ISSN 0253-2778

CN 34-1054/N

2023 Vol. 53, No. 4

2023-4 Contents
2023, 53(4): 1-2.
2023-4 Abstract
2023, 53(4): 1-2.
Information Science and Technology
Coded computing for distributed graph-based semi-supervised learning
Siqi Tan, Li Chen, Weidong Wang
2023, 53(4): 0401. doi: 10.52396/JUSTC-2022-0133
Semi-supervised learning (SSL) has been applied to many practical applications over the past few years. Recently, distributed graph-based semi-supervised learning (DGSSL) has been shown to have good performance. Current DGSSL algorithms usually have the problems of inefficient graph construction and the straggler effect. This paper proposes a novel coded DGSSL (CDGSSL) to solve these problems. We first provide a novel parallel and distributed solution of matrix completion for efficient graph construction. Then, we develop the CDGSSL algorithm based on coding theory. Specifically, the proposed algorithm consists of two parts separately designed based on the maximum distance separable (MDS) code. In general, the proposed coded distributed algorithm is efficient and straggler tolerant. Moreover, we provide an optimal parameter design for the proposed algorithm. The results of the experiments on the Alibaba Cloud elastic compute service (ECS) demonstrate the superiority of the proposed algorithm.
Low-complexity energy-aware sensor selection for noise reduction in distributed microphone networks
Jie Zhang, Lu-Zhen Xu, Li-Rong Dai
2023, 53(4): 0402. doi: 10.52396/JUSTC-2022-0121
Noise reduction (NR) is a necessary front-end in many audio applications for improving signal quality. It was shown that sparsity-promoting sensor selection potentially makes a trade-off between energy consumption and NR performance, which is rather important for large-scale wireless acoustic sensor networks (WASNs), where many sensors contribute negligibly to NR but energy consumption affects the lifetime of WASNs. This paper presents a sensor selection approach for beamforming-based NR by minimizing the total energy consumption and constraining the output noise variance. Motivated by the optimal semi-definite programming (SDP) solution and the utility-based method, we propose three low-complexity selection metrics: weighted utility, gradient, and weighted input signal-to-noise ratio (SNR). It is shown that the proposed weighted utility and gradient-based methods are near-optimal in performance but much faster than the SDP-based method, and the weighted SNR method has the lowest time complexity with a tiny performance sacrifice. Numerical results using a simulated WASN validate the superiority of the proposed approaches over conventional methods.
Information Science and Technology / Management
Confidence intervals for high-dimensional multi-task regression
Yuanli Ma, Yang Li, Jianjun Xu
2023, 53(4): 0403. doi: 10.52396/JUSTC-2022-0115

Regression problems among multiple responses and predictors have been widely employed in many applications, such as biomedical sciences and economics. In this paper, we focus on statistical inference for the unknown coefficient matrix in high-dimensional multi-task learning problems. The new statistic is constructed in a row-wise manner based on a two-step projection technique, which improves the inference efficiency by removing the impacts of important signals. Based on the established asymptotic normality for the proposed two-step projection estimator (TPE), we generate corresponding confidence intervals for all components of the unknown coefficient matrix. The performance of the proposed method is presented through simulation studies and a real data analysis.

Information Science and Intelligence Technology
Learning attention-based strategies to cooperate for multi-agent path finding
Jinchao Ma, Defu Lian
2023, 53(4): 0404. doi: 10.52396/JUSTC-2022-0048
Multi-agent path finding (MAPF) is a challenging multi-agent systems problem where all agents are required to effectively reach their goals concurrently with not colliding with each other and avoiding obstacles. In MAPF, it is a challenge to effectively express the observation of agents, utilize historical information, and effectively communicate with neighbor agents. To tackle these issues, in this work, we proposed a well-designed model that utilizes the local states of nearby agents and outputs an optimal action for each agent to execute. We build the local observation encoder by using residual attention CNN to extract local observations and use the Transformer architecture to build an interaction layer to combine local observations of agents. With the purpose of overcoming the deficiency of success rate, we also designed a new evaluation index, namely extra time rate (ETR). The experimental results show that our model is superior to most previous models in terms of success rate and ETR. In addition, we also completed the ablation study on the model, and the effectiveness of each component of the model was proved.
Engineering & Materials
Advanced functional safeguarding composites with enhanced anti-impact and excellent thermal properties
Wenhui Wang, Sheng Wang, Shuai Liu, Jianyu Zhou, Junshuo Zhang, Fang Yuan, Min Sang, Xinglong Gong
2023, 53(4): 0405. doi: 10.52396/JUSTC-2022-0089
Personal safety protection has played an important role in daily life. Developing advanced functional safeguarding composites with enhanced anti-impact and excellent thermal properties will be a significant development for body armor. Herein, Kevlar fiber reinforced polymers (KFRP) were fabricated by introducing short Kevlar fibers (KFs) into a shear stiffening elastomer (SSE). The storage modulus of KFRP with 15 wt% KFs (KFRP-15%) increased from 222.8 kPa to 830.8 kPa when the shear frequency varied from 0.1 Hz to 100 Hz. KFRP-15% achieved a higher tensile strength (2.65 MPa) and fracture toughness (11.95 kJ/m2) than SSE in the vertical type, showing superior tear resistance. Additionally, KFRP-15% exhibited promising anti-impact properties, which could dissipate the drop hammer impact force from 1.74 kN to 0.56 kN and remained intact after 10 consecutive impacts. Moreover, KFRP-15% also presented excellent stab-resistant performance. In addition, KFRP-15% also showed improved heat transfer properties, flame retardancy, and smoke suppression capabilities. Finally, functional bracers based on KFRP-15% for protection, thermal-dissipation, and flame-retardant were successfully prepared.
Multiobjective optimization of morphologies and performance of Q355C gas metal arc welding based on the NSGA-Ⅱ
Huajing Weng, Meiyan Feng, Jibin Jiang, Changrong Chen, Guofu Lian
2023, 53(4): 0406. doi: 10.52396/JUSTC-2022-0112
This work studied the influence law of gas-metal-arc welding process parameters on the morphologies and performance to improve the morphologies and performance. The mixed orthogonal surfacing test was carried out by taking the preheating temperature, welding voltage, current, speed, and wire extension as GMAW process parameters. The aspect ratio decreased with increasing welding voltage, and it first increased and then decreased with increasing welding current. The hardness increased with increasing preheating temperature and welding speed and decreased with increasing welding voltage, current, and wire extension. Residual stress increased with the increased preheating temperature. In addition, it first decreased and then increased with increasing welding voltage and speed. Based on the regression model, the nondominated sorting genetic algorithm II (NSGA-II) was used for multiobjective optimization. After that, experiments were conducted to verify the noninferior solutions among the aspect ratio, hardness, and residual stress. Errors between the predicted and experimental results by the three output indices were all less than 10%, indicating the feasibility of the optimization method. The research results provide a theoretical direction for multiobjective optimization and refined applications of arc welding.