Search

Earthquake-triggered landslides (ETLs) are characterized by their extensive occurrences, having wide distributions. The conventional human–computer interaction extraction method is often time-consuming and labor-intensive, failing to meet the demands of disaster emergency response. There is a pressing need for a swift detection of ETLs. In this study, we introduce an ETLs extraction method (MAST) combining morphological analysis edge recognition with a Swin-Transformer (SWT) deep learning model, which is specifically designed for landslide extraction. The MAST model adopts a hierarchical construction approach akin to convolution neural networks, aiding in tasks such as target detection and semantic segmentation. To enhance the accuracy of landslide edge extraction, we incorporate an edge recognition algorithm based on the morphological analysis into the MAST model. This algorithm leverages morphological operations to extract the features of landslide boundaries. It effectively addresses issues such as discretization and irregularization of the extracted landslide boundaries, leading to more precise delineation of landslide boundaries. Drawing on UAV data collected from Wan Dong Village, De Tou Town, Sichuan Luding, China, during the 2022 Ms 6.8 Luding Earthquake, we conducted automated extraction of ETLs utilizing the MAST model. Experimental results demonstrate the superior performance of the MAST model compared to the traditional full convolution neural network (FCN) model and normal SWT model. The MAST model exhibits enhanced value in landslide extraction. Notably, it demonstrates a significant advantage in boundary extraction. Employing the Boundary IoU metric to evaluate the accuracy of ETLs extraction, the MAST model outperforms the SWT and FCN models at various distances.

During tungsten inert gas welding process, arc length will change due to thermal deformation, uneven surface of workpiece and tungsten electrode loss. Considering the large fluctuation of arc length in all-position annular seam welding under the action of multi-variable coupling, a mathematical model of arc length and arc voltage was established in this paper, and an isolated and high frequency arc voltage acquisition circuit was designed. An improved sliding filtering algorithm was proposed, and after filtering the real-time arc voltage data, fuzzy control algorithm was used to realize the online tracking control of arc length in annular seam welding, in order to improve the reliability and stability of welding quality of tube-sheet. The experimental results show that the dynamic response time of the proposed arc length tracking compensation method is not more than 40 ms, and the arc length tracking deviation is less than 0.2 mm, which can fully meet the requerements of welding seam tracking of tube-sheet welding.

This paper studies the factors affecting injury severities involving motorcyclists of different age groups (under 25, 25-55, 55 and above) based on the random parameter logit model. Data collected from motorcycle crashes in the UK between 2017 and 2020 are utilized. The motorcyclist injury severity outcomes are categorized as follows: minor injury, severe injury, and fatal injury. The results of the likelihood ratio tests showed that transferability in multi-vehicle motorcycle crash injury severity involving motorcyclists of different age groups. The results of the modeling revealed significant variations in the factors that impact motorcycle crashes among three different age groups, including rider characteristics (such as male riders), motorcycle and non-motorcycle characteristics (such as vehicle running status preceding collision, age of the vehicle, and non-motorcycle vehicle type), roadway and environmental conditions (such as weather condition, speed limit, and road type), temporal-related characteristics (such as day of the week), and crash-related characteristics (such as crash types). The models demonstrate the existence of unobserved heterogeneity for three statistically significant variables, including the truck-involved, rear-end collision, and morning off-peak hours indicator in the under 25 age group crash model, and vehicle straight movement in the 25–55 age group crash model, and passenger car and sideswipe in the 55 and above age group crash model. Further, there are substantial differences in injury severity probabilities involving motorcyclists of different age groups by comparing prediction results based on out-of-sample prediction simulation. This paper emphasized the importance of revealing different age groups’ crash transferability and heterogeneity. The statistically significant differences involving age group crash injury severity models highlight the importance of age-targeted policies for motorcycle safety.

When metro systems need to be temporarily closed due to an unexpected disruption, the metro operators have to arrange shuttle buses as alternative transport for trains, while passengers have to find new routes to complete their journeys. However, the additional shuttle bus lines might be redundant if the well-known phenomenon of the “Braesss Paradox” exists. To address this issue, this paper proposes a bi-level programming model in which the objective of the upper-level model is to minimize the number of shuttle bus lines, while the lower-level model is a UE (User Equilibrium) model that simulates the route choice of passengers. To solve this bi-level programming model, the Overload Arc-oriented Heuristics (OAH) algorithm is developed to find an initial solution, and then we develop a BS (Bi-direction Search) algorithm to obtain the optimal solution to our model. Numerical cases of the partial Beijing Metro network have proved the efficiency and effectiveness of our method, and the result shows that: (1) compared with the standard bus bridging routes, the additional shuttle bus lines can significantly reduce the number of stranded passengers; (2) the number of stranded passengers at each station can be controlled by two strategies: adjust the departure frequency of each shuttle bus line and control the passenger flow at each station.

