The widespread application of Generative Artificial Intelligence (Gen-AI) is transforming educational practices and driving pedagogical innovation. While cultivating higher-order thinking (HOT) represents a central educational goal, its achievement remains an ongoing challenge. Current evidence regarding the impact of Gen-AI on HOT is relatively fragmented, lacking systematic integration, particularly in the analysis of moderating variables. To address this gap, a meta-analysis approach was employed, integrating data from 29 experimental and quasi-experimental studies to quantitatively assess the overall impact of Gen-AI on learners’ HOT and to examine potential moderating factors. The analysis revealed that Gen-AI exerts a moderate positive effect on HOT, with the most significant improvement observed in problem-solving abilities, followed by critical thinking, while its effect on creativity is relatively limited. Moderation analyses further indicated that the impact of Gen-AI is significantly influenced by experimental duration and learners’ self-regulated learning (SRL) abilities: effects were strongest when interventions lasted 8–16 weeks, and learners with higher SRL capacities benefited more substantially. Based on the research findings, this study proposed that Gen-AI should be systematically integrated as a targeted instructional tool to foster HOT. Medium- to long-term interventions (8–16 weeks) are recommended to enhance learners’ problem-solving and critical thinking abilities. At the same time, effective approaches should also be explored to promote creative thinking through Gen-AI within existing pedagogical frameworks. Furthermore, individual learner differences should be accounted for by adopting dynamic and personalized scaffolding strategies to foster SRL, thereby maximizing the educational potential of Gen-AI in cultivating innovative talents.

High-frequency (HF) surface-wave radar has a wide range of applications in marine monitoring due to its long-distance, wide-area, and all-weather detection ability. However, the accurate detection of HF radar vessels is severely restricted by strong clutter and interference, causing the echo of vessels completely submerged by clutter. As a result, the target cannot be detected and tracked for a period of time under the influence of strong clutter, which causes broken trajectories. To solve this problem, we propose an HF radar-vessel trajectory-prediction method based on a multi-scale convolutional neural network (MSCNN) that combines a gated recurrent unit and attention mechanism (GRU-AM) and a fusion with an autoregressive (AR) model. The vessel’s latitude and longitude information obtained by the HF radar is sent into the convolutional neural network (CNN) with different window lengths in parallel, and feature fusion is performed on the extracted multi-scale features. The deep GRU model is built to learn the time series with the GRU structure to preserve historical information. Different weights are given to the features using the temporal attention mechanism (AM), which helps the network learn the key information. The linear information on latitude and longitude at the current timestep is forecast by combining the AR model with the trajectory output from the AM to achieve a combination of linear and nonlinear prediction models. To make full use of the HF radar tracking information, the broken trajectory prediction is carried out by forward and backward computation using data from before and after the fracture, respectively. Weights are then assigned to the two predicted results by the entropy-value method to obtain the final ship trajectory by weighted summation. Field experiments show that the proposed method can accurately forecast the trajectories of vessels concealed in clutter. In comparison with other mainstream methods, the new method performs better in estimation accuracy for HF radar vessels concealed in clutter.

High-frequency surface wave radar (HFSWR) is an important marine monitoring technology, and this new regime of radar plays an important role in large-scale, continuous early-warning monitoring at sea. In particular, shipborne HFSWR has wider applications in detecting interesting sea areas, with the advantages of flexible deployment and extended detection capability. Due to the large amount of sea clutter accompanying the echo signals of shipborne HFSWR and the spread of sea clutter due to platform motion, the detection of targets in clutter regions is extremely difficult. In this paper, a multi-frame time-frequency (TF) analysis–based target-detection method is proposed. First, the sea clutter spreading area in the HFSWR echo signal is modeled, and the effects of platform motion and currents on the sea clutter spread are analyzed to determine the sea clutter coverage area; this paper focuses on frequency modeling. Then the TF image (TFI) of each range cell is obtained by TF analysis of the cells within a certain range of the echo signal, and the range cells of possible target points are determined by binary classification of the TFI through a convolutional neural network. Finally, the location of the final target point is obtained by correlation of multi-frame TFIs. Shipborne HFSWR field experiments show that the proposed detection method performs well in detecting targets concealed by sea clutter.

Shipborne high-frequency surface wave radar (HFSWR) has a wide range of applications and plays an important role in moving target detection and tracking. However, the complexity of the sea detection environment causes the target signals received by shipborne HFSWR to be seriously disturbed by sea clutter. Sea clutter increases the difficulty of azimuth estimation, resulting in a challenging problem for shipborne HFSWR. To solve this problem, a novel azimuth correction method based on adaptive boosting error feedback dynamic weighted particle swarm optimization extreme learning machine (APO-ELM) is proposed to improve the azimuth estimation accuracy of shipborne HFSWR. First, the sea clutter is modeled and simulated. Then, we study its characteristics and analyze the influence of its characteristics on the first-order clutter spectrum and target detection accuracy, respectively. In addition, the proposed improved particle swarm optimization (PSO) and adaptive neuron clipping algorithm are used to optimize the input parameters of the ELM network. Then, the network performs error feedback based on the optimized parameter performance and updates the feature matrix, which can give a minimum clutter-error estimation. After that, it iteratively trains multiple weak learners using the adaptive boosting (AdaBoost) algorithm to form a strong learner and make strong predictions. Finally, after error compensation, the best azimuth estimation results are obtained. The sample sets used for the APO-ELM network are obtained from field shipborne HFSWR data. The network training and testing features include the wind direction, sea current, wind speed, platform speed, and signal-to-clutter ratio (SCR). The experimental results show that this method has a lower root-mean-square error than the back-propagation neural network and support vector regression (SVR) azimuth correction methods, which verifies the effectiveness of the proposed method.

Maritime surveillance heavily relies on high-frequency surface wave radar (HFSWR) systems. However, clutter and interference make it difficult to accurately detect vessel targets using a single-frame detection method. This study introduces an improved time-frequency analysis (TFA) algorithm to enhance the features in single-frame detection. In this article, TFA, multiframe correlation, and deep neural networks are integrated to develop a three-stage detection framework. Firstly, faster R-CNN is customized for the preprocessing stage to identify sea clutter regions. Then, based on the range-Doppler (RD) spectrum, suspicious targets are swiftly identified amidst clutter in the initial stage. Subsequently, the improved TFA algorithm is applied to adjacent range cells of suspicious targets to generate multiframe TF images, forming a three-dimensional data block structured as time-range-Doppler frequency. To reduce computational complexity, a TFA method using multisynchrosqueezing transform is employed, enhancing detection accuracy for targets within cluttered regions. In the final stage, a 3DResnet model is utilized to leverage the differences in features between clutter and targets across three dimensions. This allows for distinguishing genuine targets from false targets using time series information from multiple frames. Comparative analysis against classical target detection algorithms demonstrates the superior detection performance of the proposed framework within clutter regions. This showcases its potential for enhancing the maritime surveillance capabilities of HFSWR.