Pub Date : 2021-07-18DOI: 10.1109/ICPSAsia52756.2021.9621683
Z. Weng, Jinghua Zhou, Xiaotong Song, Jin Li
Unlike grid-connected microgrids, isolated microgrids are more susceptible to internal equipment capacity changes and external dispatching strategies, so it is necessary to analyze microgrid reliability from the perspective of capacity changes. Firstly, a time series model of equipment life process, a PV model with Beta distribution, a load model with time variability and stochasticity, and an energy storage model with operational constraints are studied. Secondly, the reliability index is used to characterize the system reliability level, and the power scheduling and load shedding strategy with minimum load shedding is proposed. Finally, the improved RBTS BUS6 F4 system is used as an example to deeply analyze the interaction mechanism of each device capacity on the reliability of the microgrid and to show the effectiveness of the proposed strategy.
{"title":"Reliability Evaluation of PV-ESS Microgrid System","authors":"Z. Weng, Jinghua Zhou, Xiaotong Song, Jin Li","doi":"10.1109/ICPSAsia52756.2021.9621683","DOIUrl":"https://doi.org/10.1109/ICPSAsia52756.2021.9621683","url":null,"abstract":"Unlike grid-connected microgrids, isolated microgrids are more susceptible to internal equipment capacity changes and external dispatching strategies, so it is necessary to analyze microgrid reliability from the perspective of capacity changes. Firstly, a time series model of equipment life process, a PV model with Beta distribution, a load model with time variability and stochasticity, and an energy storage model with operational constraints are studied. Secondly, the reliability index is used to characterize the system reliability level, and the power scheduling and load shedding strategy with minimum load shedding is proposed. Finally, the improved RBTS BUS6 F4 system is used as an example to deeply analyze the interaction mechanism of each device capacity on the reliability of the microgrid and to show the effectiveness of the proposed strategy.","PeriodicalId":296085,"journal":{"name":"2021 IEEE/IAS Industrial and Commercial Power System Asia (I&CPS Asia)","volume":"28 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-07-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"129906129","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
With the wide application of distributed generation and advanced power electronic technology, AC / DC hybrid distribution network provides a new idea for the development of distribution network. Therefore, the AC-DC hybrid distribution network planning is of great significance. This paper presents a bi-level programming model of AC / DC distribution network considering the randomness of DC type of bus and connecting line. The model uses Monte Carlo method to simulate the uncertain output curve of load demand and photovoltaic distributed generation, and divides the planning scheme into upper planning model and lower optimization model: the upper planning model uses genetic membrane algorithm, with the lowest grid construction cost and operation cost as the objective function; the lower optimization model uses interior point method, with the lowest operation cost as the objective function. The upper and lower layer models are solved alternately to obtain the optimal network configuration.
{"title":"Bi-level Programming of AC / DC Hybrid Distribution System","authors":"Chunyi Wang, Zijia Ding, Xingyou Zhang, Zhaohui He, Huali Liu, Zhendong Xu","doi":"10.1109/ICPSAsia52756.2021.9621429","DOIUrl":"https://doi.org/10.1109/ICPSAsia52756.2021.9621429","url":null,"abstract":"With the wide application of distributed generation and advanced power electronic technology, AC / DC hybrid distribution network provides a new idea for the development of distribution network. Therefore, the AC-DC hybrid distribution network planning is of great significance. This paper presents a bi-level programming model of AC / DC distribution network considering the randomness of DC type of bus and connecting line. The model uses Monte Carlo method to simulate the uncertain output curve of load demand and photovoltaic distributed generation, and divides the planning scheme into upper planning model and lower optimization model: the upper planning model uses genetic membrane algorithm, with the lowest grid construction cost and operation cost as the objective function; the lower optimization model uses interior point method, with the lowest operation cost as the objective function. The upper and lower layer models are solved alternately to obtain the optimal network configuration.","PeriodicalId":296085,"journal":{"name":"2021 IEEE/IAS Industrial and Commercial Power System Asia (I&CPS Asia)","volume":"33 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-07-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"123764563","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2021-07-18DOI: 10.1109/ICPSAsia52756.2021.9621679
Zhang Xingyou, Wang Yuejiao, Chen Jian, Yu Peng
Based on common characteristic that liquid hydrogen storage system(LHSS) and superconducting magnetic energy storage(SMES) has similar requirement for refrigeration, This paper analyzes advantages of joint operation of LHSS combined SMES system based on their common characteristics. Consider current situation of renewable energy utilization difficulties in recent years, this paper studies the optimal capacity configuration of LHSS combined SMES hybrid energy storage system to maximum use renewable energy. Based on above scene, a time series based method is proposed to simulate the operation scenario of a high penetration renewable energy system, and a capacity optimization method which objects to maximize the economic benefit of hybrid energy storage system is also proposed in this paper. Also, the effectiveness of the proposed strategy is verified by a simple case.
