Mental health, which has as equally important effects on people’s life as physical health, is receiving more and more attention nowadays, especially with a significant increase of pressure brought by the fast-paced evolution of technology and society. The diagnosis of mental health symptoms, however, mostly relies on the interpretation of languages and behaviors by experienced psychologists, who are not accessible for the great population. Depression causes cognitive and motor changes that affect speech production: reduction in verbal activity productivity, prosodic speech irregularities, and monotonous speech have all been shown to be symptomatic of depression. In this study, we aim to provide a deep learning-based model that could give an initial diagnosis of mental health problems for individuals and screen the risk of developing mental health issues. This AI-driven model focuses on the understanding and analysis of people’s daily public comments/posts and captures the peoples’ mental health status embedded in the semantic and syntactic structure in those online posts.

Airfoils are one of the most important factors in determining an airplane’s ability to fly. As such, their optimization holds important ramifications for the continual need to improve airplane performance. In this paper, the optimization of airfoil design is explored through numerical computational fluid dynamics. This paper first investigates the effects of Angle of Attack on airfoil lift/drag ratios, which is shown to have a significant influence on the performance of airfoils. Then, the effects from thickness are explored, and it is found that the influence of thickness is highly dependent on the specific shape of airfoil. In addition, in order to further improve airfoil aerodynamic performance, a type of modified airfoil with an additional tip is developed and simulated. The results indicate that airfoils with additional tips can perform better, but only with the proper combination of tip parameter values. For example, the modified airfoil with a tip of 0.03 L length at a 25° tip angle bested that same airfoil without an additional tip. Finally, it is found that the Backpropagation machine-learning algorithm, when trained with prior simulation data, can quickly predict lift and drag coefficients of airfoils. The results in this paper could facilitate the further optimization of airfoil design.

浮空器悬停或飞行控制设计时,通常将风作为干扰项或阻力,螺旋桨需要消耗大量的能源克服风的阻力,而浮空器携带的能源有限,因此浮空器设计时存在欠能源问题.根据平流层风场的特点,采用副气囊控制浮空器的高度,实现平流层不同高度的风场利用,可以减小浮空器的能源消耗.建立浮空器高度控制模型,采用反步法设计浮空器高度控制器,利用状态观测器对浮空器的模型误差和输入误差进行估计,并进行了仿真分析,证明所设计的控制器能有效控制浮空器的高度.建立浮空器高度控制时其囊体内外压差变化模型,仿真分析表明,浮空器高度变化后,其囊体最终内外压差与初始压差相同.

In this paper, a performance prediction method is proposed for the design of a stratospheric propeller. The Spalart–Allmaras (S–A) model was used to calculate the airfoil performance of FX63, and the polynomial fitting method was utilized to establish the airfoil database of the lift and drag coefficient. A computational fluid dynamics (CFD) model was applied at different altitudes to prove the feasibility of the method. The CFD results were compared with the results of the vortex theory and prediction; the prediction result accuracy was improved compared with that of the vortex theory over a greater range of advance ratios. The airfoil performance data requirements and the number of iterative calculations were reduced. These results indicate that the proposed propeller design meets the requirements of stratospheric airship propulsion systems.

Path planning is one of the key technologies for autonomous flight of Unmanned Aerial Vehicle. Traditional path planning algorithms have some limitations and deficiencies in the complex and dynamic environment. In this article, we propose a deep reinforcement learning approach for three-dimensional path planning by utilizing the local information and relative distance without global information. UAV can obtain the limited environmental information nearby in the actual scenario with limited sensor capabilities. Therefore, path planning can be formulated as a Partially Observable Markov Decision Process. The recurrent neural network with temporal memory is constructed to address the partial observability problem by extracting crucial information from historical state-action sequences. We develop an action selection strategy that combines the current reward value and the state-action value to reduce the meaningless exploration. In addition, we construct two sample memory pools and propose an adaptive experience replay mechanism based on the frequency of failure. The simulation experiment results show that our method has significant improvements over Deep Q-Network and Deep Recurrent Q-Network in terms of stability and learning efficiency. Our approach successfully plans a reasonable three-dimensional path in the large-scale and complex environment, and has the perfect ability to avoid obstacles.in the unknown environment.

According to the requirements of a foreign high-altitude airship project, the Single-Rotation-Propeller (SRP) was designed under the maximum endurance factor criterion. The spacing influence on the Contra-Rotation-Propeller (CRP) was studied by numerical simulation analysis, the performance of front and rear propeller was compared with SRP. The result indicated that CRP was more efficient than SPR, the CRP has enormous potential in enhancing the aerodynamic performance.

