To address the critical challenge of enabling generalizable and smooth path planning for heterogeneous construction vehicles (loader-dump truck) in building and mining operations, this paper proposes a novel collaborative planning framework where the core innovation is the DSN-GRU module-a learning and prediction component that integrates a Domain Separation Network for feature disentanglement and Gated Recurrent Units for sequential modeling. This is synergized with a dual-path fusion mechanism (Bi-elementary and CC-Steer curves) for continuous-curvature trajectory generation, and an improved GA-MPSO algorithm that reduces the number of iterations in the key-point generation process. By explicitly decoupling shared motion features from vehicle-specific parameters, our method demonstrates strong generalization across heterogeneous combinations, accommodating wheelbase differences up to 17.3% and steering variations up to 18.5%. Extensive simulations and experiments confirm consistent improvements, achieving over 10% reduction in both path length and travel time compared to existing methods, thereby providing a scalable and efficient solution for real-world mixed fleet operations.

Accurate prediction of the dynamic response of excavator hydraulic systems is a key prerequisite for model-based control applications. However, existing studies have not established a direct mapping from cylinder outputs to joystick signals. To address this issue, this study proposes an end-to-end predictive model that integrates a joystick-signal baseline with data-driven modeling. Specifically, the joystick-signal baseline is jointly determined by the mechanism-based relationship between cylinder motion and hydraulic flow and the predefined electro-hydraulic control mappings, providing an initial prediction reference for the model. Subsequently, a two-stage long short-term memory (LSTM) network is introduced for data-driven modeling. In the first stage, sensitivity-based weighting dynamically adjusts the weights of the input variables to capture the dynamic nonlinear characteristics of the system. In the second stage, the weighted input features from the first stage are combined with the joystick-signal baseline to form a new input sequence, which is then modeled using a multilayer LSTM. In addition, a delay-aware structure is designed to account for the non-fixed delays between joystick signals and cylinder states. Experimental results show that the prediction errors remain within 10 % under in-distribution validation, while R2 values above 0.88 are achieved on an independently collected gravel excavation dataset from another excavator of the same model, demonstrating good prediction performance and generalization capability. The proposed architecture requires only 4 MB of storage and achieves an inference time of less than 2 ms per prediction, indicating strong potential for real-time applications.

Under extreme operating conditions, coordinated steering and braking control is critical for vehicle safety. However, variations in road adhesion coefficients lead to differences in tire force utilization limits, necessitating dynamic adjustment of steering and braking control weights. To address this challenge, this study proposed a trajectory tracking control method with adhesion-aware dynamic weight allocation for steering and braking. Firstly, a lateral-longitudinal coupled vehicle dynamics model is developed, integrating front wheel steering angle and braking deceleration as control inputs. Secondly, a unified MPC framework is designed for trajectory tracking, where fuzzy logic dynamically adjusts the weighting coefficients between control effort and tracking error in the cost function based on real-time vehicle speed and reference trajectory curvature. Furthermore, to account for the influence of road adhesion coefficients on tire force utilization efficiency and driving stability, the mathematical model of weight coefficient correction is established. This two-stage adaptation mechanism integrates fuzzy logic and empirical rules based on mathematical models, enhances steering and braking maneuverability and trajectory tracking accuracy while rigorously guaranteeing stability across the adhesion-variant operating envelope. Simulation results demonstrate that the proposed controller achieves adaptive redistribution of steering and braking authority across different road surfaces, significantly improving handling stability and trajectory tracking precision compared to fixed-weight benchmarks. This method enhances the safety envelope of autonomous vehicles in adhesion-constrained emergency scenarios.

Purpose-Cam mechanisms are critical in modern automation equipment, but existing cam curve design methods ignore surface micro-features, limiting accurate micro-contact analysis. This study aims to propose a cam curve design method integrating micro-characteristics for precise control of contact performance. Design/methodology/approach-A cam curve is established via high-order differential interpolation combined with the W-M fractal function to characterize surface roughness. A modified micro-contact model is developed, and the control variable method is used to analyze the influence of high-order interpolation nodes (displacement, velocity and acceleration) on contact performance. Findings-Displacement and acceleration node variations induce unidirectional offset of contact performance extrema, while velocity nodes cause bidirectional offset; the sensitivity order is velocity > displacement > acceleration. The proposed curve enables accurate control of contact performance distribution. Originality/value-This study integrates high-order differential interpolation with the W-M fractal function, bridging macro kinematic design and micro-surface characterization, and provides theoretical/methodological support for "contact performance-controllable" cam design.

