This paper takes the native vegetation in Hanzhong City as the research object, and constructs a multiobjective linear programming model to optimize the distribution of the suitability of the native vegetation in Hanzhong City. The ArcGIS software was used to test the sample consistency and screen the environmental variables of the native vegetation data in Hanzhong City represented by alfalfa, and the model in the software was used to predict the distribution of alfalfa’s suitability area. Based on the prediction results, this paper constructs a multi-objective linear planning model with economic and ecological benefits as the objective function and the land area of different utilization types as the decision variables to optimize the distribution of the suitability of native vegetation in Hanzhong. At the same time, the fuzzy mathematical planning method was used to solve the constructed model. After the model optimization, the area of fitness distribution of native vegetation in Hanzhong City increased significantly, and the growth of the fitness distribution area of each vegetation by 2080 was 49.61%, 35.51%, 36.41%, 28.11%, 15.36%, 24.75%, 27.92%, 28.40%, 31.22%, and 31.52%, respectively. In addition, the optimization of the distribution of native vegetation suitability using the model of this paper can produce obvious economic and ecological benefits, which fully demonstrates the effectiveness of the model of this paper.

Intelligent thermoregulation clothing as a new type of functional clothing, the design and development of which is receiving more and more attention. PID algorithm, as a kind of classical control algorithm, realizes the precise control of the clothing temperature regulation system by adjusting the three parameters of proportionality, integration and differentiation. The control system is firstly constructed according to the principle of PID control. Then the PID controller parameters are optimized by BP neural network to improve the response speed and stability of the temperature control system. Finally, the intelligent thermoregulation garment with physical therapy and health care and portable storage is designed. Experimental verification of the parameter self-tuning PID control based on BP neural network, the BP neural network can make the temperature better maintained near the set value, the control effect is more satisfactory. The final design of the smart thermoregulation garment has a body surface temperature retention rate of 98.35% after 30 minutes at -10°C and with the heating function on. The thermal sensation evaluation of the intelligent thermoregulation garment by the subjects in different states is concentrated between “0-2”, indicating that the garment can play a more ideal temperature control effect.

Since the introduction of fractal geometry, it has set off a wave of research in the scientific community, and it has been widely used in many fields. This paper firstly introduces the landscape modeling and generating technology based on fractal geometry, and proposes the virtual landscape generating method based on fractal geometry through the study of the regular characteristics of fractal geometry. Combined with the game development of virtual landscape generation diversity, complexity needs, in the fractal Brownian motion model on the basis of the proposed optimization of the generation process for game development. In the simulation experiments of virtual landscape generation, the NME value of virtual landscape generation under the method of this paper is the smallest, which is distributed between 3 and 6, and the generation time is reduced by 31ms and 38ms compared with the average time of the traditional generation method and the SEM method, which shows that the designed virtual landscape generation is able to generate the virtual landscape more realistically. The study concludes with strategies and recommendations for the application of fractal geometry to virtual landscape generation in game development, with a view to contributing to the promotion of virtual generation technology.

The argument of the article comes from the rapid development of digital technology and the urgent need for the digital protection and restoration of traditional paper horse art. For this reason, this paper proposes a method of digital protection and restoration of traditional paper horse art based on graphics processing technology. The traditional paper horse art image is collected, the image is denoised using mean filtering, the paper horse image is decomposed in gray scale through spatial conversion, and then its double histogram equalization is processed to obtain the color-enhanced image. Combined with the convolutional image restoration strategy, the paper horse art is digitally displayed. The method of this paper can enhance the color of the paper horse art image and retain the original details, and at the same time, in terms of the clarity effect, the method of this paper improves the comparison method by 25.27%~339.39%. In addition, the method in this paper has better image restoration quality with subjective evaluation rating ≥ 4 and higher PSNR and SSIM. What’s more, the scores on the evaluation dimension of digital preservation and restoration effect ranged from 4.02 to 4.48, and the overall effect performance was relatively good.

Under the dual background of the construction of the “new liberal arts” and the digital wave, the interdisciplinary practice of combining humanities and technology continues to develop. Taking a number of Chinese language and literature works as examples, this paper selects language features from the vocabulary and sentence levels, analyzes the syntactic structure of the selected Chinese language and literature works with the help of natural language processing technology and numerical measurement method of language features improved TF-IDF method, and realizes the discussion of the lexical categories of literary works, such as word length, word frequency, word class distribution and word density, as well as the study of sentence categories such as average sentence length, sentence dispersion and sentence class distribution. It is found that most of the utterances of the selected literary works are monosyllabic words and polysyllabic words, the cumulative proportion of both of them is more than 90%, the highest frequency of occurrence is nouns and verbs, both of them are more than 22%, the average sentence length and sentence dispersion do not differ much, and the overall readability of the selected literary works is better, with a free change of syntactic structure and a stronger narrative of the text.

