Aiming at the problems of false detection, missing detection and imbalance of positive and negative samples in video abnormal behavior detection, a multi-modal feature fusion method for abnormal behavior detection is proposed. Firstly, a cross-modal sensing module is designed, which uses the cross-attention mechanism for feature fusion to improve the expression ability of cross-modal data features, and reduces the number of network parameters by sharing parameter strategies. Then, the improved binary cross entropy loss function is used to train the network. In the training process, the weight is reduced dynamically for the easily distinguishable samples, and the larger weight is focused on the difficult to distinguish samples, which improves the processing ability of unbalanced and difficult to classify data and improves the detection accuracy of abnormal behavior. Finally, through the strategy of sample batch selection, more abnormal fragments are filtered out by statistical analysis method to effectively solve the problem of missing abnormal fragments selection. Tests were conducted on XD-Violence, Shanghai-Tech open data set and self-made data set. The AP value of XD-Violence data set reached 85.32%, and the AUC value of Shanghai-Tech data set and self-made data set were 96.84% and 81.73%, respectively. Experimental results fully prove the effectiveness and generalization ability of this method.

Accumulated water film on the pavement is unevenly distributed due to several factors, which may lead to an aircraft veer-off accident. The single index of water film thickness lacks the ability to characterize the water film distribution condition. In this paper, a theoretical model of the deviation behavior of aircraft hydroplaning was established. A hydroplaning simulation of an aircraft wheel group was carried out on the basis of the American full-scale test result of pavement loading. The transverse water film uneven distribution was quantified section by section. The differences in the frictional effect between both sides of aircraft wheel group were compared. It was suggested to use a new type of evaluation index called the critical length of unevenly distributed water film. The rationality and applicability of such an index were verified through a correlation study between critical length and aircraft taxing performance variation by using ADAMS-based kinematic analysis. The result indicated that notable differences exist between the time-history curve of the frictional effect between both sides of the wheel group in the case that the aircraft lands on the pavement with a transversely unevenly distributed water film. The entire landing procedure raised the likelihood of a veer-off mishap because asymmetric deviation forces constantly impact the aircraft's taxing performance. The critical length of uneven distributed water film derived from three study cases were merely 6.6% differed from theoretical results. The proposed evaluation index may take the accumulative impact of deviation forces due to transversely unevenly distributed water film into consideration, which may provide useful reference for take-off and landing safety protection and water distribution control on the pavement.

To rapidly predict the ascent trajectory of launch vehicles, high-accuracy analytical solutions for ascent trajectory were proposed under high angle-of-attack (AOA) conditions. First, a simplified longitudinal-plane dynamics model with mass as the independent variable was developed. The sine of AOA was innovatively expressed as a polynomial of mass. Due to the high AOA, the simplified dynamics model remains highly nonlinear, which prevents it from being analytically solved directly. Approximate polynomials are introduced to replace the strongly nonlinear but relatively small terms in the original equations through force analysis. Furthermore, the difference between the true values and the approximations was treated as a minor perturbation. The dynamics model was divided into analytically solvable subsystems according to the perturbation theory. Analytical solutions for velocity, flight-path angle, downrange, and altitude were derived by solving the subsystems. Simulation results confirm that the proposed solutions are at least 85% more accurate than existing solutions under high AOA conditions.