The fast and accurate locating of the tube-sheet annular seam is the key factors to ensure the high efficiency and quality of automatic welding of pressure vessels. Considering the problems of welding sheet deformation and tubes deviation caused by machining process and welding thermal, this paper builds an automatic welding robot platform, which is equipped with a measurement system coordinated by multiple lasers and visual sensors, and put forward a real-time welding distance measurement method based on cross lasers and visual sensors, at the same time, the multi spot laser assisted vision sensor is used to quickly and accurately measure the center of the tube, so as to guide the adjustment of the welding robot and posture of the welding torch in real time. Then, through the measurements of system error, static and dynamic measurement error, it is known that the dynamic accuracy of tube sheet welding distance is not less than 0.22mm, and the positioning accuracy of welded tube center is not less than 0.219mm, which can meet the accuracy requirements of automatic tube sheet welding.

Spatio-temporal forecasting is of great importance in a wide range of dynamic systems applications, such as earth science, transport planning, etc. These applications rely on accurate predictions of spatio-temporal structured data reflecting real-world phenomena. A stunning characteristic is that the dynamical system is not only driven by some physics laws but also impacted by the localized factor in spatial and temporal regions. One of the major challenges is to infer the underlying causes, which generate the perceived data stream and propagate the involved causal dynamics through the distributed observing units. Another challenge is that the success of machine learning-based predictive models requires massive annotated data for model training. However, the acquisition of high-quality annotated data is objectively manual and tedious as it needs a considerable amount of human intervention, making it infeasible in fields that require high levels of expertise. To tackle these challenges, we advocate a spatio-temporal physics-coupled neural networks (ST-PCNN) model to learn the underlying physics of the dynamical system and further couple the learned physics to assist the learning of the recurring dynamics. To deal with data-acquisition constraints, an active learning mechanism with Kriging for actively acquiring the most informative data is proposed for ST-PCNN training in a partially observable environment. Our experiments on both synthetic and real-world datasets exhibit that the proposed ST-PCNN with active learning converges to near-optimal accuracy with substantially fewer instances.

An interleaved zero-voltage zero-current switching (ZVZCS) high step-up DC-DC converter is proposed for modern photovoltaic power generation systems. The proposed converter can achieve high voltage conversion gain without operating under extreme duty cycles. The voltage conversion gain can be readily adjusted by selecting different numbers of the diode-capacitor multiplier (DCM) cells in the circuit at the design stage. Moreover, with the passive snubber circuit in the proposed converter, all power switches and diodes in the DCMs can achieve zero-voltage turn-off and zero-current turn-on, which effectively reduces switching losses and increases the overall converter efficiency. The operational principles and analysis of the performance analysis with two DCMs are presented for demonstrative purpose. The effectiveness of the proposed ZVZCS DC-DC converter is validated using an 800W experimental prototype.

The contradiction between transport capacity and passenger demand in urban rail transit is usually prominent during peak hours in some megacities of China, and some passenger flow control measures have been adopted to alleviate passenger congestion. To better save passengers’ travel time when taking passenger flow control measures, this paper proposes an integrated optimization method of bus route adjustment with network-level passenger flow control for urban rail transit, in which the controlled passengers can freely choose to shift to bus or to retain in urban rail transit for pursuing a lower travel cost. With the objectives of minimizing average additional travel time for all affected passengers and maximizing the operating revenue of urban rail transit, an integer non-linear programming model is formulated to determine the inbound passenger volumes and bus adjustment schemes. To solve this proposed model effectively, a multi-objective particle swarm optimization based on dual-population co-evolution is designed. Finally, three sets of numerical experiments, including an integrated optimization experiment and two independent optimization experiments of passenger flow control, are implemented to demonstrate the feasibility and benefits of the proposed method.