Teacher-student interaction, as the most important way of classroom interaction, its level directly affects the quality of classroom teaching. The study selected three English listening classes, three English reading and writing classes, and three English exercise classes, totaling nine English classes in a university for video recording. With the help of the Improved Flanders Interaction Analysis System (iFIAS), the study utilized classroom observation and multiple regression analysis to investigate the effectiveness of teacher-student interactions in the classroom and their influencing factors. It was found that the average value of students’ classroom discourse ratio (40.3%) was smaller than the average value of teachers’ classroom discourse ratio (48.1%), and that a reasonable structure of teacher-student language ratio was more conducive to the formation of benign interactions in the classroom and the enhancement of the overall classroom effectiveness. In addition, teaching ability, learning style, learning motivation and classroom environment all positively affect the effectiveness of English teachers’ classroom interaction in colleges and universities. Therefore, it is necessary to start from these four aspects to adjust the language ratio structure, create a positive classroom atmosphere, and enhance the integration of information technology and the classroom.

The traditional English teaching mode in colleges and universities has many problems in cultivating students’ language ability. This paper introduces information technology into task-based English teaching in colleges and universities and constructs a task-based English teaching mode based on SPOC technology. With the orientation of improving students’ language ability, it implements the improvement of English teaching mode in colleges and universities. Using principal component analysis to comprehensively evaluate the relevant indicators of students’ language proficiency in the process of task-based English teaching in colleges and universities, and quantify the effect of the combination of information technology and task-based English teaching on the improvement of students’ language proficiency. Ten classes of students majoring in English in a university were selected and divided into experimental and control groups, and the data related to students’ language proficiency were collected and analyzed at the end of the experiment. The data were downscaled using principal component analysis, and the principal components were extracted according to the eigenvalues and cumulative contribution rate. The comprehensive score of students’ language proficiency is calculated by the comprehensive evaluation function of students’ language proficiency constructed in this paper. The language proficiency of students in the experimental group and the control group is significantly different after the experiment, and the comprehensive scores of students in the experimental group are 53.96% and 61.96% higher than those before the experiment, respectively. It reveals that the introduction of information technology into task-based teaching of English in colleges and universities has a significant effect on the enhancement of students’ language proficiency.

In educational research, more and more scholars recognize the importance of teaching interaction network for learning, and they find that “interaction” is not only the method of learning, but also the learning process itself. Social network analysis provides a new way to study teaching interaction. Through the study of social network analysis, this paper proposes the construction method of teaching interaction network for physical education. In this paper, we take four real physical education courses in L school as the research object to conduct in-depth research, obtain the physical education classroom teaching interaction behavior data, and construct the teaching interaction network. The results of the study show that in the interaction network of the four physical education teaching courses, the teaching behaviors of the community network of physical education classroom 1 are significantly concentrated in B3, B4, B5, and B6, course 2 is concentrated in B4, B5, B6, B9, and B10, the teaching interactive behaviors of physical education classroom 3 are significantly concentrated in B2~B6, and the significant physical education teaching interactive behaviors of course 4 are concentrated in B2, B4, B5, and B6.From the degree-centeredness analysis, there are 33 marginal learners with the number of stored interactions less than or equal to 2 in physical education teaching interactions, which indicates that in this paper’s study of physical education teaching interactions, teachers do not pay enough attention to teaching interactions in a comprehensive way. By summarizing the theoretical basis and practical significance of teaching interaction and social network analysis, it proves that the network construction of teaching interaction in this paper is effective, and at the same time, it also provides a new idea for physical education teaching courses.

In order to strengthen the construction of network security defense system and effectively respond to new types of threat attacks appearing in the network environment, this paper constructs a network security threat prediction model using data mining algorithms. The network security threat posture needs to be assessed before the security threat prediction. Accordingly, this paper assesses the four security threat postures of services, vulnerabilities, weaknesses, and hosts on the basis of the quantitative assessment method of hierarchical security threat posture. After that, a network security threat prediction model is constructed based on the support vector mechanism, and a genetic algorithm is used to optimize the parameters of the model. The three evaluation index values of MAE, RMSE and MAPE for the GA-SVM-based cybersecurity posture prediction method proposed in this paper are 0.0106, 0.0133 and 0.0222, respectively, which are better than those of the ABC-SVM-based and PSO-SVM-based prediction methods. It indicates that the method in this paper has smaller error and higher accuracy in cyber security posture prediction. This shows that the method in this paper usually achieves better accuracy in cyber security threat posture prediction.

In this paper, we understand the shortcomings of the current mainstream IoT privacy protection methods through analysis, and in this way, we propose an evolutionary and signaling game model for IoT privacy protection. The model analyzes the stabilization trend of IoT platform penalty coefficients on privacy protection and provides protection strategies. Combining the implications of the signaling game model, the degree of IoT privacy protection is measured using the Bayesian equilibrium solving algorithm. Simulation experiments are conducted to evaluate the specific effect of the model on IoT privacy protection. The increase in the detection rate of the model accelerates the convergence of the probability of malicious nodes, e.g., when the detection rate increases from 0.7 to 0.9, the convergence time is reduced by about two stages. The larger the penalty amount of the IoT platform, the model recommends more aggressive protection strategies, and the probability increases from 0.16 to about 0.4. The game parameters of the model reflect the malicious behavior in IoT, and the trust level affects the game parameters. The model in this paper reduces the attack gain by 4% to 10% compared with the comparison model when the fixed defense gain is 1500, which can better reflect the influence of protection signals on the attacker’s actions.