{"title":"The Utility Analysis of LHSS Combined SMES System Considering Capacity Configuration","authors":"Zhang Xingyou, Wang Yuejiao, Chen Jian, Yu Peng","doi":"10.1109/ICPSAsia52756.2021.9621679","DOIUrl":"https://doi.org/10.1109/ICPSAsia52756.2021.9621679","url":null,"abstract":"Based on common characteristic that liquid hydrogen storage system(LHSS) and superconducting magnetic energy storage(SMES) has similar requirement for refrigeration, This paper analyzes advantages of joint operation of LHSS combined SMES system based on their common characteristics. Consider current situation of renewable energy utilization difficulties in recent years, this paper studies the optimal capacity configuration of LHSS combined SMES hybrid energy storage system to maximum use renewable energy. Based on above scene, a time series based method is proposed to simulate the operation scenario of a high penetration renewable energy system, and a capacity optimization method which objects to maximize the economic benefit of hybrid energy storage system is also proposed in this paper. Also, the effectiveness of the proposed strategy is verified by a simple case.","PeriodicalId":296085,"journal":{"name":"2021 IEEE/IAS Industrial and Commercial Power System Asia (I&CPS Asia)","volume":"220 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-07-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"121940398","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Distribution terminal unit (DTU) is an indispensable device that monitors and controls other secondary-side devices in distribution network. With the large-scale integration of DTU into power system, limited numbers of maintenance engineers can hardly cope with the pressure brought about by the massive increment of DTU devices. Fault of DTU may lead to severe impact on normal operation of distribution network and power system. Therefore, researching DTU fault mechanisms and taking measures to lower fault incidence are crucial for ensuring the stable and safe operation of distribution network. In this paper, we proposed a FP-growth algorithm-based fault analysis method for DTU. Firstly, different possible factors that lead to DTU fault are analyzed, fault types are also introduced. Then the main idea of FP-growth algorithm is illustrated. Afterwards, a DTU fault analysis method based on FP-growth algorithm is put forward. Experiment is conducted on realistic DTU fault database to prove the practicality of this method.
{"title":"A FP-Growth Algorithm Based Fault Analysis Method for Distribution Terminal Unit","authors":"Xuecen Zhang, Qiang Liu, Yi Tang, Guofeng Liu, Xin Ning, Jiankun Chen","doi":"10.1109/ICPSAsia52756.2021.9621594","DOIUrl":"https://doi.org/10.1109/ICPSAsia52756.2021.9621594","url":null,"abstract":"Distribution terminal unit (DTU) is an indispensable device that monitors and controls other secondary-side devices in distribution network. With the large-scale integration of DTU into power system, limited numbers of maintenance engineers can hardly cope with the pressure brought about by the massive increment of DTU devices. Fault of DTU may lead to severe impact on normal operation of distribution network and power system. Therefore, researching DTU fault mechanisms and taking measures to lower fault incidence are crucial for ensuring the stable and safe operation of distribution network. In this paper, we proposed a FP-growth algorithm-based fault analysis method for DTU. Firstly, different possible factors that lead to DTU fault are analyzed, fault types are also introduced. Then the main idea of FP-growth algorithm is illustrated. Afterwards, a DTU fault analysis method based on FP-growth algorithm is put forward. Experiment is conducted on realistic DTU fault database to prove the practicality of this method.","PeriodicalId":296085,"journal":{"name":"2021 IEEE/IAS Industrial and Commercial Power System Asia (I&CPS Asia)","volume":"7 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-07-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"121281694","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2021-07-18DOI: 10.1109/ICPSAsia52756.2021.9621391
Encheng Dong, W. Cong, Jian Chen, Ming Chen, Tianyou Yang, Zhen Wei
Aiming at the optimal utilization problem of integrated energy for commercial users, the stepped utilization of energy was firstly analyzed in typical combined cooling, heating and power(CCHP) systems. Secondly, considering the coupling structure of the system and energy utilization characteristics of commercial users, a physical framework of integrated energy system(IES) was set up for multi-energy optimal utilization. Devices of energy production, conversion and storage was modelled on the source side, and meanwhile, on end-user side, load models were built according to space cooling, space heating, hot water and electric. Then, taking into account the time-of-use and costs of operation and maintenance, a mathematical model of IES optimal utilization was established, which was solved by YALMIP/Gurobi. Finally, Case study proves the economy and effectiveness of the proposed optimization strategy.