According to the requirements of a foreign high-altitude airship project, the Single-Rotation-Propeller (SRP) was designed under the maximum endurance factor criterion. The spacing influence on the Contra-Rotation-Propeller (CRP) was studied by numerical simulation analysis, the performance of front and rear propeller was compared with SRP. The result indicated that CRP was more efficient than SPR, the CRP has enormous potential in enhancing the aerodynamic performance.

Target detection and tracking can be widely used in military and civilian scenarios. Unmanned aerial vehicles (UAVs) have high maneuverability and strong concealment, thus they are very suitable for using as a platform for ground target detection and tracking. Most of the existing target detection and tracking algorithms are aimed at conventional targets. Because of the small scale and the incomplete details of the targets in the aerial image, it is difficult to apply the conventional algorithms to aerial photography from UAVs. This paper proposes a ground target image detection and tracking algorithm applied to UAVs using a revised deep learning technology. Aiming at the characteristics of ground targets in aerial images, target detection algorithms and target tracking algorithms are improved. The target detection algorithm is improved to detect small targets on the ground. The target tracking algorithm is designed to recover the target after the target is lost. The target detection and tracking algorithm is verified on the aerial dataset.

A safe trajectory generation method for quadrotor flight in three-dimension dense environments is proposed in this letter. We propose a path search algorithm called Safe A* to search a safe, collision-free initial path in cluttered environments. Subsequently, a safe and smooth trajectory with minimum snap is obtained by solving a quadratic problem in closed form. Our method is tested in various simulations, results turn out that our method can significantly increase the success rate of trajectory generation, and generate a safer trajectory as compared to the standard A* algorithm.

As a sampling-based pathfinding algorithm, Rapidly Exploring Random Trees (RRT) has been widely used in motion planning problems due to the ability to find a feasible path quickly. However, the RRT algorithm still has several shortcomings, such as the large variance in the search time, poor performance in narrow channel scenarios, and being far from the optimal path. In this paper, we propose a new RRT-based path find algorithm, Fast-RRT, to find a near-optimal path quickly. The Fast-RRT algorithm consists of two modules, including Improved RRT and Fast-Optimal. The former is aims to quickly and stably find an initial path, and the latter is to merge multiple initial paths to obtain a near-optimal path. Compared with the RRT algorithm, Fast-RRT shows the following improvements: (1) A Fast-Sampling strategy that only samples in the unreached space of the random tree was introduced to improve the search speed and algorithm stability; (2) A Random Steering strategy expansion strategy was proposed to solve the problem of poor performance in narrow channel scenarios; (3) By fusion and adjustment of paths, a near-optimal path can be faster found by Fast-RRT, 20 times faster than the RRT* algorithm. Owing to these merits, our proposed Fast-RRT outperforms RRT and RRT* in both speed and stability during experiments.

During the long term of flight task, to the structure of wing fuselage, the aircraft will cause multi-site damage due to fatigue load and other factors, which will reduce the residual strength of the structure rapidly and result in serious flight accidents. It is very important to do research on the reliability of multi-site damage. on the basis of single crack, the interaction between multi-site cracks is introduced, and the theory of single crack is extended to the propagation of multi-site cracks, calculating the stress intensity factors and interaction factors in various modes of multiple cracks.[1],[2]

In this article, experiments and numerical simulation on the tensile loading properties of three-dimensional five-directional braided composites are studied. Herein, axial tensile loading tests were performed on braided composite and aluminum joints for propeller. A repeated unit cells model was established to analyze the micromechanical properties of the braided joint, a progressive damage model was established to evaluate the damage in the joint. In contrast to aluminum joint, which exhibited a tensile failure mode in only one direction, the braided joint exhibited both longitudinal and transverse damage failure modes, of which the transverse damage was the primary failure mode. The load-displacement curves and damage morphology of the numerical simulation were in a good agreement with the experimental results.

In order to solve the problem that the existing reinforcement learning algorithm is difficult to converge due to the excessive state space of the three-dimensional path planning of the unmanned aerial vehicle, this article proposes a reinforcement learning algorithm based on the heuristic function and the maximum average reward value of the experience replay mechanism. The knowledge of track performance is introduced to construct heuristic function to guide the unmanned aerial vehicles' action selection and reduce the useless exploration. Experience replay mechanism based on maximum average reward increases the utilization rate of excellent samples and the convergence speed of the algorithm. The simulation results show that the proposed three-dimensional path planning algorithm has good learning efficiency, and the convergence speed and training performance are significantly improved.

