This paper demonstrates the effectiveness of integrating computational intelligence to enhance the reliability of millimeter wave technology as a detection device for hazardous chemicals. The research explores the use of millimeter wave as an efficient and dependable alternative technology for chemical detection with the aid of machine learning to further improve its reliability and accuracy. This advancement is crucial in enabling security agencies, and authorities to remotely identify hazardous chemicals, minimizing risks to human lives and properties. The millimeter wave relies on natural non-ionizing radiation, which is of low power and considered safe for human exposure. The millimeter wave region used in this study is 77–81 GHz that offers short-pulse transmission capabilities, producing a wide spectrum of frequencies. These short pulses serve as the source for collecting the broadband spectral identity of chemicals, and the subsequent detection is post-processed with machine learning to increase the level of accuracy. The result of this study shows that by using computational intelligence models such as decision tree, k-nearest neighbor, support vector machine, and random forest, enhances the overall device reliability, and achieves higher detection accuracy based on the received reflected power. This result is comparable to an X-ray system device.

Speech is one of the most sophisticated human motor skills. Speaker identification is the ability of a software component or hardware to acquire a speech signal, recognize the speakers included in the signal, and identify the speaker after the audio signals have been received. This study proposes a fluctuating equation inversion method using feature extraction for broadcast hosting. Feature extraction aims to provide useful signal features from natural audio that can be applied to various downstream processes, including recitation, evaluation, and categorization. Initially, data were collected from the CASIA dataset. This study evaluated the experimental outcomes of the proposed approach using mel-frequency cepstral coefficients, gammatone frequency cepstral coefficients, and linear frequency cepstral coefficients. The proposed technique was tested on a publicly accessible dataset, and the findings showed that it performed better in terms of recognition accuracy (98%), precision (97%), recall (96.05%), sensitivity (92.56%), and F1-score (95.09%) than the conventional feature extraction methods. The proposed approach can be utilized to improve audio signal quality and user experience across broadcast-hosting applications.

In this study, we propose YOLOv5s-PGD algorithm for dense pedestrian detection, which can improve the recall and reduce the number of parameters compared with YOLOv5s. First, a minimum scale detection layer has been added to deepen the pyramid’s depth and enhance detection accuracy. Second, ghost convolution has been employed to replace standard convolution to increase real-time performance of the algorithm. Finally, depth separable convolution has been used to address issues of high parameters and large computational complexity that lower the efficiency of the algorithm. Experiment results demonstrate that the detection accuracy of the YOLOv5s-PGD algorithm on the CrowdHuman public dataset is up to 85.1%, which is 2.2% higher than that of YOLOv5s. Furthermore, the number of parameters has decreased by 19.7%, and the calculation burden has decreased by 2.5%. Consequently, the proposed YOLOv5s-PGD algorithm better satisfies the requirements of real-time detection, model optimization, and terminal deployment in dense pedestrian scenarios.

Employing Chinese A-share market data, this study explores how news text and economic policy uncertainty (EPU) can be combined to predict a company’s unanticipated earnings using the XL (extra long) Transformer and long short term memory (LSTM) models. The results show that adding news text features or the EPU index can improve the model’s predictive performance. However, adding the EPU index improves the model prediction performance by a tiny amount. Next, news headlines have better predictive performance relative to news content. Meanwhile, as a supplement to news headlines, news content can further improve predictive performance. Finally, the XL-Transformer model has better predictive performance than the LSTM model, but the improvement in the effect is limited.

The premature newborn receives specialized medical care in the neonatal intensive care unit (NICU), where various medical devices emit excessive light and sound stimulation, and those prolonged exposures to stimuli may cause stress and hinder the development of the newborn’s nervous system. The formation of their biological clock or circadian rhythm, influenced by light and sound, is crucial for establishing sleep patterns. Therefore, it is essential to investigate how the NICU environment affects a newborn’s sleep quality and rhythms. Brazelton’s classification criteria measure the sleep-wake state of newborns, but the visual classification is time-consuming. Therefore, we propose a method to reduce the burden by automatically classifying the sleep-wake state of newborns from video images. We focused on videos of whole-body and face-only videos of newborns and classified them into five states according to Brazelton’s classification criteria. In this paper, we propose and compare methods of classifying whole-body and face-only videos separately using a three-dimensional convolutional neural network (3D CNN) and combining the two results obtained from whole-body and face-only videos with time-series smoothing. Experiments using 16 videos of 8 newborn subjects showed that the highest accuracy of 0.611 and kappa score of 0.623 were achieved by weighting the time-series smoothed results from whole-body and face-only videos by the output probabilities from the 3D CNN. This result indicated that the time-series smoothing and combining the results based on probabilities is effective.