{"title":"Optimal Utilization Strategy of Integrated Energy for Commercial Users Based on CCHP Coupling","authors":"Encheng Dong, W. Cong, Jian Chen, Ming Chen, Tianyou Yang, Zhen Wei","doi":"10.1109/ICPSAsia52756.2021.9621391","DOIUrl":"https://doi.org/10.1109/ICPSAsia52756.2021.9621391","url":null,"abstract":"Aiming at the optimal utilization problem of integrated energy for commercial users, the stepped utilization of energy was firstly analyzed in typical combined cooling, heating and power(CCHP) systems. Secondly, considering the coupling structure of the system and energy utilization characteristics of commercial users, a physical framework of integrated energy system(IES) was set up for multi-energy optimal utilization. Devices of energy production, conversion and storage was modelled on the source side, and meanwhile, on end-user side, load models were built according to space cooling, space heating, hot water and electric. Then, taking into account the time-of-use and costs of operation and maintenance, a mathematical model of IES optimal utilization was established, which was solved by YALMIP/Gurobi. Finally, Case study proves the economy and effectiveness of the proposed optimization strategy.","PeriodicalId":296085,"journal":{"name":"2021 IEEE/IAS Industrial and Commercial Power System Asia (I&CPS Asia)","volume":"10 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-07-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"125524812","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
As it is a severe challenge to accommodate the increasing new energy generation integrated to the power grid, electricity spot market can serve to address this issue by producing real-time time-of-use price signals and guiding consumers to adjust their electricity consumption behaviors. To resolve the dilemma of new energy consumption, this paper first designed a provincial bilateral electricity spot market, and put forward the trading mechanism considering both power plants and consumers. Then, a comparative analysis was conducted on the bilateral market clearing results and the unilateral market clearing results considering only power plants through case study. The results show that integrating consumers in the spot market can help to encourage them to consume new energy in an active manner. The bilateral electricity spot market proposed in this paper can provide a reference for improving the new energy consumption in regions with high new energy penetration.
{"title":"Design of Provincial Bilateral Electricity Spot Market in China under High New Energy Penetration","authors":"Ruhai Hao, Xujun Zhang, Zhenhuan Chen, Zan Yang, Tianrui Zhang, Xiaoting Chen","doi":"10.1109/ICPSAsia52756.2021.9621605","DOIUrl":"https://doi.org/10.1109/ICPSAsia52756.2021.9621605","url":null,"abstract":"As it is a severe challenge to accommodate the increasing new energy generation integrated to the power grid, electricity spot market can serve to address this issue by producing real-time time-of-use price signals and guiding consumers to adjust their electricity consumption behaviors. To resolve the dilemma of new energy consumption, this paper first designed a provincial bilateral electricity spot market, and put forward the trading mechanism considering both power plants and consumers. Then, a comparative analysis was conducted on the bilateral market clearing results and the unilateral market clearing results considering only power plants through case study. The results show that integrating consumers in the spot market can help to encourage them to consume new energy in an active manner. The bilateral electricity spot market proposed in this paper can provide a reference for improving the new energy consumption in regions with high new energy penetration.","PeriodicalId":296085,"journal":{"name":"2021 IEEE/IAS Industrial and Commercial Power System Asia (I&CPS Asia)","volume":"12 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-07-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"126429119","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2021-07-18DOI: 10.1109/ICPSAsia52756.2021.9621660
Xiaoming Xiao, Ye Cai, Ying Liu, Zheng-yi Li, Chen Xue
Peak shaving on the user side is an effective means to promote the consumption of clean energy. This study proposes a user-side participation in peak shaving potential evaluation model considering the promotion of clean energy consumption in order to give full play to the advantages of user-side participation in peak shaving, fast response and high economy and to improve user participation enthusiasm. Aiming at the problem of multi-time scale volatility that traditional static methods cannot reflect the peak shaving potential index, this study proposes a dynamic incentive evaluation method by extracting the multi-day typical characteristics of multiple types of users participating in peak shaving. First, interval weights were constructed to reduce the impact of user load fluctuations on evaluation indicators; the idea of acceleration was introduced to describe the trend of changes in the observed values of indicators at different time scales; second, the gain range of the advantages and disadvantages of each type of indicator was calculated. In addition, assigning corresponding ”reward” and ”punishment” values can reflect the current state of the evaluated object more objectively and truly, thereby obtaining dynamic incentive index values; finally, based on interval weights and dynamic incentive index values, an interval-type decision matrix pair was formed. The potential of various types of users to participate in peak shaving was scored and sorted. Taking the load of five types of users in a certain area as an example, the results show that this method can effectively evaluate the peak shaving potential of different types of loads.
