With the arrival of the third wave of development of Artificial Intelligence, Artificial Intelligence has entered a new stage, and the research of Emotional Intelligence is of great significance to the development of Artificial Intelligence. Emotional Intelligence has a wide range of application prospects in Intelligent Robots, Intelligent Virtual Assistants and other fields. Emotional Intelligence quantifies people's psychological state through Computational Psychophysiology, and Emotional Intelligence can enable machines to better recognize and understand human emotions, and produce better emotional interaction with users. The concepts of Emotional Intelligence and Computational Psychophysiology are expounded, the opportunities and key problems in the development of Emotional Intelligence analyzed, and the typical applications of Computational Psychophysiology in physiological signal acquisition, biological information feedback intervention and emotion recognition introduced based on EEG, speech, eye movement, expression and posture. Emotional Intelligence and Computational Psychophysiology have a high recognition rate in emotion recognition, which can use a variety of modals for emotion recognition, and have a wide range of development and application prospects in many fields such as medical treatment, education, entertainment, production and so on.

The ultimate goal of constructing the database of the knowledge graph should be the isomorphism of human brain knowledge growth, and it should not be a passive solidified storage body but an actively growing hatching body. The idea of intelligent reproduction library is proposed, which is an updated version of the Knowledge graph, a multi-level embedded graph database system. Each concept node is a window. Open the window, and there is a sub-library produced. All concepts are coded hierarchically for fast query. A concept node is called a bud if the concept has multiple factors taking it as domain. Open the window of the bud, and a cause and effect analysis table is prepared, on which, factor space theory can perform algorithms to make intelligent judgment, reasoning, evaluation, decision-making and planning, by which the database becomes a living intelligent incubator. The Factorial Query Language (FQL) is also established, which may be used in the preparation of the intelligent incubation civil intelligent engineering, with guidance of the mechanistic general artificial intelligence theory.

The rapid development of mobile information technology has further innovated the forms of advertising in social media. Interactive social media advertising has won the favor of enterprises. Focusing on the unique interactive behavior and psychology of users and based on the hierarchy of effects model and cognitive-developmental theory, a model of user empathy expression and purchase intention is constructed and verified by 337 valid samples. The research results show that two social psychological factors (fear of missing out and social influence) and two advertising content factors (content quality and source credibility) significantly positively affect users’ advertising cognition; that users' advertising cognition is passed ultimately affects empathy expression and purchase intention through affecting advertising attitude, and that users’ empathy expression intention significantly affects purchase intention. Guidance for companies is provided to conduct interactive social media advertising, and research directions and suggestions put forward for mobile social media interactive advertising research.

A closure algorithm based on KD45 logic is proposed. It can improve group role assignment and solve the problem of human-machine collaboration via task assignment under the constraints of cooperation and conflict, when cooperative information is in opacity. The main contributions include: through KD45 logic, the collaboration and conflict relationship set is improved, and the group role assignment model is improved; A KD45 closure algorithm is proposed to improve the set of constraint relations; IBM ILOG CPLEX optimization package (ILOG) is used to conduct tens of thousands of simulation experiments under different scales. The experimental results show that it is feasible and effective to improve the group role assignment constraint relation set by using KD45 logic closure algorithm, and then the improved group role assignment constraint relation set is applied to the optimal group role assignment method. This group role assignment method can effectively solve the problem of man-machine collaborative task assignment under the constraints of collaboration and conflict.

To study the influence of different behaviors of operators and managers in similar engineering management systems on the system level profit, an analysis method of system profit based on space fault network (SFN) and game theory is proposed. Firstly, the significance, definition and relationship of operators and managers in the system are discussed. The feasibility of SFN representing game process is shown. The basic parameters of the method, the relationship between the basic parameters and the event comprehensive profit, the game logic relationship and the system profit are determined. Operators have safe and unsafe behaviors, and managers have punishment and reward behaviors. After the operators and managers act on the event comprehensive profit, the system profit is finally obtained through and/or logical evolution, and the winner of the profit game is judged. Finally, the flow, function and applicability of the method are illustrated by an example. A method is provided for studying the process of system profit game.

Based on the nature of opinion leaders, the opinion mining technology is used to study the identification of microblog opinion leaders in emergencies in order to provide a reference for the governance of online public opinions. A three-step opinion leaders identification framework is proposed: first, using a literature analysis, an indicator model is built to evaluate the influence of microblog bloggers; second, a text subjective and objective classification model is built to determine whether high-influence bloggers publish event-related opinions and information, and to identify opinion blog posts; then an emotional polarity classification model is built based on the comment text of the opinion blog post to calculate the support degree of the blog post opinions and the support rate of the blogger’s opinions so that those high-influence bloggers who have first output their opinions and then received more support can be rated as the event "Opinion Leaders". These methods are used in the following text to analyze the microblog public opinion data of the typical practical cases, identifying the microblog opinion leaders of the incident, with their characteristics and public opinion participation behaviors observed and analyzed, and the results compared with those of social network analysis methods and the expert manual analysis method to verify the scientificity and effectiveness of this method. The results show that this method can identify microblog opinion leaders who truly have the guiding power of public opinions and can have governance value in the evolution of public opinions in emergencies.

Through the study of feature levels in convolutional neural networks, this paper found that high-level feature have stronger semantic information and low resolution, and low-level features have strong resolution and weaker semantic information. Aiming at these problems, a object detection algorithm based on secondary feature fusion is proposed. The algorithm reuses transitional features and performs secondary feature fusion on the basis of Feature Pyramid Networks to supplement the rich low-level feature information to the top. Finally, the average accuracy of AP, AP₅₀, and AP₇₅ on the COCO2014 data set reach 35.3%, 57.5%, and 36.6%, respectively. Compared with the unused feature fusion method and the traditional feature fusion method, the average accuracy is increased by 2.4%, 3.7% and 2.4%, which significantly improves the missed detection and the detection of small targets.