In allusion to the problem that teachers not being able to timely grasp student dynamics during online classroom, resulting in poor teaching quality, this paper proposes an online learning status analysis method that combines facial emotions with fatigue status. Specifically, we use an improved ResNet50 neural network for facial emotion recognition and quantify the detected emotions using the pleasure-arousal-dominance dimensional emotion scale. The improved network model achieved 87.51% and 75.28% accuracy on RAF-DB and FER2013 datasets, respectively, which can better detect the emotional changes of students. We use the Dlib’s face six key points detection model to extract the two-dimensional feature points of the face and judge the fatigue state. Finally, different weights are assigned to the facial emotion and fatigue state to evaluate the students’ learning status comprehensively. To verify the effectiveness of this method, experiments were conducted on the BNU-LSVED teaching quality evaluation dataset. We use this method to evaluate the learning status of multiple students and compare it with the manual evaluation results provided by expert teachers. The experiment results show that the students’ learning status evaluated using this method is basically matched with their actual status. Therefore, the classroom learning status detection method based on facial expression recognition proposed in this study can identify students’ learning status more accurately, thus realizing better teaching effect in online classroom.

Innovation and green development are important aspects of a country’s economic and social development. Exploring the impact of innovation on green development from a spatial perspective can help local governments suggest measures to promote green development. The super-efficiency slack-based measure model and stochastic frontier analysis model were used separately to measure green development and innovation. The coupling coordination degree model was used to measure the interaction of “double efficiency.” Based on the spatial panel Durbin model, the spatial impact of innovation behavior on green development was studied, and regional heterogeneity was further studied. The study found that in the eastern region, provinces with a high degree of innovation and a developed financial industry have a negative effect on the green development of their neighboring provinces. In the central and western regions, the higher the innovation level of a region, the more negative the green development of neighboring regions is. A high-level financial industry promotes the green development of neighboring regions.

The declining trend in China’s fertility rate is very pronounced, and since 2023, the population has entered a phase of negative growth, significantly constraining economic development. At the same time, income inequality, which creates many problems, remains a serious issue. Existing research does not discuss the impact of income inequality on fertility rates in China currently. This paper depends on both macro and micro perspectives to examine the relationship between income inequality and fertility. We used the macro perspective to study the impact of urban-rural income inequality on the birth rate. The results show that the widening income gap between urban and rural areas will reduce the birth rate, but this effect declines with increasing of the birth rate; this negative effect is the strongest in the eastern region and the weakest in the western region. We used micro perspective to study the effect of the Gini coefficient on fertility motivation. It was found that for every one percent increase of the Gini coefficient, the fertility motivation decreased by about 0.08%, indicating that income inequality also impacts fertility behavior negatively. Comparing different income groups, income inequality has no impact on the fertility motivation of low-income groups, but has a significant negative and positive impact, respectively, on the middle and high-income groups. The mediating effect model shows that income inequality can negatively affect fertility behavior by reducing social trust and subjective well-being. Therefore, the empirical results from China suggest that income inequality negatively affects fertility behavior.

The anti-intereference ability can be strengthened and the BER can be decreased when space-time block code (STBC) is applied to MIMO systems. The existing quasi-orthogonal code TBH can effectively improve the system capacity, but the effect is not good at high signal-to-noise ratio (SNR). In this paper, we introduce an improved quasi-orthogonal code, derive system capacity formula not with space-time coding but with TBH and improved quasi-orthogonal STBC, and simulate the relationship curve of the SNR and capacity in these systems based on MATLAB. The simulation shows that the system capacity with the improved code is bad in low SNR, but much better than that of the existed code with the increasing SNR, and the advantage on increasing capacity is outstanding, especially when the SNR is high.

