Partially defective fingerprint image (PDFI) with poor performance poses challenges to the automated fingerprint identification system (AFIS). To improve the quality and the performance rate of PDFI, it is essential to use accurate segmentation. Currently, most fingerprint image segmentations use methods with ridge orientation, ridge frequency, coherence, variance, local gradient, etc. This paper proposes a method of XFinger-Net for segmenting PDFIs. Based on U-Net, XFinger-Net inherits its characteristics. The attention gate with fewer parameters is used to replace the cascaded network, which can suppress uncorrelated regions of PDFIs. Moreover, the XFinger-Net implements a pixel-level segmentation and takes non-blocking fingerprint images as an input to preserve the global characteristics of PDFIs. The XFinger-Net can achieve a very good segmentation effect as demonstrated in the self-made fingerprint segmentation test.

Purpose: Due to the market trend of low-volume and high-variety, the manufacturing industry is paying close attention to improve the ability to hedge against variability. Therefore, in this paper the assembly line with limited resources is balanced in a robust way that has good performance under all possible scenarios. The proposed model allows decision makers to minimize a posteriori regret of the selected choice and hedge against the high cost caused by variability.Design/methodology/approach: A generalized resource-constrained assembly line balancing problem (GRCALBP) with an interval data of task times is modeled and the objective is to find an assignment of tasks and resources to the workstations such that the maximum regret among all the possible scenarios is minimized. To properly solve the problem, the regret evaluation, an exact solution method and an enhanced meta-heuristic algorithm, Whale Optimization Algorithm, are proposed and analyzed. A problem-specific coding scheme and search mechanisms are incorporated.Findings: Theory analysis and computational experiments are conducted to evaluated the proposed methods and their superiority. Satisfactory results show that the constraint generation technique-based exact method can efficiently solve instances of moderate size to optimality, and the performance of WOA is enhanced due to the modified searching strategy.Originality/value: For the first time a minmax regret model is considered in a resource-constrained assembly line balancing problem. The traditional Whale Optimization Algorithm is modified to overcome the inferior capability and applied in discrete and constrained assembly line balancing problems.

The working characteristics of wet clutches have an important impact on the safety performance of vehicles. In order to obtain the thermoelastic instability characteristics of wet clutch separate plate, a finite element modeling method is proposed. The temperature field calculation model of separate plate and its thermo-hydro-mechanical coupling relationship are constructed. The distribution law of high-temperature hot spots on the surface of separate plate is obtained and the thermoelastic instability mechanism is revealed. Effectiveness of the simulation model is verified by road test, and surface topography of separate plate is observed by scanning electron microscope. A thermoelastic instability calculation model considering different material parameters is established. The temperature field distribution law is reviewed under different elastic moduli, specific heat capacities, thermal conductivities, and thermal expansion coefficients. Results show that increasing the specific heat capacity and thermal conductivity of the separate plate, decreasing the elastic modulus and thermal expansion coefficient can improve the stability of the system. The thermal conductivity and thermal expansion coefficient have important effects on the thermoelastic instability. The specific heat capacity has a certain effect, and the elastic modulus’ effect is the least. The research results of this paper can provide theoretical support for optimizing the structure of wet clutch and improving the stability of the system.

In automatic parking scenarios, obstacles on both sides of a target parking space (TPS) will degrade the feasibility and comfort of passengers boarding and alighting (CPBA) when the target vehicle maintains a standard final parking state (FPS). Additionally, the presence of dynamic obstacle vehicles (DOV) in the parking lot severely challenges the adaptability of path - planning algorithms. To tackle these issues, this paper presents a method for FPS optimization and two-stage path planning in dynamic scenarios while considering CPBA. First, a CPBA characterization variable is established using the allowed opening angle of the vehicle door (OAVD) restricted by adjacent obstacles. Then, to enhance path quality, the basic motion curve (BMC) of the vehicle is formulated for the design of queuing, parking, and transition paths. A two-stage path planning framework that integrates the parking lot system and the automatic parking system (APS) of the target vehicle is proposed. It includes a Dijkstra algorithm with embedded occupancy dynamic time windows of corridor nodes (DDTW) for queuing path planning and a parking path planning method based on the multi-objective particle swarm optimization (MOPSO) algorithm, aiming to improve CPBA and parking efficiency. Simulation experiments using real-world parking lot and vehicle parameters show that the proposed method can effectively reduce queuing time and enhance comfort, validating its effectiveness. A comparison with other parking path planning methods further demonstrates the superiority of the path performance achieved by the proposed approach.

Parking in narrow spaces presents significant challenges, often resulting in deviations between vehicle’s final parking state (FPS) and normal FPS. These deviations can cause partial intrusion into the target parking spots, increasing the risk of collisions in the parking process and potentially making the parking spots unusable. To address these issues, this paper proposes an optimization method for both the parking path and FPS in narrow and non-ideal vertical parking scenarios. Initially, a partition-based calculation method for the minimum distance between the vehicle and obstacles (DBVO) was developed to quantitatively assess the impact of intrusion on parking safety. Following this, a predefined geometric set of clothoids is used to smooth curvature discontinuities in the parking path, and a four-phase parking path pattern is devised. Subsequently, a preferred method for parking manners is proposed by analyzing the effects of parking direction and maneuvers on spatial requirements. Finally, a two-step parking path optimization method is presented with the framework integrating both online and offline calculations, using the minimum DBVO field and a predefined path pattern. Comparative experiments demonstrate that this method could enhance the proportion of available intrusion scenarios, increase the success rate of effective path acquisition, and improve the overall quality of parking paths.