{"title":"Evaluation Method of Peak Shaving Potential for Multi-type Users Considering the Promotion of Clean Energy Consumption","authors":"Xiaoming Xiao, Ye Cai, Ying Liu, Zheng-yi Li, Chen Xue","doi":"10.1109/ICPSAsia52756.2021.9621660","DOIUrl":"https://doi.org/10.1109/ICPSAsia52756.2021.9621660","url":null,"abstract":"Peak shaving on the user side is an effective means to promote the consumption of clean energy. This study proposes a user-side participation in peak shaving potential evaluation model considering the promotion of clean energy consumption in order to give full play to the advantages of user-side participation in peak shaving, fast response and high economy and to improve user participation enthusiasm. Aiming at the problem of multi-time scale volatility that traditional static methods cannot reflect the peak shaving potential index, this study proposes a dynamic incentive evaluation method by extracting the multi-day typical characteristics of multiple types of users participating in peak shaving. First, interval weights were constructed to reduce the impact of user load fluctuations on evaluation indicators; the idea of acceleration was introduced to describe the trend of changes in the observed values of indicators at different time scales; second, the gain range of the advantages and disadvantages of each type of indicator was calculated. In addition, assigning corresponding ”reward” and ”punishment” values can reflect the current state of the evaluated object more objectively and truly, thereby obtaining dynamic incentive index values; finally, based on interval weights and dynamic incentive index values, an interval-type decision matrix pair was formed. The potential of various types of users to participate in peak shaving was scored and sorted. Taking the load of five types of users in a certain area as an example, the results show that this method can effectively evaluate the peak shaving potential of different types of loads.","PeriodicalId":296085,"journal":{"name":"2021 IEEE/IAS Industrial and Commercial Power System Asia (I&CPS Asia)","volume":"7 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-07-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"127661092","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2021-07-18DOI: 10.1109/ICPSAsia52756.2021.9621468
D. Xia
In order to analyze the local overheating of single phase three-legged transformer under AC/DC condition, the transient magnetic field and eddy current loss in the iron core of transformer are studied. This paper builds a coupling model of electric circuit and magnetic circuit considering nonlinearity of ferromagnetic materials and iron core topology. The distribution and variation of magnetic field and eddy current loss of single phase three-legged transformer are researched. The results show that the amplitudes of magnetic field and eddy current loss in the central leg are greater than those in two side legs without DC, and they grow nonlinearly as DC enlarges. The magnetic field amplitudes in the central leg grow faster than those in two side legs, while the eddy current loss amplitudes in two side legs grow faster than those in the central leg.
