[ Objective] The study aimed to analyze and predict the suitable period of laver along the coastal areas of Jiangsu Province. [ Method ] TO enhance the ability of meteorology to service laver culture, using the modern...[ Objective] The study aimed to analyze and predict the suitable period of laver along the coastal areas of Jiangsu Province. [ Method ] TO enhance the ability of meteorology to service laver culture, using the modern weather forecast technology, agricultural weather prediction was de- veloped according to the demands for meteorological conditions during laver production. [ Result] From south to north, there were certain differ- ences in the suitable periods of laver culture, breeding and harvesting, with slight variations. The forecast of the first and last days of certain water temperature could provide scientific references for the meteorological service of laver culture, and the service effect became better after it was modi- fied with the aid of the medium-term and long-term weather prediction. [ Conclusion] The research can offer theoretical bases for the culture of laver along the coastal areas of Jiangsu Province.展开更多
A laboratory leaching experiment with samples of different grades was carried out, and an analytical method of concentration of leaching solution was put forward. For each sample, respectively, by applying phase space...A laboratory leaching experiment with samples of different grades was carried out, and an analytical method of concentration of leaching solution was put forward. For each sample, respectively, by applying phase space reconstruction for time series of monitoring data, the saturated embedding dimension and the correlation dimension were obtained, and the evolution laws between neighboring points in the reconstructed phase space were revealed. With BP neural network, a prediction model of concentration of leaching solution was set up and the maximum error of which was less than 2%. The results show that there exist chaotic characteristics in leaching system, and samples of different grades have different nonlinear dynamic features; the higher the grade of sample, the smaller the correlation dimension; furthermore, the maximum Lyapunov index, energy dissipation and chaotic extent of the leaching system increase with grade of the sample; by phase space reconstruction, the subtle change features of concentration of leaching solution can be magnified and the inherent laws can be fully demonstrated. According to the laws, a prediction model of leaching cycle period has been established to provide a theoretical foundation for solution mining.展开更多
We propose the pseudo-periodicity method and its quantitative prediction indexes for the occurrence time of earlier strong aftershock. We conducted tests of regressive prediction, and the R-value of the tests is 0.45,...We propose the pseudo-periodicity method and its quantitative prediction indexes for the occurrence time of earlier strong aftershock. We conducted tests of regressive prediction, and the R-value of the tests is 0.45, indicating that this method is effective for prediction.展开更多
针对目前大多数机器学习模型预测材料性质时需要大量的先验知识以及特征向量筛选困难的问题,基于电子轨道矩阵和元素周期表法两种描述符,通过特征融合的方式,设计了一种卷积神经网络模型OPCNN(Orbital of electron and Periodic table C...针对目前大多数机器学习模型预测材料性质时需要大量的先验知识以及特征向量筛选困难的问题,基于电子轨道矩阵和元素周期表法两种描述符,通过特征融合的方式,设计了一种卷积神经网络模型OPCNN(Orbital of electron and Periodic table CNN)。实验数据表明,OPCNN与其他预测模型相比,在带隙、生成热以及形成能数据集上都有着更好的性能,平均绝对误差分别为0.26 eV、0.037 KJ/mol和0.073 eV/atom,且R^(2)都达到了91%以上。OPCNN在保证了预测准确性的同时对先验知识的要求更低,只需要元素周期表中的信息即可预测材料性质,特征融合的思想可以让特征设计更加灵活,有利于新材料体系快速和准确的预测。展开更多
基金Supported by National Research Fund for Public Welfare (Meteorology) of China (GYHY201006029)Meteorological Scientific Research Open Fund of Jiangsu Province,China (ZD201108)
文摘[ Objective] The study aimed to analyze and predict the suitable period of laver along the coastal areas of Jiangsu Province. [ Method ] TO enhance the ability of meteorology to service laver culture, using the modern weather forecast technology, agricultural weather prediction was de- veloped according to the demands for meteorological conditions during laver production. [ Result] From south to north, there were certain differ- ences in the suitable periods of laver culture, breeding and harvesting, with slight variations. The forecast of the first and last days of certain water temperature could provide scientific references for the meteorological service of laver culture, and the service effect became better after it was modi- fied with the aid of the medium-term and long-term weather prediction. [ Conclusion] The research can offer theoretical bases for the culture of laver along the coastal areas of Jiangsu Province.
基金Project(51374035)supported by the National Natural Science Foundation of ChinaProject(2012BAB08B02)supported by the National“Twelfth Five”Science and Technology,ChinaProject(NCET-13-0669)supported by New Century Excellent Talents in University of Ministry of Education of China
文摘A laboratory leaching experiment with samples of different grades was carried out, and an analytical method of concentration of leaching solution was put forward. For each sample, respectively, by applying phase space reconstruction for time series of monitoring data, the saturated embedding dimension and the correlation dimension were obtained, and the evolution laws between neighboring points in the reconstructed phase space were revealed. With BP neural network, a prediction model of concentration of leaching solution was set up and the maximum error of which was less than 2%. The results show that there exist chaotic characteristics in leaching system, and samples of different grades have different nonlinear dynamic features; the higher the grade of sample, the smaller the correlation dimension; furthermore, the maximum Lyapunov index, energy dissipation and chaotic extent of the leaching system increase with grade of the sample; by phase space reconstruction, the subtle change features of concentration of leaching solution can be magnified and the inherent laws can be fully demonstrated. According to the laws, a prediction model of leaching cycle period has been established to provide a theoretical foundation for solution mining.
文摘We propose the pseudo-periodicity method and its quantitative prediction indexes for the occurrence time of earlier strong aftershock. We conducted tests of regressive prediction, and the R-value of the tests is 0.45, indicating that this method is effective for prediction.
文摘针对目前大多数机器学习模型预测材料性质时需要大量的先验知识以及特征向量筛选困难的问题,基于电子轨道矩阵和元素周期表法两种描述符,通过特征融合的方式,设计了一种卷积神经网络模型OPCNN(Orbital of electron and Periodic table CNN)。实验数据表明,OPCNN与其他预测模型相比,在带隙、生成热以及形成能数据集上都有着更好的性能,平均绝对误差分别为0.26 eV、0.037 KJ/mol和0.073 eV/atom,且R^(2)都达到了91%以上。OPCNN在保证了预测准确性的同时对先验知识的要求更低,只需要元素周期表中的信息即可预测材料性质,特征融合的思想可以让特征设计更加灵活,有利于新材料体系快速和准确的预测。
