A machine learning-based strategy for predicting the mechanical strength of coral reef limestone using X-ray computed tomography

Different sedimentary zones in coral reefs lead to significant anisotropy in the pore structure of coral reef limestone(CRL),making it difficult to study mechanical behaviors.With X-ray computed tomography(CT),112 CRL samples were utilized for training the support vector machine(SVM)-,random forest(RF)-,and back propagation neural network(BPNN)-based models,respectively.Simultaneously,the machine learning model was embedded into genetic algorithm(GA)for parameter optimization to effectively predict uniaxial compressive strength(UCS)of CRL.Results indicate that the BPNN model with five hidden layers presents the best training effect in the data set of CRL.The SVM-based model shows a tendency to overfitting in the training set and poor generalization ability in the testing set.The RF-based model is suitable for training CRL samples with large data.Analysis of Pearson correlation coefficient matrix and the percentage increment method of performance metrics shows that the dry density,pore structure,and porosity of CRL are strongly correlated to UCS.However,the P-wave velocity is almost uncorrelated to the UCS,which is significantly distinct from the law for homogenous geomaterials.In addition,the pore tensor proposed in this paper can effectively reflect the pore structure of coral framework limestone(CFL)and coral boulder limestone(CBL),realizing the quantitative characterization of the heterogeneity and anisotropy of pore.The pore tensor provides a feasible idea to establish the relationship between pore structure and mechanical behavior of CRL.

A Combined Method of Temporal Convolutional Mechanism and Wavelet Decomposition for State Estimation of Photovoltaic Power Plants

Time series prediction has always been an important problem in the field of machine learning.Among them,power load forecasting plays a crucial role in identifying the behavior of photovoltaic power plants and regulating their control strategies.Traditional power load forecasting often has poor feature extraction performance for long time series.In this paper,a new deep learning framework Residual Stacked Temporal Long Short-Term Memory(RST-LSTM)is proposed,which combines wavelet decomposition and time convolutional memory network to solve the problem of feature extraction for long sequences.The network framework of RST-LSTM consists of two parts:one is a stacked time convolutional memory unit module for global and local feature extraction,and the other is a residual combination optimization module to reduce model redundancy.Finally,this paper demonstrates through various experimental indicators that RST-LSTM achieves significant performance improvements in both overall and local prediction accuracy compared to some state-of-the-art baseline methods.

Sustainable Seawater Desalination and Energy Management:Mechanisms,Strategies,and the Way Forward

Solar-driven desalination systems have been recognized as an effective technology to address the water crisis.Recently,evaporators prepared based on advanced manufacturing technologies have emerged as a promising tool in enhancing ocean energy utilization.In this review,we discussed the thermal conversion,energy flow,salt deposition mechanisms,and design strategies for solar-driven desalination systems,and explored how to improve the desalination performance and energy use efficiency of the systems through advanced manufacturing technologies.In future perspectives,we determined the feasibility of coupling solar-driven solar desalination systems with multi-stage energy utilization systems and emerging artificial intelligence technologies,for which conclusions are given and new directions for future desalination system development are envisioned.Finally,exciting opportunities and challenges in the face of basic research and practical implementation are discussed,providing promising solutions and blueprints for green and novel desalination technologies while achieving sustainable development.

Revealing the Distribution of Aggregation-Induced Emission Nanoparticles via Dual-Modality Imaging with Fluorescence and Mass Spectrometry

Aggregation-induced emission nanoparticles(AIE NPs)are widely used in the biomedical field.However,understanding the biological process of AIE NPs via fluorescence imaging is challenging because of the strong background and poor penetration depth.Herein,we present a novel dual-modality imaging strategy that combines fluorescence imaging and label-free laser desorption/ionization mass spectrometry imaging(LDI MSI)to map and quantify the biodistribution of AIE NPs(TPAFN-F127 NPs)by monitoring the intrinsic photoluminescence and mass spectrometry signal of the AIE molecule.We discovered that TPAFN-F127 NPs were predominantly distributed in the liver and spleen,and most gradually excreted from the body after 5 days.The accumulation and retention of TPAFN-F127 NPs in tumor sites were also confirmed in a tumor-bearing mouse model.As a proof of concept,the suborgan distribution of TPAFN-F127 NPs in the spleen was visualized by LDI MSI,and the results revealed that TPAFN-F127 NPs were mainly distributed in the red pulp of the spleen with extremely high concentrations within the marginal zone.The in vivo toxicity test demonstrated that TPAFN-F127 NPs are nontoxic for a long-term exposure.This dual-modality imaging strategy provides some insights into the fine distribution of AIE NPs and might also be extended to other polymeric NPs to evaluate their distribution and drug release behaviors in vivo.

“Living”fluorophores:thermo-driven reversible ACQ-AIE transformation and ultra-sensitive in-situ monitor for dynamic Diels-Alder reactions

Organic fluorophores play essential roles in both academic and applied fields.Most of the fluorescent molecules can be divided into aggregation-caused quenching(ACQ)and aggregation-induced emission(AIE)types based on the diverse emission properties in solution and aggregated states.Currently,a large part of studies focuses on the ACQ-to-AIE one-way transformation and the complex synthesis of chemical bonds is inevitable in all existing methods.To maximize the advantages of ACQ and AIE types fluorophores and avoid complex chemosynthesis,we propose a facile strategy first realizing the reversible ACQAIE transformation with the dynamic Diels-Alder(DA)reactions.Besides,the fluorescent platform can monitor DA reactions in microscale ultra-sensitively and quantitively.The dynamic covalent bonds can help to develop novel fluorophores creatively,and the reversible ACQ-AIE platform is expected to offer fresh insights into the dynamic covalent chemistry.