文章提出了一种基于长短期记忆网络(long short term memory network,LSTM),面向光伏发电功率预测的数字孪生模型,并通过迁移学习将此模型应用到其他投入运行时间较短、数据不足的光伏系统发电功率预测中。光伏发电功率由于受到太阳辐...文章提出了一种基于长短期记忆网络(long short term memory network,LSTM),面向光伏发电功率预测的数字孪生模型,并通过迁移学习将此模型应用到其他投入运行时间较短、数据不足的光伏系统发电功率预测中。光伏发电功率由于受到太阳辐照度、温度和一些随机因素的影响,具有较强的间歇性和波动性,因此很难进行精确的光伏功率预测;所提出的数字孪生模型,实现了与光伏系统物理实体的同步和实时更新,因此获得比传统预测方法更准确的预测结果,同时利用从历史数据充足的光伏系统中学到的知识来辅助历史数据有限的光伏系统建立发电功率预测数字孪生模型,不仅可以得到精确的预测结果而且节省了模型训练时间。文中通过Queensland大学开源网站中3个不同站点以及山西晋能清洁能源公司的光伏历史数据验证了所提方法的有效性。展开更多
The emergence of a digital divide or a lack of internet access may hinder urbanization and adversely affect growth.However,fintech development can help to improve the accessibility and affordability of financial servi...The emergence of a digital divide or a lack of internet access may hinder urbanization and adversely affect growth.However,fintech development can help to improve the accessibility and affordability of financial services,particularly for sections of the population formerly excluded from such services.Improved access is expected to stimulate the growth essential for promoting urbanization,possibly alleviating the negative impacts of a digital divide and creating a trickle-down effect.This paper is among the first to investigate the effect of fintech development on urbanization,as indicated by labor transfer from agricultural to non-agricultural sectors in China.It is found that:(i)the digital divide hindered,while fintech development contributed to urbanization;(ii)the positive effect of fintech development on urbanization affects individuals both with and without access to the internet,yielding the so-called trickledown ejfect and thus helping to alleviate the negative consequences of a digital divide;and(Hi)regarding transmission mechanisms,fintech development helps to generate additional jobs and raise income in non-agricultural sectors,stimulating urbanization even for those without access to the internet.展开更多
文摘文章提出了一种基于长短期记忆网络(long short term memory network,LSTM),面向光伏发电功率预测的数字孪生模型,并通过迁移学习将此模型应用到其他投入运行时间较短、数据不足的光伏系统发电功率预测中。光伏发电功率由于受到太阳辐照度、温度和一些随机因素的影响,具有较强的间歇性和波动性,因此很难进行精确的光伏功率预测;所提出的数字孪生模型,实现了与光伏系统物理实体的同步和实时更新,因此获得比传统预测方法更准确的预测结果,同时利用从历史数据充足的光伏系统中学到的知识来辅助历史数据有限的光伏系统建立发电功率预测数字孪生模型,不仅可以得到精确的预测结果而且节省了模型训练时间。文中通过Queensland大学开源网站中3个不同站点以及山西晋能清洁能源公司的光伏历史数据验证了所提方法的有效性。
基金This paper was funded by the National Natural Science Foundation of China(Nos.7170308&71833003 and 71973014).
文摘The emergence of a digital divide or a lack of internet access may hinder urbanization and adversely affect growth.However,fintech development can help to improve the accessibility and affordability of financial services,particularly for sections of the population formerly excluded from such services.Improved access is expected to stimulate the growth essential for promoting urbanization,possibly alleviating the negative impacts of a digital divide and creating a trickle-down effect.This paper is among the first to investigate the effect of fintech development on urbanization,as indicated by labor transfer from agricultural to non-agricultural sectors in China.It is found that:(i)the digital divide hindered,while fintech development contributed to urbanization;(ii)the positive effect of fintech development on urbanization affects individuals both with and without access to the internet,yielding the so-called trickledown ejfect and thus helping to alleviate the negative consequences of a digital divide;and(Hi)regarding transmission mechanisms,fintech development helps to generate additional jobs and raise income in non-agricultural sectors,stimulating urbanization even for those without access to the internet.