Physical-layer security is an effective technique to achieve secure communication in wireless communication. Physical-layer security in an unmanned aerial vehicle (UAV) communication system is investigated, where a UAV base station transmits confidential information to a legitimate receiver in the presence of a potential eavesdropper. The existing works usually assume that the flight environment of the UAV is stable, and the corresponding trajectory optimization methods are offline, which are not able to adjust the trajectory according to changing environment. To solve this issue, online trajectory optimization for the UAV base station is studied to ensure the information transmission security and maximize the communication secrecy rate. Different from the existing works, a more accurate hybrid light-of-sight/non-light-of-sight channel model is adopted for the air-to-ground communication links and an online trajectory optimization method proposed based on the Q-learning algorithm to ensure the physical-layer security of UAV communication. Numerical results show that the proposed algorithm significantly improves the secrecy rate of the UAV communication system, as compared with two benchmark schemes.

A new direct SLAM (Simultaneous Localization And Mapping) system with enhanced stereo-camera constrains based on Stereo Direct Sparse Odometry (Stereo DSO) is presented. As a direct SLAM method, any image pixel with sufficient intensity gradient can be utilized, which makes it robust even in featureless areas. Two-stage checking combining SAD (Sum of Absolute Differences) with NCC (Normalized Cross Correlation) matching method is used for stereo matching and triangulated uncertainty concerned to accelerate the convergence of depth filters. To estimate the accurate scale of the environment, static stereo constrains are added to the tracking module. Our evaluation on KITTI demonstrates that the proposed system achieves the better performance than the state of the art direct SLAM systems such as LSD-SLAM2 and Stereo DSO, and achieves the comparable performance with ORB-SLAM3 which is the state of the art feature SLAM. The proposal provides mobile robots with a new direct SLAM system to explore the environment more precisely and robustly.

The development of fitness industry puts forward requirements for the improvement of thermal comfort levels of sports buildings and the thermal satisfaction needs of different gender groups. Three kinds of moderate-activity exercise levels are selected, and experiments in the climate chamber conducted. During the experiments, the subjects regulate the air temperatures according to self-thermal comfort demands. Then the indoor air temperature measurement and the statistics of human physiological parameters and subjective questionnaire voting are also conducted. Thus thermal comfort characteristics between male and female in different activities are discussed. Results show that: at the same activity level, there are no significant differences for these multiple physiological parameters and subjective questionnaire votes, except for the higher heart rate of female. At the paces of 4, 5 and 6 km/h, the preferred temperatures of female are respectively 1.6, 0.3 and 0.9 ℃ higher than the male. The thermal comfort for female is relatively lower than that for male for the same thermal sensation and also expresses more sensitivity to the thermal comfort variations. On the basis of these results, the spatial environmental parameters aiming at different activity levels are proposed based on the respective thermal demands of different gender, which would provide theoretical evidence for constructing suitable indoor environment of sports buildings.

Road marking material is a kind of marking material that transmits traffic information such as guidance, restriction and warning to road traffic participants, which plays an important role in controlling and guiding traffic. Reflective road marking materials are widely used in road engineering in recent years. The reasonable combination of titanium dioxide, glass beads and other additives can significantly improve the night reverse reflection brightness coefficient of road marking materials, improve the guiding efficiency of road marking materials for driving at night, and effectively reduce the incidence of traffic accidents at night. However, the existing road reflective marking materials lack reasonable technical standard requirements, especially the core material of retroreflection performance glass beads lacking perfect and unified evaluation index system and limit value requirements, which directly affects the retroreflection performance, durability and construction quality of road reflective marking. Based on this, the reflective road marking is taken as the research object, starting from the classification of road marking materials, the composition and characteristics of different types of road marking materials are analyzed and the influencing factors of the retroreflective performance of existing road marking materials investigated. Based on the retroreflective performance of glass beads, the requirements of domestic and foreign standards for glass beads related technical indices are comprehensively analyzed and the common proportion and particle size composition range of glass beads defined. The future development of reflective road markings is prospected.

A numerical simulation of a two-stage traveling-wave thermoacoustic engine with reducers is investigated. The results showed that the variation of different inner diameter, length, and position of the reducer affect the temperature difference and output performance of heat exchanger between two thermoacoustic engines. When both thermoacoustic engines operate under the same constant power, the reducers affect the temperature difference and output performance of the system by adjusting the phase at different positions and inner diameters that result in low-temperature difference and high energy output.

The addition of waste stone powder to fibre mortar or fibre concrete can improve its mechanical property, but the fresh property would be decreased. In order to explore the governing factors of the fresh property of the polypropylene fibre mortar with the waste stone powder replacing the equal volume of the standard sand, 20 fibre mortar mixes with different water cement ratios and different volume contents of the waste stone powder are produced for flowability, cohesiveness, adhesiveness and packing density measurement. Based on the packing density results, the average liquid thickness (ALT) and average paste thickness (APT) of the fibre mortar mixes are calculated. The results showed that the ALT and APT have good correlation with the test results of flow rate, flow spread, cohesiveness and adhesiveness. According to the results of regression analysis, it is found that the ALT is the major factor governing the flow rate and adhesiveness, and the ALT and APT are the major factors governing the flow spread and cohesiveness.