Current research generally focuses on the application of gamification to improve the engagement of learning. However, the research on gamification in specific fields such as color education is not sufficient, and there is a lack of analysis on the gamification elements and influence factors of learning effects. For these problems, a game model for training color recognition was designed. Firstly, two different ways of playing were designed with same core gameplay but different interaction modes. Then, the same virtual reward was added in both playing ways. Finally, the effects of two playing ways on learning effect with or without virtual reward were compared, and the effect of virtual reward in the same playing way were compared. The results show that gameplay design mainly affects learning efficiency, and virtual reward mainly affects engagement.

To improve the bad performance of crowd counting in surveillance videos and images caused by the scale and perspective variation, a crowd counting model, named Multi-scale Multi-column Convolutional Neural Network (MsMCNN) was proposed. Before extracting features with MsMCNN, the dataset was processed with the Gaussian filter to obtain the true density maps of images, and the data augmentation was performed. With the structure of multi-column convolutional neural network as the backbone, MsMCNN firstly extracted feature maps from multiple columns with multiple scales. Then, MsMCNN was used to generate the estimated density map by combining feature maps with the same resolution in the same column. Finally, crowd counting was realized by integrating the estimated density map. To verify the effectiveness of the proposed model, experiments were conducted on Shanghaitech and UCF_CC_50 datasets. Compared to the classic methods:Crowdnet, Multi-column Convolutional Neural Network (MCNN), Cascaded Multi-Task Learning (CMTL) and Scale-adaptive Convolutional Neural Network (SaCNN), the Mean Absolute Error (MAE) of MsMCNN respectively decreases 10.6 and 24.5 at least on Part_A and UCF_CC_50 of Shanghaitech dataset, and the Mean Squared Error (MSE) of MsMCNN respectively decreases 1.8 and 29.3 at least. Furthermore, MsMCNN also achieves the better result on the Part_B of the Shanghaitech dataset. MsMCNN pays more attention to the combination of shallow features and the combination of multi-scale features in the feature extraction process, which can effectively reduce the impact of low accuracy caused by scale and perspective variation, and improve the performance of crowd counting.

To deal with shortcomings of a single node or the simple neighbor relations in the research of cascading failures, a cascading failure model was proposed considering the local characteristics of node-community structure. The model gave each node dynamic initial load value based on the community property of the node, and adopted different strategies to attack the Western States Power Grid of the United States, US Air lines, IEEE118 standard grid and ScaleF-ree Network (SFN) to simulate the process of cascading failures. The simulation results show that these nodes within community lead to relative minor faults when community factor dominated in initial load, but some special nodes connecting multiple communities will cause serious cascading failures. It also indicates that the property of the number of neighbor nodes is more relevant than other properties by calculating Pearson correlation coefficients of different properties.

Concerning the difficulties in solving the continuous functions of general Harmony Search (HS) algorithm, an improved HS algorithm was proposed. With analogies to the concept of the simulated annealing algorithm, the way of updating parameter was redesigned. And it limited the number of identical harmonies stored in the harmony memory to increase the diversity of solutions. Simulation results of the proposed algorithm were compared with other HS approaches. The computational results reveal that the proposed algorithm is more effective in enhancing the solution quality and convergence speed than other HS approaches.

Based on input layer,latent layer and output layer correlation rule,the structure learning method of stochastic fuzzy neural network considered the effects on the latent layer function,which was very important for engineering applications.One key problem of this algorithm is how to get the best node number of latent layer.In this paper,a normal method to confirm the best node number was proposed.According to this theory,one simulation was presented.

The real-time property of the serial communication program for UAV(Unmanned Aerial Vehicle) flight controller was studied. The hard real-time serial receiving program which dealt with totally four different data frames was designed with four techniques: data frame picking-up in interrupt serve program, finite state machine theory, buffer sharing and code optimizing. Test results show that the program has reasonable time consumption and ideal hard real-time property. Meanwhile, it has an outstanding data frame picking-up ratio with high reliability and satisfies the requirement to control the UAV craft by serial communication.