We report on a simple approach of time-delayed reservoir computing (RC) based on a two-element phased laser array for image identification. Here the phased laser array with optical feedback and injection is trained according to the representative characteristics extracted through histograms of oriented gradients. These characteristic vectors are multiplied by a random mask signal to form input data, which are subsequently trained in the reservoir. By optimizing the parameters of the RC, we achieve an identification accuracy of 97.44% on the MNIST dataset and 85.46% on the Fashion-MNIST dataset. These results indicate that our proposed RC indeed allows accurate classification of handwritten digit and fashion production. Moreover, we further forecast an RC scheme based on a larger-scale phased laser array, which is expected to tackle more complex tasks at a high speed. Our work offers a possibility for advanced image processing using highly integrated neuromorphic photonic systems.

By the negative influence of parasitic parameters from the components, the voltage step-up capacity of basic dc-dc converters is limited at extremely high duty cycle. Which makes the traditional dc-dc converters are difficult to be applied to renewable energy generation systems such as fuel cell and photovoltaic (PV) systems. In this study, a single-switch high step-up Zeta converter is proposed based on the basic Zeta converter and corresponding coat circuit. By using the coat circuit, the proposed converter not only can achieve higher voltage gain, but also acquire lower voltage stress on the semiconductor devices. Therefore, the devices with lower conduction resistances can be used to improve the efficiency. Furthermore, the coat circuit is composed of passive components, which does not change the continuous output current characteristic of the basic Zeta converter. And the proposed converter has only one active switch, the drive and control circuits are simple. The operation principle and the performance analysis of the proposed Zeta converter are described in detail, and a 300 W closed-loop experimental prototype has been developed for verification.

There are many uncertain factors, which are critical for the efficient operation of steam power systems (SPSs). Due to the modelling inaccuracy and the changing environment, it is challenging to directly formulate the model for SPS operation. In this paper, a method is proposed to alleviate the influence of small-scale fluctuation by discretizing the uncertainty working condition on the SPS. Moreover, the exergy analysis method is proposed to evaluate the quality of different energy and develop a mixed-integer nonlinear model of SPS considering the economic and exergy objectives. Finally, the $\varepsilon $ -constraint method is used to obtain the Pareto front of the multi-objective optimization problem, and the fuzzy decision is used to determine the optimal strategy of the system. By introducing the multi-objective optimization of the SPS model with demand side uncertainty, a flexible scheduling scheme with good economic and exergy objectives is obtained. Comparing with the optimal scheduling without the demand-side uncertainty, numerical results indicate the feasibility and effectiveness of the model.

Fully dynamic multi-key fully homomorphic encryption (FHE) that allows an unlimited number of homomorphic operations for unconstrained parties. That is to say, it supports performing as many computational procedures on inputs (which are encrypted by an unrestricted number of parties) as needed. The existed fully dynamic multi-key FHE scheme is based on the learning with errors (LWE) problem. However, the LWE problem suffers from the intricate and laborious Gaussian sampling which extremely weakens the schemes’ efficiencies. What’s more, it has been lately revealed that Gaussian sampling creates lots of possible side-channel vulnerabilities that result in the disclosure of secret keys. The paper proposed a fully dynamic multi-key FHE that based on the learning with rounding (LWR) problem which leaves out the time-consuming Gaussian sampling required in the LWE problem while sustaining almost the same security level.

With the increasing needs of and rapid developments in digital and/or smart cities, an effective method for the management of various types of complex and massive buildings for the creation of a three-dimensional (3D) photorealistic city has become crucial for high-quality “digital/smart city” construction. To this end, this paper proposes a data model called the “SCSG-OSM”, which combines spatial constructive solid geometry (SCSG) and the object-based spatial model (OSM). The SCSG-OSM assumes an object can be represented by four element types: point, line, face and body. The SCSG applies an extension of the dimensionally extended nine-intersection model (DE-9IM) to describe the spatial topological relationship. The OSM consists of nine definitions that define the points, lines, faces, and bodies of an object and applies point sets instead of node sets and face sets instead of surface sets to model buildings. Two data sets, depicting Denver, Colorado, USA, and Zurich, Switzerland, are employed to assess the storage space and the time consumption averages during the modelling of 3D buildings using the proposed data model. The experimental results demonstrate that the proposed SCSG-OSM is able to save storage space and reduce the computational time compared with the CSG-BR model, TIN model, POLYGON model, Patch model, SSM and CityGML model. Therefore, it can be concluded that the method proposed in this paper is a simple and effective method for the 3D photorealistic visualization of a large city.

