Complexity最新文献

In deep learning, the choice of activation function plays a vital role in enhancing model performance. We propose AHerfReLU, a novel activation function that combines the rectified linear unit (ReLU) function with the error function (erf), complemented by a regularization term 1/(1 + x²), ensuring smooth gradients even for negative inputs. The function is zero centered, bounded below, and nonmonotonic, offering significant advantages over traditional activation functions like ReLU. We compare AHerfReLU with 10 adaptive activation functions and state-of-the-art activation functions, including ReLU, Swish, and Mish. Experimental results show that replacing ReLU with AHerfReLU leads to 3.18% improvement in Top-1 accuracy on the LeNet network for the CIFAR100 dataset, 0.63% improvement on CIFAR10%, and 1.3% improvement in mean average precision (mAP) on the SSD300 model in the Pascal VOC dataset. Our results demonstrate that AHerfReLU enhances model performance, offering improved accuracy, loss reduction, and convergence stability. The function outperforms existing activation functions, providing a promising alternative for deep learning tasks.

This paper designs an optimal control strategy for the auto-berthing control (ABC) problem of underactuated surface vessels. The purpose is to achieve accurate berthing of USVs in complex environments. In the traditional ABC problem of surface vessels, the underactuated characteristics of USVs complicate the control design. This underactuated phenomenon is manifested in the lack of lateral driving force on the boat, which restricts its motion to a specific direction. In order to solve the problem of the lack of lateral driving force of USVs, this paper first uses diffeomorphism transformation to transform the original nonholonomic constraint system into a chain structure. Through this transformation, the problem of the state of the vessel being restricted during motion is avoided. On this basis, this paper further describes the optimal control problem of the USV on the horizontal plane based on the kinematic and dynamic models of the USV. In order to solve the optimal control problem, this paper adopts the Gaussian pseudospectral method (GPM). By discretizing the continuous optimal control problem into a nonlinear programming problem, the computational complexity is effectively reduced and the solution efficiency is improved. The optimal navigation trajectory of the USV system and the corresponding optimal control input are then obtained. In order to verify the feasibility and effectiveness of the proposed control strategy, numerical simulation is carried out. The simulation results show that the optimal control strategy can achieve stable and accurate berthing of USVs.

Tracking athletes in high-speed outdoor sports like alpine skiing causes substantial difficulties because of ever-changing movements, environmental variability, and the limitations of traditional tracking technologies, such as intrusive sensors and single-view camera setups. This study proposes a hybrid approach for tracking alpine skiing activities by combining YOLO-v8 with an evolutionary version of the chimp optimization algorithm (CHOA-EVOL) for optimizing hyperparameters. The primary goal of this research is to enhance the CHOA to optimally adjust the hyperparameters of YOLO-v8, consequently addressing the drawbacks of outdoor sports tracking technology. This hybrid model integrates data from unmanned aerial vehicles (UAVs) and terrestrial cameras to better understand athletes’ rapid rotating motion. The suggested approach is extensively tested and validated using advanced algorithms with the UAV123 dataset and a recently developed alpine skiing dataset (ASD). The results have shown that our proposed approach can achieve high precision and robustness.

The existence of clearance in the joint causes positional deviation and reduces the accuracy of robotics, in which nonlinear factors such as friction and lubrication inside the joint seriously affect its dynamic behaviors. On the other hand, the effectiveness of control algorithms is largely dependent on the dynamic complexity of the robotics, while its complexity analysis is an important prerequisite for achieving high-performance control. In this paper, we provide a method for analyzing the dynamic complexity of a planar manipulator with clearance and lubrication in joints based on a nonlinearity measure. The influences of joint parameters such as friction coefficient, lubricant viscosity, and gap radius are quantified on the dynamic complexity of the kinetics, which can identify the collision state of the joints effectively in the robot’s motion. First, a dynamical model of a planar manipulator is established by integrating the contact-separation model and the force model at the joint, and a method for calculating the dynamic complexity based on the nonlinearity measure is proposed. Second, the effects of different joint parameters on the dynamic behaviors of the robotic system are analyzed, and the relationships between the lubricant viscosity, gap radius, and the dynamic complexity under different friction coefficients are established to analyze the impacts of the joint parameters on the dynamic complexities. The results show that the dynamic complexity of the robotic system can be decreased significantly by using the small gap radius and high viscosity of the lubricant, which can help to realize the better control performance. Especially, this method is more sensitive to state changes such as the collision state of the joint relative to the sample entropy method.