This study seeks to enhance the classification performance of breast ultrasound images, addressing the challenges of difficult and costly collection of breast ultrasound datasets as well as the discrepancies in feature distribution of the collected datasets. Performance is enhanced by using a mix of generative adversarial networks (GAN) and domain adaptive networks. First, an adaptive layer is first added to the basic model of the gradually vanishing bridge (GVB), to better match the feature distributions of the source and target domains of the dataset. The multi-kernel maximum mean discrepancy (MK-MMD), which is the most efficient of existing adaptive approaches, is implemented in the fully connected layer of the original model’s feature extraction network. Finally, through the process of fine-tuning, the model that has the highest level of overall performance is determined. In experiments, the proposed method surpassed the conventional unsupervised domain adaptation (DDC) and adversarial domain adaptation (MK_DAAN, GVB) in performance, achieving 85.11% accuracy, 97.48% recall, and 0.92 F1-score.

Coordinating the relationship between the digital economy, green innovation, and the urban-rural income gap is conducive to promoting common prosperity in high-quality development. This study aims to show that, in the short term, the development of the digital economy will promote the level of green innovation, which, in turn, will promote the development of the digital economy, but will also widen the urban-rural income gap. The study uses panel data of 273 prefecture-level cities from 2011 to 2019, and adopts the panel vector error correction model for quantitative analysis, combined with the theoretical analysis of the long- and short-term causal relationship among the digital economy, green innovation, and urban-rural income gap. The results reveal that, in the long term, the digital economy, green innovation, and urban-rural income gap demonstrate a double-circular causality that is positively promoting each other. The robustness test validates the conclusions. Therefore, while promoting the development of the digital economy and green innovation, the government should control the urban-rural income gap within a reasonable range, to provide theoretical and practical support for promoting the sustainable development of China’s economy and realizing the goal of common prosperity.

The carbon market was established to reduce carbon dioxide emissions. The traditional fossil energy market, new energy market, and carbon market have interrelated effects such as substitution, demand, and production inhibition, which can potentially lead to risk transmission. This study examines the nonlinear volatility correlation between China’s carbon market, China’s new energy market, and the international crude oil futures market. Seven submarkets within these three markets are selected for analysis. By measuring volatility risk through the conditional heteroscedasticity of returns, the analysis of nonlinear Granger causality networks reveals that, from a nonlinear perspective, risk primarily spills over through the paths of “International crude oil futures market → China’s carbon market” and “International crude oil futures market → China’s new energy market → China’s carbon market.” China’s carbon market serves as a recipient of risk, with minimal spillover effects. Therefore, further optimization is needed for the framework of China’s carbon market to enhance its asset allocation function and promote its spillover influence. Investors in China’s carbon market should consider both linear and nonlinear risks from China’s new energy market and the international crude oil futures market, and take appropriate measures to facilitate the sustainable growth of Chinese enterprises.

This study defines financially distressed enterprises based on stock delisting risk warnings and uses the annual data of A-share listed companies on the Shanghai and Shenzhen stock exchanges from 2008 to 2021 to examine the impact and mechanism of financial distress on the capital expenditures of non-distressed enterprises in the same city. The results indicate that if financially distressed enterprises exist within a city, the capital expenditure of non-distressed enterprises within the same city will subsequently decrease. The conclusions hold after multiple robustness and endogeneity tests. Mechanism tests show that financially distressed enterprises reduce the operating performance and cash flow of non-distressed enterprises in the same city through business performance contagion, thereby reducing their intrinsic motivation for capital expenditure. However, financial distress enhances creditors’ credit risk perceptions of non-distressed enterprises in the same city through signal transmission effects, prompting creditors to tighten credit contracts or directly intervene in corporate capital expenditure decisions, thus suppressing corporate capital expenditure. Heterogeneity tests indicate that the smaller the asset size of non-distressed enterprises, the larger the scale of financially distressed enterprises relative to non-distressed enterprises in the same city, or the more severe the agency problem of non-distressed enterprises or degree of financial distress, the more significant the negative externality of financial distress on the capital expenditures of local enterprises. The economic consequences test shows that the reduction effect of financial distress on the capital expenditure of non-distressed enterprises in the same city ultimately improves their capital expenditure efficiency.