The conventional buck-boost converter has the advantages of simple structure, low cost, and the capability to achieve both voltage step-up and down. However, due to the negative impacts of the parasitic parameters of the device, the voltage conversion gain of the conventional buck-boost converter is greatly limited. A low-voltage stress buck-boost converter with a high voltage conversion gain based on a coat circuit is proposed in this paper to address the problem. Similar to a coat that can enhance human's resistance to cold weather, by adding the proposed coat circuit to the conventional buck-boost converter, not only the range for the voltage conversion can be extended, but also the voltage stresses of the semiconductor components are effectively reduced. In this work, comprehensive analysis on the working principles and performance characteristics of the coat converter are provided. Experimental results are obtained and analyzed to validate of the theoretical analysis based on a 300W closed-loop prototype platform.

With excellent mobility and flexibility, mobile manipulators have great potential for loading and unloading tasks of numerical control machine tools (CNC) in manufacturing workshops. However, because of the rough and oily ground, dynamic obstacles and the convex plate of a CNC, harsh manufacturing workshop poses a huge challenge to the localization system of an autonomous mobile manipulator. To address the above problem, this paper presents a hybrid maps enhanced localization system which mainly consists of a global localization method and a pose tracking method. Hybrid maps including hybrid grid map, multi-resolution likelihood fields (MLFs) and hybrid point map are constructed to efficaciously model the harsh environment and to improve localization performance. Our global localization method employs the convex hull sampling to spares dense Lidar data and the MLFs based branch and bound (BnB) search to speed up global search. To achieve real-time localization reliably and accurately, our pose tracking method seamlessly combines the BnB search and the adaptive Monte Carlo localization, and the Iterative Closest Point (ICP) based scan matching using the hybrid point map is adopted for higher accuracy. In addition, a distance filter improved by unscented transform is integrated into the pose tracking process to mitigate the influence of dynamic obstacles. The developed localization system is evaluated through different experiments including two weeks of loading and unloading tasks in a real manufacturing scenario, resulting in superior localization performance.

Facial beauty plays an important role in many fields today, such as digital entertainment, facial beautification surgery and etc. However, the facial beauty prediction task has the challenges of insufficient training datasets, low performance of traditional methods, and rarely takes advantage of the feature learning of Convolutional Neural Networks. In this paper, a transfer learning based CNN method that integrates multiple channel features is utilized for Asian female facial beauty prediction tasks. Firstly, a Large-Scale Asian Female Beauty Dataset (LSAFBD) with a more reasonable distribution has been established. Secondly, in order to improve CNN's self-learning ability of facial beauty prediction task, an effective CNN using a novel Softmax-MSE loss function and a double activation layer has been proposed. Then, a data augmentation method and transfer learning strategy were also utilized to mitigate the impact of insufficient data on proposed CNN performance. Finally, a multi-channel feature fusion method was explored to further optimize the proposed CNN model. Experimental results show that the proposed method is superior to traditional learning method combating the Asian female FBP task. Compared with other state-of-the-art CNN models, the proposed CNN model can improve the rank-1 recognition rate from 60.40% to 64.85%, and the pearson correlation coefficient from 0.8594 to 0.8829 on the LSAFBD and obtained 0.9200 regression prediction results on the SCUT dataset.

In this paper, a detailed model of motion stage driven by iron-less permanent magnet linear synchronous motor (PMLSM) and supported by rolling guides are presented. This stage is a prototype of the feeding system for a precise turning machine tool. In this model, there are two categories disturbances including slow-varying ones mainly produced by the thrust ripple of PMLSM and abrupt-changing ones caused by nonlinear friction of rolling guide. To overcome these disturbances, a simple, effective and practical servo control method based on the general-purpose two-degree-of-freedom PID (2DoF-PID) type controller integrated with the linear extended state observer (LESO) and the friction feedforward (FFF) controller is presented. In this control framework, the linear ESO is designed to estimate and compensate for the slow-varying disturbance and the FFF controller is adopted to eliminate the abrupt-changing nonlinear friction effects. Finally, comparative simulations indicate that the 2DoF-PID-LESO-FFF is effective and achieves better performance compared with traditional controllers, and experiment results demonstrate that the tracking error of the proposed method can be kept in the bounds of 1.5 μm under different conditions.

