At present, drilling fluid leakage in oil and gas drilling engineering in complex formations is a worldwide technical problem. The study explains the mechanism of dense pressure-bearing plugging at the bottom of the fracture, explores the influencing factors of the pressure-bearing capacity of the leakage prevention and plugging working fluid, and establishes a mathematical model by using multivariate nonlinear regression analysis. Based on the machine learning technology, the support vector machine algorithm is selected as the prediction method of the particle size of the working fluid for leakage prevention and plugging, and the system model of the ultra-high-temperature dense pressurized leakage prevention and plugging working fluid is constructed. It is found that the established multivariate nonlinear regression analysis has good fit and accuracy, and the average relative error is only 2.9%, and the seam width (-0.694) and formation pressure (0.502) have the greatest influence on the pressure-bearing capacity of the working fluid for leakage prevention and plugging. The prediction accuracy of the support vector machine model for the working fluid particle size was 95.36%, and the prediction F1 values on multiple datasets were all greater than 0.9, showing excellent prediction results. The constructed mathematical model can be used to guide the field operation, which is conducive to the long-term stable plugging and scientific leakage prevention of fissure-based leakage.

Chinese oil painting art is an important carrier of contemporary Chinese cultural identity features, the identification and quantitative study of the color and texture of the picture can help to understand the characteristics of the oil painting works more deeply. Therefore, this paper proposes a feature recognition method for oil painting art based on deep learning method. The Otsu threshold method and DeeplabV3+ network model based on DeeplabV3+ are selected for image graying and segmentation processing. The global color histogram and ring LBP are used to extract the color and texture features of the picture respectively, and the oil painting feature recognition is completed based on the regularized limit learning machine. In several sets of quantitative results, the methods in this paper all have better oil painting color and texture feature recognition, among which the RELM algorithm has the highest detection accuracy at low correlation features. It shows that the deep learning based Chinese oil painting art and cultural identity feature recognition method can effectively extract oil painting features and realize the quantitative research on oil painting.

The article firstly establishes a mathematical model of the FMS shop floor planning process problem, and combines the rescheduling strategy and rolling scheduling strategy for solving the FJSP problem. Subsequently, the simulated annealing genetic algorithm is improved by relying on genetic algorithm, simulated annealing algorithm and particle swarm optimization algorithm, and the application of hybrid optimization algorithm in problem solving is proposed. The simulated annealing algorithm is incorporated into the crossover and mutation operations of the genetic algorithm to strengthen the local search capability, and then the global annealing operation is incorporated into the new individuals obtained. The overall design of the mixed reality-based FMS virtual simulation system was tested with a view to optimizing the external tool library tool limitation problem in the FMS shop floor planning process. The results of the simulation experiments show that although the algorithm of this paper, SaDE and CoDE algorithms can reach the optimal solution, the convergence speed of the algorithm proposed in this paper is significantly better than the other two algorithms. Based on the experimental results, the article finally constructs a mixed reality-based FMS virtual simulation system to solve the external tool library tool limitation problem in the FMS shop floor planning process.

Agent technology is widely used in intelligent manufacturing and digital workshop as a new method to solve complex, dynamic and distributed artiϐicial intelligence application problems. This paper ϐirstly summarizes the application steps of Agent technology in 3 aspects of modeling, simulation and monitoring of intelligent manufacturing system on the basis of a brief description of multi-agent system. Then, based on reinforcement learning theory, a multi-agent collaborative algorithm SRL_M3DDPG based on state representation learning is proposed.Finally, the algorithm model is tested and applied to the smart shop scheduling problem. The learning curve of the SRL_M3DDPG algorithm in the example remains relatively stable after the 3400th round, and the maximum completion time of the scheduling is 29. Comparing with other composite scheduling rules, the delay rate of this paper’s algorithmic model is the lowest, which is only 15.47%, which indicates that the algorithm is able to signiϐicantly reduce the delay rate of the workpiece. In addition, this paper’s algorithm achieves better results in adaptive intelligent manufacturing workshop scheduling, ϐinding the shortest machining completion time of 221 unit time, which can adapt to the dynamic intelligent manufacturing workshop environment.

