为了提升机动目标的航迹预测精准度,提出了一种粒子群(Particle Swarm Optimization,PSO)算法与长短周期记忆单元(Long Short Term Memory,LSTM)神经网络相结合的PSO-LSTM目标航迹预测模型。针对LSTM神经网络中的超参数难以人工最优化...为了提升机动目标的航迹预测精准度,提出了一种粒子群(Particle Swarm Optimization,PSO)算法与长短周期记忆单元(Long Short Term Memory,LSTM)神经网络相结合的PSO-LSTM目标航迹预测模型。针对LSTM神经网络中的超参数难以人工最优化的问题,通过PSO算法进行自动调参,对诸如LSTM隐藏层规模、学习率、训练集批次规模等参数进行优化。使用PSO-LSTM航迹预测模型在真实航迹数据集上进行预测,实验结果表明,PSO-LSTM航迹预测模型在MAE、RMSE等指标上优于传统模型,有着较好的准确性与稳定性。展开更多
The physical-mechanical,chemical,and durability characteristics of alkali-activated materials(AAMs)have been widely investigated.However,a critical gap in the literature is the lack of a comprehensive overview of rece...The physical-mechanical,chemical,and durability characteristics of alkali-activated materials(AAMs)have been widely investigated.However,a critical gap in the literature is the lack of a comprehensive overview of recently published literature regarding the life cycle assessment(LCA)of these binders.This study aims to fill that gap by conducting a systematic literature review of globally published literature on the topic.This paper consolidates knowledge by searching different databases,focusing on LCA studies that used AAMs as pastes,mortars,concretes,bricks,and rammed earth/soil blocks.The selected articles were reviewed and categorized based on precursors,alkaline activators,functional units,system boundaries,life cycle inventory databases,allocation,impact methodologies,and software used.Additionally,this paper also critically analyzes the key challenges of LCA for AAMs.The major challenges were identified as selecting a functional unit,subjectivity in boundary systems,and data interpretation.This work concludes that AAMs show substantial advantages in global warming potential compared to ordinary Portland cement-based materials;however,the average of other categories such as marine ecotoxicity and ozone layer depletion has been reported to be higher than for the reference samples.展开更多
An effective and flexible rotation and compensation scheme is designed to improve the accuracy of rotating inertial navigation system （RINS）. The accuracy of single-axial R1NS is limited by the errors on the rotatin...An effective and flexible rotation and compensation scheme is designed to improve the accuracy of rotating inertial navigation system （RINS）. The accuracy of single-axial R1NS is limited by the errors on the rotating axis. A novel inertial measurement unit （IMU） scheme with error compensation for the rotating axis of fiber optic gyros （FOG） RINS is presented. In the scheme, two couples of inertial sensors with similar error characteristics are mounted oppositely on the rotating axes to compensate the sensors error. Without any change for the rotation cycle, this scheme improves the system＇s precision and reliability, and also offers the redundancy for the system. The results of 36 h navigation simulation prove that the accuracy of the system is improved notably compared with normal strapdown INS, besides the heading accuracy is increased by 3 times compared with single-axial RINS, and the position accuracy is improved by 1 order of magnitude.展开更多
文摘为了提升机动目标的航迹预测精准度,提出了一种粒子群(Particle Swarm Optimization,PSO)算法与长短周期记忆单元(Long Short Term Memory,LSTM)神经网络相结合的PSO-LSTM目标航迹预测模型。针对LSTM神经网络中的超参数难以人工最优化的问题,通过PSO算法进行自动调参,对诸如LSTM隐藏层规模、学习率、训练集批次规模等参数进行优化。使用PSO-LSTM航迹预测模型在真实航迹数据集上进行预测,实验结果表明,PSO-LSTM航迹预测模型在MAE、RMSE等指标上优于传统模型,有着较好的准确性与稳定性。
基金supported by the Alexander von Humboldt Foundation,International Climate Protection Fellowship(Ref 3.5—1157991-IRN-IKS)This financial support is gratefully appreciated by Morteza Nikravan.Rafia Firdous and Dietmar Stephan highly acknowledge Bundesministerium fur Wirtschaft und Energie(BMWi)for funding number 16KN046744.
文摘The physical-mechanical,chemical,and durability characteristics of alkali-activated materials(AAMs)have been widely investigated.However,a critical gap in the literature is the lack of a comprehensive overview of recently published literature regarding the life cycle assessment(LCA)of these binders.This study aims to fill that gap by conducting a systematic literature review of globally published literature on the topic.This paper consolidates knowledge by searching different databases,focusing on LCA studies that used AAMs as pastes,mortars,concretes,bricks,and rammed earth/soil blocks.The selected articles were reviewed and categorized based on precursors,alkaline activators,functional units,system boundaries,life cycle inventory databases,allocation,impact methodologies,and software used.Additionally,this paper also critically analyzes the key challenges of LCA for AAMs.The major challenges were identified as selecting a functional unit,subjectivity in boundary systems,and data interpretation.This work concludes that AAMs show substantial advantages in global warming potential compared to ordinary Portland cement-based materials;however,the average of other categories such as marine ecotoxicity and ozone layer depletion has been reported to be higher than for the reference samples.
基金supported by the National Natural Science Foundation of China (No.40904018)the Key Laboratory Foundation of the Ministry of Education of China (No.201001)the Doctoral Innovation Foundation of Naval University of Engineering (No.BSJJ2011008)
文摘An effective and flexible rotation and compensation scheme is designed to improve the accuracy of rotating inertial navigation system （RINS）. The accuracy of single-axial R1NS is limited by the errors on the rotating axis. A novel inertial measurement unit （IMU） scheme with error compensation for the rotating axis of fiber optic gyros （FOG） RINS is presented. In the scheme, two couples of inertial sensors with similar error characteristics are mounted oppositely on the rotating axes to compensate the sensors error. Without any change for the rotation cycle, this scheme improves the system＇s precision and reliability, and also offers the redundancy for the system. The results of 36 h navigation simulation prove that the accuracy of the system is improved notably compared with normal strapdown INS, besides the heading accuracy is increased by 3 times compared with single-axial RINS, and the position accuracy is improved by 1 order of magnitude.