Inspired by the genetic evolution mechanism of DNA, a hybrid DNA Gray Wolf Optimizer (hDNA-GWO) is proposed to develop an accurate kinetic model. This algorithm incorporates innovative DNA encoding, selection, crossover, and mutation operators inspired by genetic processes. We adopt the roulette-wheel method to select individuals with greater environmental adaptability from the current population to form the next population. The crossover operation involves swapping gene segments between paired chromosomes to create new individuals and maintain the population diversity. The mutation operator can maintain the diversity of the population, avoid the phenomenon of “premature” convergence, and effectively improve the local search capability. The performance of hDNA-GWO is investigated on typical benchmark functions compared to GWO, PSO, GWO-PSO, and GWO-GA. In addition, the superior search capabilities of our model are validated by kinetic parameter estimation using experimental data from supercritical water oxidation processes. The results indicate that the hDNA-GWO can overcome premature convergence and obtain higher-quality global optimal solutions.

Using the matching data of the UN COMTRADE, China Customs, and China Industrial Enterprise database from 2000 to 2013, we explore the relationship between product density and the export technological complexity at the enterprise level. We find product density at the firm level has a significant positive effect on the export technological complexity of Chinese manufacturing enterprises. The result remains still robust after adopting instrumental variable method, replacing explanatory and explained variables and substituting more detailed data for the proxy variables. Both specialized agglomeration and diversified agglomeration have significant positive effects on the export technological complexity. The results of moderating effect test indicate that specialized industry agglomeration has significantly reinforced the abovementioned effects as a moderator. However, diversified agglomeration significantly inhibits the technological upgrading effect. Heterogeneity tests show that product density is more conducive to the upgrading of the export technological complexity of enterprises in the eastern region, and so are to the foreign-funded, capital-intensive, and technology-intensive enterprises.

To suppress the frequency oscillation phenomenon that occurs in the parallel control system of multiple virtual synchronous generators (multi-VSG) during load mutation, this paper proposes a multi-VSG parallel control strategy based on sliding mode linear active disturbance rejection (SM-LADRC). Initially, mathematical modeling of the multi-VSG parallel control system is conducted to analyze the mechanism by which load mutation affect frequency. Subsequently, based on the rotor motion equation of the VSG, linear active disturbance rejection control (LADRC) is applied to the angular frequency, and an extended state observer (ESO) is constructed to estimate and compensate for the system’s frequency state and load mutation in real time, thereby enhancing the system’s disturbance rejection capability. Concurrently, an integral sliding mode linear state error feedback (SM-LSEF) control law is formulated to rapidly adjust the frequency error control quantity, eliminating the reaching phase and accelerating the system’s response speed. Moreover, the integral sliding mode, by introducing an integral term, continuously approximates the switching function, making the sliding mode surface smoother, which effectively suppresses sliding mode chattering and improves the system’s robustness. Finally, simulation comparisons validate the correctness and effectiveness of the proposed control strategy, providing a theoretical and simulation experimental basis for engineering applications.