This paper presents the control scheme of a pseudo-satellite in situations where error constraints, uncertainties, and external disturbances occur. The control scheme of a pseudo-satellite includes planar path following control scheme and altitude control scheme. In order to control the altitude of a pseudo-satellite, the backstepping technique is used to control the quality of air in the air balloon. To deal with the error constrained requirements of a pseudo-satellite’s position, a tan-type barrier Lyapunov function (TBLF) is proposed and incorporated with the guidance control schemes. Adaptive fuzzy disturbance observers are presented to estimate uncertainties and external disturbances. Rigorous stability analysis proves that the altitude error of the pseudo-satellite can be uniformly bounded, and the position error of which can be maintained within the range of constrained requirements, while all closed-loop signals are uniformly bounded regardless of uncertainties and disturbances.

Path-following controller of a stratospheric satellite is designed under circumstances of wind currents utilization, parameter uncertainty, input value errors, and external disturbances. In order to reduce the energy consumption of the stratospheric satellite, the stratospheric east (west) wind currents are utilized by the stratospheric satellite to track the desired path. To utilize the wind currents, an air balloon as well as altitude control model are designed. Fuzzy disturbances observers are presented to estimate uncertainties and external disturbances. Input values are quantized by logarithmic quantizers. Rigorous stability analysis proves that path-following errors of the stratospheric satellite can be reduced to zero, and all closed-loop signals are uniformly bounded regardless of uncertainties and disturbances. Simulation results show that the stratospheric satellite can track the desired path well and the proposed control strategy can reduce the energy consumption of the stratospheric satellite effectively.

Path planning is important in robot field and in this field, many researchers have done a lot of work. This paper proposes an improved ant colony algorithm as traditional ones have a shortage of slowly convergence and easily falling into local optimum. On the basis of traditional ones, the dynamic random statistical analysis and extraction of each generation of Ant Colony are performed the optimal, average and worst ant information constitutes an adaptive operator for adaptive updating of local pheromones. Simulation results demonstrate that it is effective in equilibrium increasing convergence rate and getting into the contradiction of local optimal solution.

This research deals with the coverage path problem (CPP) in a given area with several known obstacles for agriculture Unmanned Aerial Vehicles (UAVs). The work takes the geometry characteristics of the field and obstacles into consideration. A practical method of the coverage path planning process is established. An obstacle avoidance path planning is used to find a coverage path for agriculture UAVs. The method has been tested with an Android application and is already applied in reality. The results turn out that the method is complete for this kind of coverage path planning problem.

随着大数据、云计算、物联网、移动互联网等信息技术的迅猛发展与广泛应用,新的作战模式不断涌现,以任务分布式指控流程为核心的云作战成为一种全新的跨域全维作战样式.在分析作战云与云作战特征的基础上,结合传统作战仿真流程提出了云作战体系仿真流程,并提出了云作战构造型仿真平台框架的总体方案设计与系统功能设计.通过云作战构造型仿真示例,对比了传统作战样式与云作战样式的观察—调整—决策—行动(OODA)循环,结果表明,云作战样式能够有效缩短OODA循环时间.

Considering the entire skin coating of aerostat envelop materials as one material,the effect of absorption and emission rate of aerostat envelop materials on helium temperature differences between day and night is investigated.In order to further reduce helium temperature differences between day and night,in this paper,aerostat envelop materials are divided into illuminated side with materials of low absorption rate and backlight side with materials of high emission rate.Under the established thermal analysis model,material properties in different parts of aerostat envelop materials are optimized with the method of Kriging model.It holds the thoughts that aerostat envelop materials can be divided into 48 parts,Latin hyper-cube method is used to do sampling,and sample response can be obtained through thermal analysis so as to build a Kriging approximate model.As the result,it shows that the helium temperature difference between day and night is reduced to 28.6 K,which is 7.7％ less than the traditional ways of analysis.

An autonomous vehicle landing control algorithm of a quadrotor is investigated for the situation when the quadrotor hovers above the vehicle in this paper. To facilitate the controller design, the problem of autonomous landing is converted from general trajectory tracking problem of a quadrotor to a stabilization problem of relative motion. A four-degrees-of-freedom (4-DOF) nonlinear relative motion model with four control inputs is estimated. An adaptive radial basis function neural network (RBFNN) is developed to estimate the unknown disturbance and is applied to design the controller through a backstepping technique. It is proved that all the states in the closed-loop system are uniformly ultimately bounded and the error converges to a small neighborhood of origin. Numerical simulation results illustrate the good performance of the proposed controller.