Imitation learning which uses only expert demonstrations is suitable for safety-crucial tasks, such as autonomous driving. However, causal confusion is a problem in imitation learning where, with more features offered, an agent may perform even worse. Hence, we aim to augment agents’ imitation ability in driving scenarios under sequential setting, using a novel method we proposed: sequential masking imitation learning (SEMI). Inspired by the idea of Granger causality, we improve the imitator’s performance through a random masking operation on the encoded features in a sequential setting. With this design, the imitator is forced to focus on critical features, leading to a robust model. We demonstrated that this method can alleviate causal confusion in driving simulations by deploying it the CARLA simulator and comparing it with other methods. The experimental results showed that SEMI can effectively reduce confusion during autonomous driving.

Occlusion handling is a key technical issue in augmented reality research. This paper proposes a new occlusion algorithm based on object contour detection to address issues such as poor real-time occlusion processing, high computational complexity in comparing the depth values of virtual and real objects, and the presence of jagged, blurry, and hollow edges in occluded areas. First, based on the depth and color information, we obtained aligned images of real scenes. Second, we extracted the maximum closed contour of the real object in the scene and overlaid it with the aligned image. Subsequently, we generated a virtual object and obtained a depth map of the virtual object. Finally, by comparing the depth values of the stacked images with the virtual objects, masks are generated in real time and optimized to present the occlusion processing results. Experimental comparisons demonstrated that the algorithm presented in this study not only improves real-time performance but also enhances accuracy at the intersection edges of virtual and real images. Simultaneously, it is no longer limited by the size of real scene images and can achieve real-time virtual and real occlusion effects.

Down syndrome is a lifelong cognitive impairment characterized by lower mental skills and intelligence quotient (IQ) compared to their typical peers. The profile is not curable. However, research has been conducted to supplement and improve cognitive functioning through computing and software applications. Conventional cognitive applications and IQ scales pose significant challenges as they are not developed based on specific cognitive guidelines. Therefore, such methods often fail to accurately assess cognitive profiling, resulting in uncertainty. To overcome these challenges, Interactive Mental Learning Activity Software utilizes tailored guidelines incorporating fuzzy logic rules, ensuring accurate cognitive profiling for Down syndrome. Fuzziness was applied within the machine learning framework across three groups structured based on IQ levels. A total of N=200 individuals with Down syndrome participated in the IQ assessment. The findings revealed that individuals with mild impairment demonstrated a higher degree of improvement in cognitive abilities compared to moderate and severe levels. However, the severe category appears to have an unrealistic probability, leading to a standstill in progress. The implementation of the specific guided system led to improvements of 6%, 5%, and 5% in individuals with mild, moderate, and severe cases, respectively.

Factor market distortion hinders economic growth, and digital economy may provide an impetus to help alleviate the misallocation. Based on China’s provincial panel data from 2013 to 2021, this study applies a dynamic evaluation method to measure the comprehensive level of digital economy, and constructs a spatial econometric model to investigate its impacts on factor market distortion, as well as the differences before and during the COVID-19 pandemic. The empirical findings indicate that China’s digital economy development shows a steady upward trend at the provincial level, has a spatial spillover effect on factor market distortion, and plays a dissipating role, which embodies the potential of digital economy in improving the optimal allocation of resources. At the same time, there is a phenomenon of unbalanced digital economy development in different areas, reflecting the “digital divide” problem. Moreover, during the COVID-19 pandemic, various mobility restrictions were launched, which promotes continuous penetration of the digital economy in various fields of the economy, and its effects on alleviating the factor market distortion are manifested. Overall, this study provides some enlightenments to better implement digitalization strategies.

This paper investigates a robust detection for demagnetization fault in the permanent-magnet synchronous motor (PMSM). A proportional-integral fading sliding-mode observer (PIFSMO) is proposed, which can solve the problem of PMSM performance degradation caused by the permanent-magnet demagnetization and resistance disturbance increase during long-term operation. A PMSM demagnetization model is established, and an improved PIFSMO is proposed to detect the demagnetization flux linkage. A fading integral term is added to the traditional sliding-mode observer, which can suppress the offset in the resistance disturbance estimation caused by transients, thereby achieving direct estimation of the resistance disturbance and the slowly varying resistance disturbance (e.g., ḋ≠0) in particular. Simulation and experimental results demonstrate the feasibility and effectiveness of the proposal.

As the real estate industry expands with time, the personalized needs of users for indoor space layouts have become increasingly complex. Traditional indoor space layout design methods can no longer meet the needs of large market groups because of their complex steps and low levels of specialization. Therefore, this study first analyzes the problematic factors in indoor space layout design. Second, an interactive genetic algorithm is introduced to solve the multifactor optimal selection problem; the process is optimized and improved using a differential evolution algorithm. A comprehensive spatial layout model combining interactive genetic and differential evolution algorithms is proposed. The experimental results show that the model performs best with uniform variation, and its average number of iterations to find the optimal individual is 57. In addition, compared with similar layout models, the proposed model achieved the highest space utilization value of 79%, which is approximately 19% higher than that for the stacking layout model; it also required the shortest time, that is, 15 min. In summary, the proposed model provides a new intelligent method for indoor layout design, which is expected to improve the satisfaction of designers and users.

The widespread use of automobiles has revolutionized transportation and attracted a large population owing to their convenience and effectiveness. However, this widespread adoption has resulted in a significant increase in road traffic accidents. The alarming road fatalities suggest that medical responders are overwhelmed by the need to save lives in a timely manner. This is due to a lack of affordable autonomous detection and notification mechanisms. Prior work in this domain includes the use of vehicular ad hoc networks, Arduinos, and Raspberry Pis; machine-learning approaches for predictions; and smart devices using integrated sensors. These methods are either expensive to acquire, human-reliant, or require vehicular modifications. Therefore, the aim of this study is to suggest a cheap prototype that can work with smartphones. The prototype should have embedded micro-electromechanical system (MEMS) sensors that measure g-force to find car accidents and global system for mobile communications-long term evolution (GSM-LTE) to call the closest medical responders, which would be found using GPS. A prototype was developed using the .NET Multi-Platform App UI (MAUI) framework. This study applied the design science research methodology (DSRM) to produce a socially acceptable, low-cost artifact similar to existing in-vehicle systems to save lives on the road during a road traffic accident. The FEDS evaluation of the results indicated that smartphones can perform such complex tasks with reasonable accuracy compared with expensive in-vehicle systems. The prototype can be adopted by lower- to middle-class individuals as it is a cheaper alternative. This study makes a practical contribution to the society by utilizing artifacts to ensure road safety.

Electricity consumption pattern recognition is the foundation of intelligent electricity distribution data analysis. However, as the scale of electricity consumption data increases, traditional clustering analysis methods encounter bottlenecks such as low computation speed and processing efficiency. To meet the efficient mining needs of massive electricity consumption data, in this paper a parallel processing method of the density-based k-means clustering is presented. First, an initial cluster center selection method based on data sample density is proposed to avoid inaccurate initial cluster center point selection, leading to clustering falling into local optima. The dispersion degree of the data samples within the cluster is also used as an important reference for determining the number of clusters. Subsequently, parallelization of density calculation and clustering for data samples were achieved based on the MapReduce model. Through experiments conducted on Hadoop clusters, it has been shown that the proposed parallel processing method is efficient and feasible, and can provide favorable support for intelligent power allocation decisions.

This paper focuses on policy selection and scheduling of sensors and attackers in cyber-physical systems (CPSs) with multiple sensors under denial-of-service (DoS) attacks. DoS attacks have caused enormous disruption to the regular operation of CPSs, and it is necessary to assess this damage. The state estimation of the CPSs plays a vital role in providing real-time information about their operational status and ensuring accurate prediction and assessment of their security. For a multi-sensor CPS, this paper is different from utilizing robust control methods to characterize the state of the system against DoS attacks, but rather positively analyzes the optimal policy selection of the sensors and the attackers through dynamic programming ideology. To optimize the strategies of both sides, game theory is employed as a means to study the dynamic interaction that occurs between the sensors and the attackers. During the policy iterative optimization process, the sensors and attackers dynamically learn and adjust strategies by incorporating reinforcement learning. In order to explore more state information, the restriction on the set of states is relaxed, i.e., the transfer of states is not limited compulsorily. Meanwhile, the complexity of the proposed algorithm is decreased by introducing a penalty in the reward function. Finally, simulation results show that the proposed algorithm can effectively optimize policy selection and scheduling for CPSs with multiple sensors.

The precise detection of falls is essential for promptly providing first aid to individuals who are at risk of accidental injury. Presently, the predominant approach for detecting falls is through inertial measurement unit (IMU) sensors, which can capture the real-time motion of an object. However, it is difficult for the current approach to face the challenges in attaining the anticipated performance in real-world applications, owing to the diverse nature of human behavior. To tackle this concern, a fall detection approach that uses a graph convolutional neural network (GCN) with variable time windows (T-GCN) is introduced. The proposed method uses well-designed graph topologies to effectively mitigate the impact of inconsistent data dimensions. Meanwhile, variable time windows are designed to capture keyframe data and to enhance their validity. To evaluate the effectiveness of the T-GCN method, a dataset D_hard containing 12 suspected falls and four real falls is built. The experimental results show that the T-GCN method achieves an accuracy of 91.3% and a precision of 92.5%, surpassing the average accuracy and precision of conventional fall detection methods.

In this study, model predictive control (MPC) was considered for a class of discrete time-delay systems with external disturbances. Equivalent input disturbance (EID) is a commonly utilized active disturbance rejection technique that enhances the rejection of system disturbances. Any disturbance can be effectively suppressed using the EID approach, leaving no traces of the disturbance. This paper proposes a novel MPC control approach that combines the EID method with the MPC principle. An MPC law with disturbance-rejection performance was presented after obtaining and combining the EID estimates with performance indicators using the EID method. Optimizing the performance index is then converted into a semi-definite problem comprising of an objective function and linear matrix inequality through building a Lyapunov function. Based on this basis, an MPC design algorithm is proposed. A numerical simulation was conducted to confirm superiority and efficacy of the proposed method.

In the context of the new era, teachers and students in colleges and universities, as well as the general public, rely more on the Internet and social media to obtain news, express their opinions, and share information, and the dissemination of public opinion events in colleges and universities is not only related to the physical and mental health of teachers and students but also to the reform and development of colleges and universities. In this study, we took campus public opinion events as the main research object in which we selected three recent campus public opinion events to be analyzed. The public opinion data used in the research was collected from Weibo social media platforms. Firstly, we analyzed the dissemination cycle and regional dissemination patterns of a college food safety public opinion hot event through the popularity and regional distribution of public opinion data, thus revealing its formation and evolution patterns. Secondly, the LDA topic mining method is used to mine the themes of the three hot public opinion events, and then analyze the hot factors of the dissemination of each public opinion event from the massive public opinion data. This is crucial for the management department to grasp the dynamics of public opinion. Then, we used the SKEP sentiment classification method to analyze the emotional factors of the public opinion data of the three events to obtain the overall public opinion sentiment situation of the events. Finally, based on the characteristics of time, region, and gender, the evolution and diffusion rules of public topics and emotional distribution under different types of events are analyzed. The precision of the analyses associated with this paper may be limited to the effects of current mainstream as well as state-of-the-art analytical models. The analysis methods and conclusions in this paper provide a scientific theoretical basis and improvement measures for campus public opinion management, which helps to enhance the level of campus public opinion management and safeguard campus stability and public order.

