Search

Chlorophyll prediction facilitates the comprehension of red tide characteristics and enables early warning. In practice, it is formulated as a multivariate time series forecasting problem aimed at forecasting future chlorophyll concentrations by considering both exogenous factors and chlorophyll. However, the multi-step prediction of chlorophyll concentration poses a formidable challenge due to the intricate interaction between factors and the long temporal dependence between input sequences. In this work, we propose a Multi-attention Recurrent Neural Network (MaRNN) for the multi-step prediction of chlorophyll concentration. The MaRNN comprises an encoder incorporating two-stage spatial attention and a decoder employing temporal attention. The encoder first learns the significance of exogenous factors for prediction in the first phase, and subsequently captures the spatial correlation between the exogenous sequence and chlorophyll sequence in the second phase. The decoder further excavates input sequences that exhibit a strong correlation with the task through temporal attention module, thereby enhancing the prediction accuracy of the model. Experiments conducted on two real-world datasets reveal that MaRNN not only surpasses state-of-the-art methods in performance, but also offers interpretability for chlorophyll prediction. [ABSTRACT FROM AUTHOR]

NaNbO3-based ferroelectric composites are regarded as a highly promising typical energy storage material. In this work, 0.02xMnO2−(1−x)NaNbO3−xBi0.5Na0.5TiO3 composites were prepared by the solid-state sintering method and their microstructure, energy storage performance, and temperature stability were analyzed. As observed by the SEM images, the addition of Bi0.5Na0.5TiO3 to NaNbO3 in appropriate amounts can effectively suppress grain growth, which is beneficial for improving the breakdown field strength. When x = 0.3, an optimal value of 210 kV/cm is achieved. Meanwhile, as 2θ = 33.2°, the most intense XRD peak can be observed, indicating excellent solid solubility between the phase components. Consequently, 0.006MnO2–0.7NaNbO3–0.3Bi0.5Na0.5TiO3 composites demonstrated an impressive discharging storage density of 2.464 J/cm3 and a considerable energy storage efficiency of 85.8%. In addition, within the temperature range of 30–80 °C, the fluctuation in energy storage density remains below 10%. All the above performances indicate that this type of material exhibits excellent practicality in the field of pulse power supply. [ABSTRACT FROM AUTHOR]

[Objective] The effects of water and fertilizer coupling on the growth and yield of day lily were studied. [Method] This experiment took day lily as the research object, two factors experiments were set up to study the effects of water and fertilizer coupling on the growth and yield of day lily. [Result] Under the same drip amount, the yield of day lily was significantly different under different drip times, and the agronomic characters of day lily were the best under IV and V treatments, and the maximum yield was 5 983.96-6 000 kg/hm². Under the same fertilization time and different fertilization rates, S-IV treatment and S-V treatment showed the best agronomic characters such as height, number of flower buds and flower buds. [Conclusion] Daylily should be dripped 8-10 times during the whole growth period and fertilized with 525-600 kg/hm². [ABSTRACT FROM AUTHOR]

High strain with low hysteresis is crucial for commercial applications in high precision actuators. However, the clear conflict between the high strain and low hysteresis in BNT-based ceramics has long been an obstacle to actual precise actuating or positioning applications. To obtain piezoceramics with high strain and low hysteresis, it is necessary to enhance the electrostrictive effect and develop an ergodic relaxor (ER) and nonergodic relaxor (NR) phase boundary under ambient conditions. In this work, (Co0.5Nb0.5)4+ doped 76Bi0.5Na0.5TiO3-24SrTiO3 (BNST24) relaxors were fabricated using the conventional solid state reaction route. X-ray diffraction patterns revealed the B-site substitution in BNST24 ceramics. By adjusting the (Co0.5Nb0.5)4+ doping in BNST24, we effectively tuned the TNR-ER and Td close to ambient temperature, which contributed to the development of the ergodic relaxor phase and enhanced the electrostrictive effect at ambient temperature. The I-P-E loops and bipolar strain curves verified the gradual evolution from NR to ER states, while the enhanced electrostrictive effect was verified by the nearly linear S-P2 curves and improved electrostrictive coefficient of the BNST24-xCN relaxors. An enhanced strain of 0.34% (d*33 = 483 pm/V) with low hysteresis of 8.9% was simultaneously achieved in the BNST24-0.02CN relaxors. The enhanced strain was mainly attributed to the proximity effect at the ER and NR phase boundary of BNST24-0.02CN, while the improved electrostrictive effect contributed to the reduced strain hysteresis. Our work demonstrates an effective strategy for balancing the paradox of high strain and low hysteresis in piezoceramics. [ABSTRACT FROM AUTHOR]

Since residual chiral pollutants in the environment and toxic or ineffective chiral components in drugs can threat human health, there is an urgent need for methods to separation and analyze chiral molecules. Molecular imprinting technology (MIT) is a biomimetic technique for specific recognition of analytes with high potential for application in the field of chiral separation and analysis. However, since MIT has some disadvantages when used for chiral recognition, such as poor rigidity of imprinted materials, a single type of recognition site, and poor stereoselectivity, reducing the interference of conformationally and structurally similar substances to increase the efficiency of chiral recognition is difficult. Therefore, improving the rigidity of imprinted materials, increasing the types of imprinted cavity recognition sites, and constructing an imprinted microenvironment for highly selective chiral recognition are necessary for the accurate identification of chiral substances. In this article, the principle of chiral imprinting recognition is introduced, and various strategies that improve the selectivity of chiral imprinting, using derivative functional monomers, supramolecular compounds, chiral assembly materials, and biomolecules, are reviewed in the past 10 years. [ABSTRACT FROM AUTHOR]