Accurate prediction of the excavator bucket's landing point is a critical aspect of achieving intelligent operation. However, existing prediction methods generally suffer from low computational efficiency, vulnerability to environmental factors, and poor feasibility, making them inadequate to meet the dual demands of efficiency and prediction information completeness required for engineering construction. To address these issues, this paper proposes a bucket landing point prediction method that integrates point cloud information. First, IMU data is filtered, and a kinematic model of the working device is established to calculate the bucket's pose information. Then, LiDAR point cloud data is processed, and dynamic region point clouds are extracted based on the bucket's pose, followed by the reconstruction of the elevation contour, on which the key point distances are calculated. Finally, various experimental scenarios were constructed in Gazebo to comprehensively validate the proposed method through simulation. The results demonstrate that the proposed method meets the requirements for real-time and complete information in predicting the excavator bucket's landing point.

For injured and after-stroke patients who temporarily lose their hand’s grasping abilities, assisting them in regaining their index finger mobility is very important in the rehabilitation process. In this paper, a finger rehabilitation device based on one degree-of-freedom (DOF) linkage mechanism is designed, aiming to lead the index finger through the flexion–extension trajectory during grasping tasks. Two types of one-DOF mechanisms, a four-bar linkage and a Watt-I six-bar linkage, are synthesized for the task trajectory. Various algorithms such as PSO, GA, and GA–BFGS are adopted and compared for the synthesis of these two types of mechanisms, among which the Watt-I six-bar linkage obtained with GA–BFGS shows the optimal performance in accuracy. Clinical biomechanical data are utilized to perform static analyses of the mechanisms, and the feasibility of the Watt-I six-bar linkage models is tested, compared, and demonstrated. Finally, the prototype of the six-bar linkage as well as a wearable exoskeleton finger rehabilitation device are designed to show how they are applied in the finger rehabilitation scenario.

Wheel loaders are multi-function construction equipment, yet their complex kinematics and coupled front-rear axle relationships pose significant challenges in path planning and tracking control. This paper presents a synchronous path planning method and a reinforcement learning-based tracking controller for these axles. Initially, a kinematic model and tracking error model are developed based on steering radius analysis and verified through simulations. The Reeds-Shepp curve is used to plan the V-shaped path for the front axle, with the rear axle path derived from the kinematic model. Using reinforcement learning, a path tracking controller is developed with a preview model of tracking error. Performance is validated through simulations and field tests, showing a position error reduction of 25% and 37.5% for two paths, and heading error reductions of 27.8% and 27.3%. The results also indicate improved stability in heading and articulation angles during tracking.

A method for calibrating meso-parameters in numerical simulations of granular media by the discrete element method (DEM) was studied in this work. Firstly, the resonant column (RC) test was carried out to calibrate the test results of the bender element (BE) test, in order to eliminate the drawback of the near field effect affecting the subjectivity in determining the shear-wave travel time. Secondly, the BE test was simulated by DEM, which was used as the calibration criterion for the mesoparameters in the simulations. Under the condition that the simulated material was the same as the BE test, the meso-parameters were adjusted to minimize the errors of the received shear-wave travel time between the DEM simulation and the BE test. It was found that the contact model in particles and the friction coefficient were the crucial factors affecting the consistency between the DEM simulation and the BE test. In addition, local damping and modest increases in confining pressure can effectively reduce errors in DEM simulations. The received signal in the effective simulation segment was affected by the boundary reflection, but the calibration criterion was not affected significantly.