A novel design of the setup for testing super high-pressure high-speed switch valves is proposed that provides more accurate dynamic characteristics and reduces the hardware and energy expenses. After reviewing the traditional setups for testing high-speed switch valves, we introduced the architecture and the working principle of the novel setup. Subsequently, a numerical comparison study is performed to confirm that the novel setup provides the same conditions for testing the switch valves as the common setup. The test rig for the super high-pressure high-speed switch valve was concurrently built to confirm the feasibility of the novel idea. The experimental results demonstrated that the novel setup meets the standard of testing the super high-pressure high-speed switch valves and can measure the dynamic flow rate signal. An overall appraisal of the novel setup is conducted and its advantages and disadvantages are discussed. In conclusion, the advantages of the new design are significant when it is applied for repeatability and durability tests of the super high-pressure high-speed switch valves.

With the deepening penetration of renewable resources worldwide, power system operators are faced with emerging challenges, e.g., the increase of operating risks due to the volatility and uncertainty of wind and solar power. To efficiently identify the operational limit violations, a switch from deterministic to stochastic framework for assessing the system security, which could manage various types of uncertainties, has been advocated in this paper. The established model is based on an improved probabilistic load flow, which is adapted to incorporate the steady-state behavior of frequency regulation. An efficient importance sampling (IS) technique is also developed to speed up the crude Monte Carlo (MC) simulation in estimating the low probability of violations of security constraints. Extensive computational experiments on both the IEEE 14-bus test case and a simplified regional system show that the proposed IS estimator makes significant enhancement to the crude MC in the computational efficiency and has better numerical performance as compared with other IS schemes.

Four-wheel redundantly-actuated mobile robot (FRMR) offers high controllability and maneuverability for automation applications. However, the robot dynamics and actuator failures are normally omitted in existing kinematic control schemes, which may lead to steering vibration and degraded robustness. To deal with these problems, a fault-tolerant dynamic control method is developed for precise trajectory tracking of the FRMR, which utilizes a two-level structure to cover the wheel-ground interactions and possible actuator failures. In the high level, with a novel fractional-order sliding mode control, this method offers an effective way to regulate the steering angle and eliminate the rotation chattering simultaneously. For the redundantly-actuated issue, a robust allocation solution is presented in the lower level to straightly determine optimal driving torques with full considerations of actuator failures and optimization efficiency. The convergence and stability of the achieved FRMR system are guaranteed theoretically. The experimental comparative results verify that higher tracking precision and enhanced robustness can be obtained using our proposed method.

In recent years, the discovery of local communities in multi-layer networks has become an active research field of complex systems. Such as communication, social networking, sensor network the rapid development of new technologies, the amount of data generated by increased, all want to obtain the network information difficulty is big, and the network, community, mutual influence between nodes, greatly enhance the complexity of network and make local found existing multilayer network method is unable to get more accurate test results. In this paper, based on the homogeneity drive of multi-layer network and the influence relation index of multi-layer path length measurement, a local community detection model based on the influence relation of the multi-layer network is proposed by combining the direct influence relation and indirect influence relation of the network (IMLC). Compared with six real multi-layer network data sets, the algorithm has better robustness in many most advanced multi-layer methods: GL, PMM, and ML-LCD.

Sequential allocation is a decentralized mechanism for allocating indivisible objects to agents, in which agents sequentially pick their favorite objects among the remainder based on a pre-defined priority ordering of agents (a sequence). The problem of choosing the “best”sequence of agents to achieve the optimal social welfare has been investigated and conjectured to be NP-hard. We propose a simple parallel allocation protocol that is insensitive to agents' identities. In every round of the parallel allocation process, every agent asks for an object among those that remain, and every reported object will be allocated randomly to an agent reporting it. Supposing additive utilities and independence between the agents, we compare the average (and worst-case) social welfare achieved by a parallel protocol and a sequential protocol. Theoretical and empirical results show that parallel protocol outperforms a sequential protocol (even when the best sequence of agents is applied). We also show that under the parallel mechanism, some manipulation problems (e.g., finding a successful strategy for a target set and finding an optimal strategy under some scoring function) can be solved in polynomial time.

In the above article [1], the funding information is incorrect and, thus, is modified as in the following paragraph: