With the increase in emergency incidents, research on pedestrian dynamics has received widespread attention. Bidirectional pedestrian flow has been identified as a major factor in numerous global disasters over the past few decades. Conflicts can arise between individuals moving in opposite directions. Additionally, under different emergency conditions, evacuation behavior can vary significantly. This study primarily examines the impact of proactive avoidance and urgency on bidirectional pedestrian flow.

Experiments were conducted by establishing an outdoor experimental setup and recruiting participants. A controlled variable method was used to design 17 experimental trials to investigate the movement of bidirectional pedestrian flow under different levels of urgency and the presence of proactive avoidance behavior. Variations in urgency and proactive avoidance were implemented through a monetary reward and penalty mechanism. Pixel coordinates were first recorded using the Tracker software and then converted into real-world coordinates using direct linear transformation. The analysis focused on the evacuation speed, evacuation time, movement trajectories of the evacuees, maximum offset during the evacuation process, and behavior exhibited throughout the evacuation. Moreover, variance analysis was conducted to assess the stability of the experimental conditions and the validity of the experiments. Finally, a questionnaire was administered to collect some basic parameters of the participants, such as height, weight, and age. Furthermore, information on whether the participants experienced feelings of anxiety during the experiment and their perspectives on whether the provision of additional rewards influenced their sense of urgency was obtained.

Common phenomena observed during the evacuation process included overtaking, following, and side-by-side behavior. An innovative behavior termed boundary flow acceleration was identified, where individuals moving against the flow exhibited significantly higher walking speeds when adhering to the wall compared with those not using the wall for support. Notably, groups of individuals moving against the flow spontaneously formed lanes when faced with a larger number of pedestrians moving in the opposite direction, which enhanced their ability to navigate through the crowd. In emergencies, the evacuation speed of individuals with disabilities was slower than that of normal pedestrians. Therefore, individuals with disabilities need more attention during emergencies. In nonemergency situations, the average evacuation speed of individuals moving against the flow was 1.26 m/s, which increased to 1.49 m/s when proactive avoidance behavior was present. In emergencies, the average speed of individuals moving against the flow increased to 2.04 m/s. When no individuals were moving against the flow, the evacuation speed was 1.77 m/s in nonemergency situations and 3.50 m/s in emergencies. The questionnaire results indicated that approximately 82.9% of the participants experienced anxiety during the evacuation process, and 91.4% of the participants believed that providing additional rewards could enhance the urgency of the experiment. This finding validated the effectiveness of the monetary reward mechanism in simulating emergency conditions.

The results indicated that proactive avoidance behavior effectively reduced pedestrian conflicts. Moreover, the behavior improved the evacuation efficiency of individuals moving against the flow by 24.60% and increased the evacuation speed by 19.87%. Furthermore, the phenomenon of stratification became more pronounced with the introduction of proactive avoidance behavior. Compared with nonemergency situations, the evacuation speed of the larger crowd significantly increased compared with that of the smaller crowd during emergencies. These findings can provide valuable insights for managing bidirectional pedestrian flow in emergencies and offer relevant support for modeling research in this area.

The arc fault of the low-voltage distribution system is one of the primary causes of residential fires. Due to the diverse load types and complex connection methods in residential areas, arcs fault exhibit many similar and concealed characteristics, making them difficult to detect. This frequently leads to issues with arc fault protection devices, such as false alarms and missed detections. The conventional detection method, which is based on manually extracting arc fault feature vectors, is incomplete and heavily relies on expert-designed features. Consequently, this impedes the development of highly generalizable models. In addressing these challenges of multiload systems, this paper proposes a method for serial arc fault diagnosis and load recognition based on a one-dimensional dilated convolutional neural network (1D-DCNN).

First, a series of experiments on multiload arcs fault are conducted using a custom-designed experimental platform. This platform supports both single-load and dual-branch load conditions during testing. Normal and faulty current data on the main bus are sampled under various operating conditions at a rate of 500 kHz. During the data processing phase, the continuous time series data are discretized and normalized based on the half-cycle length. Subsequently, a 1D-DCNN is used to extract features from the high-sampling-rate arc fault current data. Furthermore, the scaled exponential linear unit activation functions and residual connections are introduced to address the challenges of gradient vanishing and network degradation. Moreover, the cyclic padding method is adopted to alleviate boundary effects and enhance the model's robustness to dataset shift. The arc fault detection model is developed by integrating average ensemble learning with a Softmax multiclassifier. Precision, recall, and specificity are used to assess the efficiency of the model. Finally, the accuracy of the proposed model in load classification, load state recognition, and overall accuracy is compared with that of other classical models, providing a comprehensive assessment of its efficacy.

The findings of this method were as follows: (1) The accuracy of arc detection using a recurrent neural network was considerably low, primarily due to gradient vanishing and exploding gradients, making it difficult to effectively train the model. (2) Under the condition of equal parameter count between a 1D-DCNN and 1D-CNN, the dilated convolutional operation expanded the receptive field, resulting in greater accuracy than the 1D-CNN model. (3) The proposed method achieved a remarkable accuracy of 99.67% in detecting arc faults for both single-load and mixed-load scenarios, with an accuracy of 99.95% and an overall accuracy of 99.62%.

This research presents a unique model capable of autonomously learning features from high-sampling-rate current data without requiring manual feature extraction. It efficiently detects arcs fault while identifying the type of faulty load simultaneously. The model outperforms typical convolutional neural networks on validation of the test set, thereby meeting the requirements for arc fault identification. This advancement has major implications for serial arc fault detection and load recognition applications.

The negative impact of COVID-19 pandemic has hindered the development of urban agglomerations. Because of close geographical, transportation, and economic ties, COVID-19 is more likely to be transmitted repeatedly in urban agglomerations. This paper provides references for coordinating the development and security of urban agglomerations and building "resilient cities" in the context of pandemic. By constructing a system of indicators for the urban development level and using an autoregressive integrated moving average (ARIMA) model to predict the urban development level without pandemic, this study can quantitatively assess the impact of pandemic on individual cities in the Beijing-Tianjin-Hebei urban agglomeration.

The study applied an ARIMA model to investigate the urban development mechanism in urban agglomerations in different stages of pandemic. First, indicators were selected from multiple sources based on the collection and analysis of literature. Their reliabilities were tested with the Cronbach's coefficients. Second, the indicators were assigned weights using an integrated method with the analytic hierarchy process (AHP) and the entropy weight method. Third, the stages of the COVID-19 pandemic were divided based on the monthly data collected from Weibo and other websites. Fourth, based on historical data and urban development trends before pandemic, the ARIMA model was used to predict the urban development level without the effect of pandemic. Finally, a comparison analysis was conducted between the prediction value and the real value to quantitatively assess the impact of pandemic on individual cities in the urban agglomeration.

(1) In the context of pandemic, the urban development level indicators of three cities reached peak and trough values in the same month. (2) The degree of influence was less than 0 during the outbreak period and gradually decreased to a stable trough value. (3) The degree of influence was greater than 0 in the early stage of the recovery period and gradually decreased to less than 0 in the later stage until it reached the trough point.

This study shows that: (1) the COVID-19 pandemic in the central city of the urban agglomeration affects the formulation and implementation of the overall urban agglomeration development strategy; (2) the development pattern of urban agglomeration converges because of pandemic; and (3) cities are resilient and have a certain disaster-bearing capacity. To strengthen the construction of the Beijing-Tianjin-Hebei urban agglomeration, the paper suggests that the government should start from the economy, transportation, people's livelihood, and disaster resilience to improve the urban development level.