Target detection in use-case environments is a challenging task, which is influenced by complex and dynamic landscapes, illumination, and vibrations. Therefore, this article presents a research on road target detection based on deep learning by combining image data of vision with point cloud data from light detection and ranging (LiDAR). First, the depth map of the point cloud was densified for the sparse and disordered characteristics of the point cloud, with the ground removed to create an image dataset that could serve for training. Next, considering the computational capacity of the on-board processor and the accuracy, the MY3Net network is designed by integrating Mobilenet v2, a lightweight network, as the feature extractor, and you only look once (YOLO) v3, a high-precision network, as the multiscale target detector, to implement the detection of red green blue (RGB) images and the densified depth maps. Finally, a decision-level fusion model is proposed to integrate the detection results of RGB images and depth maps with dynamic weights. Experimental results show that the proposed approach offers high detection accuracy even under complex illumination conditions.

The cam-linkage mechanism is a typical transmission mechanism in mechanical science and is widely used in various automated production equipment. However, conventional modeling methods mainly focus on the design and dimensional synthesis of the cam-linkage mechanism in the slow-speed scenario. The influence of component dimensions is not taken into consideration. As a result, the model accuracy dramatically falls when analyzing large-size cam-linkage mechanisms, especially in high-speed environments. The kinematic aspects of cam design have been investigated, but there are few studies discussing the motion characteristic and accuracy analysis models of the large-size cam-linkage mechanism under high-speed scenarios. To handle such issues, this paper proposes a parameter optimization methodology for the design analysis of the large-size high-speed cam-linkage mechanism considering kinematic performance. Firstly, the mathematical model of the cam five-bar mechanism is presented. The cam curve and motion parameters are solved forward with linkage length and output speed. Then, a particle swarm-based multi-objective optimization method is developed to find the optimal structure parameters and output speed curve to minimize cam pressure angle and roller acceleration and maximize linkage mechanism drive angle. A Monte Carlo-based framework is put forward for the reliability and sensitivity analysis of kinematic accuracy. Finally, a transverse device of a sanitary product production line is provided to demonstrate the applicability of the proposed method. With the parameter optimization, the productivity of the transverse device is doubled, from 600 pieces per minute (PPM) to 1200 PPM.

Current digging trajectory planning methods for excavator arms are limited to a single digging cycle, which does not meet the continuous excavation demands of the task. To address this issue, a real-time task-oriented continuous digging trajectory planning method for autonomous excavators is presented. The method involves optimizing digging trajectory for a single excavation cycle using multi-objective PSO method, building a PINN model using optimization results as training samples for real-time planning, and embedding the PINN model in planning framework for typical tasks. The method was validated using four different cross-section shapes of the trench. Results show that the average time taken to plan a single digging trajectory is less than 4.5ms, which is negligible compared to the time taken by PSO. The overall performance of the trajectory is only about 5% different from those planned by PSO. This method offers a more efficient and effective solution for continuous excavation tasks.

As a new kind of high-performance alloy suitable for elevated environments, refractory high entropy alloys (RHEAs) have attracted a wide range of research attention. How to process RHEAs has profound significance for their practical engineering applications. In this work, four RHEAs with different constituent phases were subjected to wire electrical discharge machining (WEDM). The effects of constituent phases and processing parameters on the WEDM performance of RHEAs were investigated. The findings have shown that the constituent phase is the primary factor that affects the WEDM performance of RHEAs, followed by the melting point. The relationship between the WEDM performance and the constituent phases of RHEAs was also established that the increase in the number of constituent phases or the content of additional phases is beneficial for the removal of workpiece materials. In addition, by exploring the effects of processing parameters on the WEDM performance, both higher cutting efficiency (CE) and lower surface roughness (Ra) in a single cutting pass were achieved, overcoming the trade-off between CE and machined quality. The present findings not only reveal the WEDM performance of RHEAs but also shed more light on the WEDM mechanisms of RHEAs with different constituent phases.

A resonant inertial impact rotary piezoelectric motor based on a self-clamping structure is designed, assembled, and tested. The designed piezoelectric motor mainly includes a rotor (two vibrators, preload mechanism, and intermediate connection mechanism), a clamping mechanism, and another auxiliary mechanism. The piezoelectric ceramic sheet on the rotor drives the vibrator to swing under the excitation of a single harmonic wave. Because there is a clamping mechanism formed by the combination of clamp baffle and fixed clamp ring, thus the half-cycle resonant rotation of the rotor can be effectively completed, and repeated harmonic excitation can realize the unidirectional continuous rotation and swing of the rotor. The whole excitation process of the motor is in a resonance state, which has significant advantages, such as low friction and simple structure, compared with the traditional quasi-static piezoelectric motor. The structure of the piezoelectric motor is designed and analyzed using COMSOL5.5 software and then the motor performance is tested and analyzed by building an experimental platform to verify the feasibility of the motor design. The final experimental results show that the optimal working frequency of the piezoelectric motor is 150 Hz, which is consistent with the characteristic frequency of the simulation. When the motor prototype is under the conditions of optimal operating frequency 150 Hz, voltage 240 Vp-p, and preload torque 7.8 N.mm, the maximum angular speed can reach 2.4 rad/s, the maximum load can reach 27.8 N mm and the maximum resolution of the movement angle can reach 0.941°.

We propose a new digital evaluation model for the negative-pressure molding mechanism, which is used to verify the feasibility of improvement measures for quality problem of the pad. The gas-solid flow in the negative pressure molding system facing the pad of diapers at the molding surface under different outlet pressures is studied using CFD-DEM bidirectional coupling simulation method. Aiming at the concentration of velocity/pressure values in the molding defect region, the uniformity coefficient is used as an important evaluation index to reflect the molding quality, and the parameters are optimised for the velocity and pressure uniformity of the gas phase in the molding defect region. For the fine flocculent discrete element, the parameter calibration is carried out, and the percentage of solid phase bearing at the screen is used as a reference index to further verify the feasibility of the improvement scheme. Finally, the model is applied to an improvement example, the results show that the molding defects of the cotton core layer are related to the velocity distribution of the molding surface. The molding quality can be significantly improved by setting a retaining ring at the outlet of the wind barn. The speed uniformity coefficient is improved by 44.2% when the retaining ring thickness s = 140 mm, inner diameter d = 1000 mm, and the relative position l = 0 mm. The molding quality has been significantly improved.

A resonant-type inertial impact linear piezoelectric motor based on coupling of driving and clamping parts was designed and manufactured. The motor mainly includes stator (coupling of driving and clamping parts), mover (slider) and auxiliary parts. The driving part works in the resonant state under the excitation of single harmonic, which mainly realizes the function of reciprocating driving. Similarly, under the single harmonic driving, the clamping part also works in the resonant state to realize the clamping function. Through the coupling between the two parts of the stator, the mover is driven to move continuously in one direction. The inertial impact piezoelectric motor works in the resonant state because the driving and clamping parts work in the resonant state respectively. Compared with the traditional quasi-static inertial impact motor, this study novelly changes the working state of the inertial impact motor. Through the finite element simulation software COMSOL 5.2 , the resonant frequency coupling of the driving and the clamping part is consistent. An experimental platform was built to verify the feasibility of the principle by testing the motor prototype. The experiment results show that: The maximum speed reaches 78 mm s −1 when the motor prototype is operated at the frequency of 810 Hz with a preload of 2 N and the working voltages of clamping and driving parts of motor were set at 80 and 220 V p-p respectively. Meanwhile, the maximum load of the motor prototype can reach 5 N. The minimum resolution of the motor prototype is 6.379 μ m.

As a potential candidate for the next generation of high-temperature alloys, refractory high entropy alloys (RHEAs) have excellent mechanical properties and thermal stability, especially for high-temperature applications, where the processing of RHEAs plays a critical role in engineering applications. In this work, the wire electrical discharge machining (WEDM) performance of WNbMoTaZrx (x = 0.5, 1) RHEAs was investigated, as compared with tungsten, cemented carbide and industrial pure Zr. The cutting efficiency (CE) of the five materials was significantly dependent on the melting points, while the surface roughness (Ra) was not. For the RHEAs, the CE was significantly affected by the pulse-on time (ON), pulse-off time (OFF) and peak current (IP), while the surface roughness was mainly dependent on the ON and IP. The statistical analyses have shown that the CE data of RHEAs have relatively-smaller Weibull moduli than those for the Ra data, which suggests that the CE of RHEAs can be tuned by optimizing the processing parameters. However, it is challenging to tune the surface roughness of RHEAs by tailoring the processing parameters. Differing from the comparative materials, the WEDMed surfaces of the RHEAs showed dense spherical re-solidified particles at upper recast layers, resulting in larger Ra values. The proportion of the upper recast layers can be estimated by the specific discharge energy (SDE). Following the WEDM, the RHEAs maintained the main BCC1 phase, enriched with the W and Ta elements, while the second BCC2 phase in the Zr1.0 RHEA disappeared. Strategies for achieving a better WEDMed surface quality of RHEAs were also proposed and discussed.