At present, product family design has become an important link in enterprise development and manufacturing. Optimization ideas and technologies are important foundations and core frameworks in product family design. Previous research on product family design has mainly been limited to optimization problems within the product domain. As an important influencing factor in the product family design process, the supply chain not only affects the cost level of the product family in the back-end of the design process, but also affects the modular structure layout of the product family in the front-end of the design process. Therefore, the optimization of the correlation between supply chain and product family design process is a crucial issue that determines the success or failure of product families. However, when researching the personalized needs of users in product family design and configuring product modules, there is very little consideration given to the optimization of supply chain correlation. To address the aforementioned issues, this article develops supply chain oriented product design optimization decision-making method based on improved CUR matrix decomposition. Firstly, based on the customer’s functional requirements C matrix and module relationship R matrix, perform customer clustering and corresponding product configuration. Then, utilizing the numerical stability of orthogonal trigonometric decomposition (QR), U matrix is constructed, which represents the inherent relationship between functional requirements and module relationships. Secondly, based on quality/character requirements, functional module levels division and initial supplier configuration are carried out. Finally, determine the supplier configuration for each module with the goal of maximizing total profit. Analyze the customer selection, classification, and product configuration process of a contractor as a case study. The research results indicate that the optimization decision method based on improved CUR matrix decomposition can effectively obtain the optimal solution of the decision problem.

This study analyzes the readability and sentiment of central bank communications across six Central and Eastern European countries. It reveals considerable variability in readability, with Moldova being the most accessible and Serbia the most complex. Notably, readability declined during the 2020 COVID-19 pandemic, reflecting the urgent and complex nature of economic communication. The study finds no direct correlation between readability and sentiment; however, the net hawkishness index significantly correlates with business cycle phases, suggesting its potential to forecast monetary policy shifts. This study underscores the intricate relationship between central bank communication, readability, sentiment, and economic conditions, advocating for enhanced clarity in central bank communication. It also highlights the importance of domain-specific sentiment analysis for interpreting and predicting the implications of monetary policy communication, providing valuable insights for policymakers and market participants.

Pure pursuit algorithm is commonly used in path tracking control of autonomous vehicle for its high real-time performance. Due to the problem of “taking shortcuts,” traditional pure pursuit algorithms usually have the problem of low path tracking accuracy in curved road scenarios. To address the issue, a path tracking control method based on improved pure pursuit algorithm is proposed. This method builds upon traditional pure pursuit theory and dynamically adjusts the look-ahead distance based on vehicle speed and road curvature radius information, allowing it to adapt to different road scenarios. This effectively addresses the problem of large path tracking errors in curved road scenarios. Furthermore, a fuzzy feedback control is employed to compensate for control variable and enhance tracking accuracy across various scenarios. Simulations and real-world experiments demonstrate that the proposed method significantly improves path tracking accuracy compared to traditional pure pursuit methods, particularly in curved road scenarios. The maximum lateral deviation is reduced by over 50%, realizing the precise tracking of autonomous vehicle on the park roads.

Based on nonlinear memory feedback, the robust admissibility problem of a class of singular nonlinear time-delay systems was investigated. First, an equivalent structural model of the system was created using the state decomposition approach. Then, sufficient criteria for the robust admissibility of the system were obtained in the form of linear matrix inequalities using the Lyapunov–Krasovskii theory and free weighted matrix method. Subsequently, the closed-loop robust admissibility of the system under nonlinear memory feedback control was investigated using similar research techniques as before, yielding corresponding results. The intended design controller was obtained by explicitly computing each component in the deconstructed structure of the nonlinear memory feedback controller. Importantly, the flexibility and viability of the planned control design could be enhanced by the algorithm for solving controller gain through this component decomposition. Finally, numerical examples confirmed the feasibility of the method.

The compound directional drilling process, including slip and rotation, is the key to realizing long-hole drilling in coal mines. First, a compound directional drilling process control scheme is proposed to realize intelligent control of the directional drilling process. An intelligent drilling process optimization method was designed to improve drilling efficiency. A robust controller for the drilling rate based on gain scheduling was designed to achieve stable control of the drilling rate during sliding deflection drilling. A robust controller based on hybrid sensitivity was designed to achieve stable control of the drilling rate during rotary inclined stabilization drilling. The results of this study can provide a theoretical basis for realizing the intelligent optimization and control of compound directional drilling in coal mines.