We discuss the relationship between environmental sustainability and system complexity. This is motivated by the fact that solutions to environmental challenges often create additional complexity in the overall socioeconomic system, at local to global levels. This increase in complexity might hamper the ultimate achievement of sustainability. This theme is over utmost importance but is overlooked in studies of environmental sustainability, environmental and climate policy, and sustainability transitions. It merits serious attention as this can provide a general basis and clarification of related topics that are currently studied in isolation—think of energy rebound, carbon leakage, green paradox (fossil fuel market responses to climate policy), circular economy, and environmental problem shifting. The relationship between complexity and sustainability is examined from thermodynamic and systemic perspectives, resulting in identifying a set of mechanisms of complexity increase and clarifying how this potentially creates barriers to meeting sustainability goals. While this issue is pertinent to all economies and countries, it is of high relevance to developing countries as their economies are likely to undergo considerable complexity increases in the near future due to further development. The question is then whether countries will be able to steer their development in a sustainable direction while simultaneously limiting a more roundabout nature of their production structure. We contend that this may require “complexity policy” and outline ideas in this regard. An important role can be played by cap-and-trade, but this will work mainly for carbon emission and not for other environmental pressures. Ultimately, a policy mix could guide different subsystem complexities in terms of environmental pressures and welfare impacts—resulting in optimizing system complexity for sustainability.

In the context of rapid development of social networks, users gather together through social network platforms (SNPs) to form social circles, which become the main carriers for users to disseminate public opinion information and emotions, and it is easy to form the phenomenon of network public opinion polarization in social circles. To prevent the polarization of network public opinions in social circles, the application of blockchain technology to SNPs can provide a credible database for the traceability of public opinions and avoid the spread of false information in social circles. Based on blockchain traceability, this paper establishes a three-party evolutionary game model consisting of government, SNPs, and social circle managers (SCMs). Through model solving and numerical simulation, the influencing factors of each party’s strategy are analyzed. The findings suggest: (1) Although the rising cost of building blockchain platforms may undermine the government’s incentive to regulate, the government will still prioritise support for their construction if they can significantly enhance social welfare. (2) The government could boost SNPs’ willingness to adopt blockchain technology through subsidy measures, but increased privacy risks could undermine this willingness. (3) Blockchain technology realizes the traceability of public opinion and improves the rational guidance tendency of SCMs. Therefore, in order to promote and strengthen the collaborative governance of network public opinion in social circles, SNPs should enhance the use of blockchain technology, while the government should give full play to its regulatory and supervisory roles. This study introduces blockchain technology for public opinion traceability; explores a collaborative network public opinion governance system based on strict regulation by the government, introduction of blockchain technology by SNPs, and rational guidance by SCMs; expands the related research on network public opinion governance; and provides a new idea for the study of network public opinion governance of social circles based on blockchain traceability.

Correct detection of plant diseases is critical for enhancing crop yield and quality. Conventional methods, such as visual inspection and microscopic analysis, are typically labor-intensive, subjective, and vulnerable to human error, making them infeasible for extensive monitoring. In this study, we propose a novel technique to detect tomato leaf diseases effectively and efficiently through a pipeline of four stages. First, image enhancement techniques deal with problems of illumination and noise to recover the visual details as clearly and accurately as possible. Subsequently, regions of interest (ROIs), containing possible symptoms of a disease, are accurately captured. The ROIs are then fed into K-means clustering, which can separate the leaf sections based on health and disease, allowing the diagnosis of multiple diseases. After that, a hybrid feature extraction approach taking advantage of three methods is proposed. A discrete wavelet transform (DWT) extracts hidden and abstract textures in the diseased zones by breaking down the pixel values of the images to various frequency ranges. Through spatial relation analysis of pixels, the gray level co-occurrence matrix (GLCM) is extremely valuable in delivering texture patterns in correlation with specific ailments. Principal component analysis (PCA) is a technique for dimensionality reduction, feature selection, and redundancy elimination. We collected 9014 samples from publicly available repositories; this dataset allows us to have a diverse and representative collection of tomato leaf images. The study addresses four main diseases: curl virus, bacterial spot, late blight, and Septoria spot. To rigorously evaluate the model, the dataset is split into 70%, 10%, and 20% as training, validation, and testing subsets, respectively. The proposed technique was able to achieve a fantastic accuracy of 99.97%, higher than current approaches. The high precision achieved emphasizes the promising implications of incorporating DWT, PCA, GLCM, and ANN techniques in an automated system for plant diseases, offering a powerful solution for farmers in managing crop health efficiently.