{"title":"The Research of Transient Magnetic Field and Eddy Current Loss of Single Phase Three-Legged Transformer under AC/DC Condition","authors":"D. Xia","doi":"10.1109/ICPSAsia52756.2021.9621468","DOIUrl":"https://doi.org/10.1109/ICPSAsia52756.2021.9621468","url":null,"abstract":"In order to analyze the local overheating of single phase three-legged transformer under AC/DC condition, the transient magnetic field and eddy current loss in the iron core of transformer are studied. This paper builds a coupling model of electric circuit and magnetic circuit considering nonlinearity of ferromagnetic materials and iron core topology. The distribution and variation of magnetic field and eddy current loss of single phase three-legged transformer are researched. The results show that the amplitudes of magnetic field and eddy current loss in the central leg are greater than those in two side legs without DC, and they grow nonlinearly as DC enlarges. The magnetic field amplitudes in the central leg grow faster than those in two side legs, while the eddy current loss amplitudes in two side legs grow faster than those in the central leg.","PeriodicalId":296085,"journal":{"name":"2021 IEEE/IAS Industrial and Commercial Power System Asia (I&CPS Asia)","volume":"123 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-07-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"127407946","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2021-07-18DOI: 10.1109/ICPSAsia52756.2021.9621550
Haiyang Liu, Xiaobo Ling, Yifan Zhu, Qiyu Tu, Piao Gao, Chenxu Chao, Yaxin Xie, Xiaodong Zheng, N. Tai
Large-scale distributed photovoltaics (DPs) are connected to the distribution network. when a fault happens in the distribution network, the bus voltage at the fault location cannot immediately decrease to zero due to the existence of DPs, which invalidates the criteria of non-voltage detection and synchronism check and impedes the automatic switching of backup power. Therefore, considering the influence of DPs, this paper puts forward a configuration optimization scheme of backup power automatic switching (BPAS) strategy, which allows the BPAS devices in the distribution network with DPs to conduct non-voltage detection and synchronism check. The example analysis shows that the optimized BPAS strategy can effectively improve the load transfer capacity and reduce the cost of BPAS devices.
{"title":"Optimization of Backup Power Automatic Switching Strategy in the Distribution Network with Distributed Photovoltaics","authors":"Haiyang Liu, Xiaobo Ling, Yifan Zhu, Qiyu Tu, Piao Gao, Chenxu Chao, Yaxin Xie, Xiaodong Zheng, N. Tai","doi":"10.1109/ICPSAsia52756.2021.9621550","DOIUrl":"https://doi.org/10.1109/ICPSAsia52756.2021.9621550","url":null,"abstract":"Large-scale distributed photovoltaics (DPs) are connected to the distribution network. when a fault happens in the distribution network, the bus voltage at the fault location cannot immediately decrease to zero due to the existence of DPs, which invalidates the criteria of non-voltage detection and synchronism check and impedes the automatic switching of backup power. Therefore, considering the influence of DPs, this paper puts forward a configuration optimization scheme of backup power automatic switching (BPAS) strategy, which allows the BPAS devices in the distribution network with DPs to conduct non-voltage detection and synchronism check. The example analysis shows that the optimized BPAS strategy can effectively improve the load transfer capacity and reduce the cost of BPAS devices.","PeriodicalId":296085,"journal":{"name":"2021 IEEE/IAS Industrial and Commercial Power System Asia (I&CPS Asia)","volume":"72 16","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-07-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"133587094","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2021-07-18DOI: 10.1109/ICPSAsia52756.2021.9621522
Yuanfan Ji, Guangchao Geng, Q. Jiang
Concept drift refers to the relation between input and target envolves over time in an online supervised learning scenario. With the development of smart grid and smart meter, mass data accessibility poses a huge challenge to learning model adaptability. To address such issue, this paper proposed a model adaptability enhancement approaches based on concept drift detection for short-term forecast model. It exploits canonical correlation analysis to measure mapping relation between input and output of the forecast model. Then the correlation coefficient vector sequence will be monitored over an adaptive window to detect concept drift. Model is updated on data over the sliding window only when a concept drift happens. Experimental results shows this method reduces memory assumption and computing resources remarkably meanwhile guarantees the forecast accuracy.
{"title":"Enhancing Model Adaptability Using Concept Drift Detection for Short-Term Load Forecast","authors":"Yuanfan Ji, Guangchao Geng, Q. Jiang","doi":"10.1109/ICPSAsia52756.2021.9621522","DOIUrl":"https://doi.org/10.1109/ICPSAsia52756.2021.9621522","url":null,"abstract":"Concept drift refers to the relation between input and target envolves over time in an online supervised learning scenario. With the development of smart grid and smart meter, mass data accessibility poses a huge challenge to learning model adaptability. To address such issue, this paper proposed a model adaptability enhancement approaches based on concept drift detection for short-term forecast model. It exploits canonical correlation analysis to measure mapping relation between input and output of the forecast model. Then the correlation coefficient vector sequence will be monitored over an adaptive window to detect concept drift. Model is updated on data over the sliding window only when a concept drift happens. Experimental results shows this method reduces memory assumption and computing resources remarkably meanwhile guarantees the forecast accuracy.","PeriodicalId":296085,"journal":{"name":"2021 IEEE/IAS Industrial and Commercial Power System Asia (I&CPS Asia)","volume":"42 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-07-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"116644352","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}