High-dimensional optimization presents a novel challenge within the realm of intelligent computing, necessitating innovative approaches. When tackling high-dimensional spaces, traditional evolutionary tools often encounter pitfalls, including dimensional catastrophes and a propensity to become trapped in local optima, ultimately compromising result accuracy. To address this issue, we introduce the Pair Barracuda Swarm Optimization (PBSO) algorithm in this paper. PBSO employs a unique strategy for constructing barracuda pairs, effectively mitigating the challenges posed by high dimensionality. Furthermore, we enhance global search capabilities by incorporating a support barracuda alongside the leading barracuda pair. To assess the algorithm’s performance, we conduct experiments utilizing the CEC2017 standard function and compare PBSO against five state-of-the-art natural-inspired optimizers in the control group. Across 29 test functions, PBSO consistently secures top rankings with 9 first-place, 13 second-place, 5 third-place, 1 fourth-place, and 1 fifth-place finishes, yielding an average rank of 2.0345. These empirical findings affirm that PBSO stands as the superior choice among all test algorithms, offering a dependable solution for high-dimensional optimization challenges.

Dynamic degradation occurs when chaotic systems are implemented on digital devices, which seriously threatens the security of chaos-based pseudorandom sequence generators. The chaotic degradation shows complex periodic behavior, which is often ignored by designers and seldom analyzed in theory. Not knowing the exact period of the output sequence is the key problem that affects the application of chaos-based pseudorandom sequence generators. In this paper, two cubic chaotic maps are combined, which have symmetry and reconfigurable form in the digital circuit. The dynamic behavior of the cubic chaotic map and the corresponding digital cubic chaotic map are analyzed respectively, and the reasons for the complex period and weak randomness of output sequences are studied. On this basis, the digital cubic chaotic map is optimized, and the complex periodic behavior is improved. In addition, a reconfigurable pseudorandom sequence generator based on the digital cubic chaotic map is constructed from the point of saving consumption of logical resources. Through theoretical and numerical analysis, the pseudorandom sequence generator solves the complex period and weak randomness of the cubic chaotic map after digitization and makes the output sequence have better performance and less resource consumption, which lays the foundation for applying it to the field of secure communication.

The brain network is an effective tool and has been widely used in the field of brain neurodegenerative disease analysis. Due to the high cost of accessing medical image data, efforts have been devoted to investigating data augmentation. However, the brain network containing topological characteristics of non-European space is different from the traditional image data, which makes it challenging to synthesize brain structural connectivity and limits the application of brain network analysis. In this paper, a Hemisphere-separated Cross-connectome Aggregating Learning (HCAL) model is proposed to synthesize realistic and diverse brain structural connectivities. Specifically, the latent representation is transformed from structural connectivity by the graph variational autoencoder (GVAE). To generate more diverse and high-quality structural connectivities, the hemisphere-separated generator with a cross-connectome aggregating mechanism is developed to first learn local topological patterns by splitting the whole brain into inter- and intra-hemispheres, then capture global topological characteristics among all the neighbors for each brain region. Also, the connectivity-aware discriminator is devised to make the adversarial training stable and enhance the disease diagnosis. Evaluations of the public Alzheimer’s Disease Neuroimaging Initiative (ADNI) dataset show that the proposed model generates brain structural networks with higher quality and more diversity than related methods. In addition, the classification performance using the augmented data achieved by our approach achieves an accuracy improvement of over 3% compared with the competing method. The proposed model provides an alternative way to synthesize brain networks for connectivity-based disease analysis.

Federated Learning (FL) uses local data to perform distributed training on clients and combines resulting models on a public server to mitigate privacy exposure by avoiding data sharing. However, further research indicates that communication overheads continue to be the primary limitation for FL relative to alternative considerations. This is especially true when training models on non-independent and identically distributed data, such as financial default risk data, where FL’s computational costs increase significantly. This study aims to address financial credit risk data by establishing a dynamic receptive field adjustment mechanism for feature extraction, efficiently extracting default features with varying distributions and attributes. Additionally, by constructing a distributed feature fusion architecture, characteristics from both local and overarching models are aggregated to attain higher accuracy with lower transmission costs. Experimental results demonstrate that the proposed FL framework can utilize dynamic receptive fields to allocate convolutional kernel weights, thereby improving feature extraction. In the feature fusion stage, the proposed Multi-Fusion strategy efficiently customizes the aggregation of features from local and global models. The final solution reduces the communication rounds in federated learning by approximately 80%.

The offspring selection strategy is the core of evolutionary algorithms, which directly affects the method’s accuracy. Normally, to improve the search accuracy in local areas, the population converges quickly around the optimal individual. However, excessive aggregation can narrow the search range of the population, and thus the population may be trapped by local optima. To overcome this problem, a bare-bones particle swarm optimization with crossed memory (BPSO-CM) is proposed in this work. The BPSO-CM contains a multi-memory storage mechanism (MSM) and an elite offspring selection strategy (EOSS). The MSM enables an extra storage space to extend the search ability of the particle swarm and the EOSS enhances the local minimum escaping ability of the particle swarm. The population is endowed with the ability of enhanced global search through the cooperation of the MSM and the EOSS. To verify the performance of the BPSO-CM, the CEC2017 benchmark functions are used in experiments, five population-based methods are selected in the control group. Finally, experimental results proved that the BPSO-CM can present highly accurate results for global optimization problems.

Using point cloud to reconstruct the 3D model of a substation is crucial for smart grid operation. Its main objective is to swiftly capture equipment point cloud data and align each device’s model within the large and noisy point cloud scene of the substation. However, substation reconstruction needs improvement due to the low efficiency of traditional noise-resistant clustering methods and challenges in accurately classifying similar-looking electrical equipment. This paper proposes an automatic modeling framework for large-scale substation point cloud scenes. Firstly, we reduce the substation scene’s dimensionality to improve clustering efficiency and establish relationships between data dimensions using a re-clustering algorithm. Next, a neural network is developed to identify various device types within clusters, even with limited subdivisions. Finally, a model library is employed to register standard models onto the target device’s point cloud, obtaining device types and orientations. Real substation data processing demonstrates the ability to rapidly extract devices from complex and noisy point cloud scenes, effectively avoiding missegmentation issues. The automatic modeling approach achieves a precise substation calculation rate of 92.86%.