基于大脑可塑性理论和连续被动康复训练运动(continuous passive motion,CPM)疗法在手部康复的运用,文章设计了一种外骨骼机械手.该机械手各手指设置一个自由度,利用双平行四边形投影机构和关节对中结构相结合的方式带动患指;为适应康复时的手部姿态,设计腕关节弯曲机构和桡尺关节翻转机构;对手指训练机构进行运动学分析和仿真,并推出机构带动手指的角度运动关系;最后针对机械手的穿戴适应性和食指的实际运动效果进行康复试验探究.结果表明:该机械手结构简单,能够避免机构对手指关节软组织的损伤;穿戴适应性良好,手指实际运动角度重复性误差小于5％.该康复外骨骼机械手能够满足脑卒中患者手部康复训练的要求.

痉挛状态等级的客观定量评定对于偏瘫康复治疗具有积极意义.针对改良Ashworth量表(modified Ashworth scale,MAS)临床评定痉挛状态等级主观性高、量表痉挛等级划分不精准的问题,文章提出一种基于层次聚类的上肢痉挛状态定量评定方法.根据痉挛状态发生过程中的动力学机理设计痉挛状态等级量化评定装置;利用临床实验采集患者上肢被动牵伸过程中加速度以及角度数据判定牵张反射阈值,计算牵张反射阈值与关节活动度比值作为痉挛状态等级量化评定指标;利用层次聚类算法对量化评定指标聚类划分,重新评估痉挛状态等级.实验结果表明,基于层次聚类的上肢痉挛状态定量评定方法可以有效消除M AS评定痉挛状态等级中的主观性因素,能更加精准评定和细化痉挛状态等级.

This paper reported a combination of powerful mechanical dispersion and chemical dispersion to solve the agglomeration of lithium iron phosphate (LiFePO4) fine powder in pulping process. The effect of the addition of dispersant fatty alcohol-polyoxyethylene ether (AEO-7) on the dispersibility of LiFePO4 slurry was compared, and the slurry prepared by traditional pulping process was also been compared. The dispersion uniformity of LiFePO4 particles and the rheological properties of the slurry were significantly improved with the addition of the dispersant AEO-7. The surface of the prepared electrode was regular, and the corresponding electrode showed a high compaction density. The LiFePO4 sample prepared with AEO-7 demonstrated excellent electrochemical performance of initial discharge capacities at 0.2, 0.5, 1, 2, and 5 C are 163.3, 160.9, 155.1, 146.2, and 126.5 mAh g−1, respectively. Furthermore, the14500 type steel shell battery prepared with different electrodes has also been studied.

Feeding and unloading operation for part manufacturing are widely applied by industrial robots. In this paper, a set of algorithms has been used to reach higher efficiency and automation of trajectory planning for a 6-DOF integrated serial kinematic manipulator. Depending on the key nodes of the joint angles calculated by a hybrid inverse kinematics algorithm, the continuous quintic B-spline curve algorithm was utilized for planning smooth trajectories of the feeding motion from peer to peer. An adaptive cuckoo search (ACS) algorithm with high efficiency and excellent stability was proposed to minimize the total motion time under strict dynamic constraints. Comparing with 5 commonly used heuristic methods, the ACS algorithm has faster convergence speed and higher accuracy based on the same fitness function. To verify the implementation effect of the strategy, a 1:5 scale experimental platform was designed and built to implement the time-optimal trajectories. The simulations and experiments indicate that these algorithms lead to efficient planning of time-optimal and smooth trajectory in joint space.