failure in traditional evaluation methods in practical engineering. Finally, the relative optimal dynamic performance of AGC units was discussed and the future trend of data-driven research in wind–thermal bundled power system was summarized.
基金Supported by National Hi-tech Research and Development Program of China(863 Program,Grant No.2015AA042101)
文摘Complex engineered systems are often difficult to analyze and design due to the tangled interdependencies among their subsystems and components. Conventional design methods often need exact modeling or accurate structure decomposition, which limits their practical application. The rapid expansion of data makes utilizing data to guide and improve system design indispensable in practical engineering. In this paper, a data driven uncertainty evaluation approach is proposed to support the design of complex engineered systems. The core of the approach is a data-mining based uncertainty evaluation method that predicts the uncertainty level of a specific system design by means of analyzing association relations along different system attributes and synthesizing the information entropy of the covered attribute areas, and a quantitative measure of system uncertainty can be obtained accordingly. Monte Carlo simulation is introduced to get the uncertainty extrema, and the possible data distributions under different situations is discussed in detail The uncertainty values can be normalized using the simulation results and the values can be used to evaluate different system designs. A prototype system is established, and two case studies have been carded out. The case of an inverted pendulum system validates the effectiveness of the proposed method, and the case of an oil sump design shows the practicability when two or more design plans need to be compared. This research can be used to evaluate the uncertainty of complex engineered systems completely relying on data, and is ideally suited for plan selection and performance analysis in system design.
基金Supported by the Key Program of Petro China Exploration&Production Company(Grant No.kt2017-17-01-1 and kt2017-17-06-1)Consulting Project of Chinese Academy of Engineering(Grant No.2019-XZ-17)
文摘Based on the traditional numerical simulation and optimization algorithms,in combination with the layered injection and production"hard data"monitored at real time by automatic control technology,a systematic approach for detailed water injection design using data-driven algorithms is proposed.First the data assimilation technology is used to match geological model parameters under the constraint of observed well dynamics;the flow relationships between injectors and producers in the block are calculated based on automatic identification method for layered injection-production flow relationship;multi-layer and multi-direction production splitting technique is used to calculate the liquid and oil production of producers in different layers and directions and obtain quantified indexes of water injection effect.Then,machine learning algorithms are applied to evaluate the effectiveness of water injection in different layers of wells and to perform the water injection direction adjustment.Finally,the particle swarm algorithm is used to optimize the detailed water injection plan and to make production predictions.This method and procedure make full use of the automation and intelligence of data-driven and machine learning algorithms.This method was used to match the data of a complex faulted reservoir in eastern China,achieving a fitting level of 85%.The cumulative oil production in the example block for 12 months after optimization is 8.2%higher than before.This method can help design detailed water injection program for mature oilfields.
基金This project was supported by the SAIC Motor Industry-University Cooperative Research Foundation(No.1912).
文摘With more commercialized automated vehicles(AVs)shortly entering the market,evaluating their fuel economy has become an important topic.The traditional fixed-profile test methods only indicate the effect of powertrain efficiency on fuel economy and cannot reflect the influences of control algorithms or driving behaviors on fuel consumption under real-world traffic conditions.Therefore,a data-driven simulation method for evaluating the real driving fuel economy of automated vehicles is developed.It utilizes naturalistic driving data to reconstruct test cycles.This method can inspect the performance of automated vehicle control algorithms under realistic traffic conditions in terms of fuel economy.The naturalistic driving data collected on urban expressways and freeways were used to model the longitudinal driving scenarios.Then,the fuel consumption of automated and human-driven vehicle was evaluated under the simulated scenarios via the Monte Carlo integration approach.It was found that the fuel consumption rate of the automated vehicle was 11.944 L/100 km under the car-following scenarios.The human-driven vehicle had a fuel consumption rate of 10.124 L/100 km under the same traffic conditions.The tested automated vehicle control algorithm is tuned to achieve better performance in terms of safety and travel efficiency and hence it tends to maintain a relatively steady time headway to the leading vehicle.It applies more frequent accelerating/braking cycles and higher average speed compared to typical human drivers.These characteristics lead to a higher fuel consumption rate of the automated vehicle.The presented method provides an accurate and efficient way to analyze the fuel economy performance of automated vehicles under practical conditions.This method can easily be scaled for large-scale traffic flow analyses.It can also be used to study the effects of human driving styles on fuel economy.