This paper constructs a comprehensive evaluation system based on the CIPP model, covering multiple dimensions such as input evaluation and outcome evaluation, in order to comprehensively measure the effect of college students’ mental health education in the new media environment. In terms of weight determination, the subjective weights are obtained by hierarchical analysis method, then the objective weights of each index are calculated by entropy value method based on the actual data, and then the combination assignment method is used to organically combine the subjective and objective weights to obtain the ϐinal indexes. The relationship matrix was constructed on the basis of a large amount of collected data, and the fuzzy comprehensive evaluation method was used to comprehensively assess the implementation effect of college students’ mental health education. The results of the study show that the overall level of the effect of college students’ mental health education is good, with the ratings of 79.54 and 78.28 for their mental health knowledge and ideological awareness evaluation, respectively, and that the mastery of mental health methodology and the awareness of proactively seeking psychological help are the main factors affecting the mental health of college students. In addition, the mastery level of college students’ mental health practice ability is average (69.52), and there is an obvious deϐiciency in their theory to practice, which also adds difϐiculty to the construction of college students’ mental health. Therefore, the fuzzy comprehensive evaluation method can be used to optimize the evaluation system of college students’ mental health education in the new media environment.

With the rapid development of artificial intelligence technology, the research on personalized learning in the field of ideological and political intelligence education is increasingly active. In this paper, an improved locust optimization algorithm is proposed, which is applied to the intelligent grouping strategy of ideological and political education. Then a knowledge state-oriented hypergraph self attention knowledge tracking model is proposed, which consists of a hypergraph module and a self attention module, and is capable of predicting students’ future interaction sequences through their past interaction sequences. In order to realize students’ personalized test question matching needs, a Civics test question recommendation algorithm based on the neural graph model is proposed, based on which a personalized Civics test question recommendation exam system is designed and implemented. The intelligent grouping strategy based on the optimized locust algorithm achieves a total score accuracy of 100% in the Civics grouping task. The knowledge tracking model accurately predicts students’ knowledge status, and the attention weights of students’ learning paths based on this paper’s recommendation algorithm are all higher than 0.5. It shows the effectiveness of this paper’s strategy of automatic generation of Civics education content based on the locust optimization algorithm and the personalized test question matching model on the students’ in-depth understanding of the Civics knowledge and improvement of learning efficiency.

In the face of the traditional agricultural marketing model is difficult to continue the status quo, agricultural marketing competition – cooperation relationship for agricultural enterprises of commodity marketing and long-term development is also increasingly important. In this paper, game theory is introduced into the study of competition and cooperation strategy of agricultural products marketing, the strategic behavior of two agricultural products enterprises in the agricultural products industry cluster is constructed into the corresponding matrix, and evolutionary dynamic stability analysis is carried out to establish the replication dynamic equations and Jacobi matrix to solve the evolutionary stability strategy (ESS), so as to provide reference for the formulation of the competition and cooperation strategy of the enterprise’s agricultural products marketing. Using simulation to explore the influencing factors of the evolutionary direction of the marketing competition and cooperation strategy of agricultural products enterprises. When the probability of winning the joint bidding is greater than 0.8, it will evolve into a cooperative strategy, and when it is less than 0.7, it will evolve into a competitive strategy, and with the increase of the allocation coefficient of the investment amount of the project construction, the agricultural products enterprise 1 and the agricultural products enterprise 2 will gradually shift from a competitive strategy to a cooperative strategy. The lower the cost allocation coefficient is, the higher the probability that enterprises will evolve to cooperative strategy. The increase of cooperative transaction cost then accelerates the evolution of enterprise 1 and enterprise 2 to competitive state.

The traditional Chinese culture contacted in history education has many common points with the Civic and Political Education, which has become a new method of value penetration of Civic and Political Education. This paper reveals the value penetration of traditional Chinese culture in Civic and political education from the perspective of innovative cultural topology, and puts forward three strategies to innovate the concept of Civic and political education, such as enhancing the effect of aesthetic connotation of Civic and political education. On this basis, variables are designed, structural equation model is constructed, and the role of teaching concept and other variables on the value penetration of traditional Chinese culture Civic and political education is analyzed through the reliability test and factor analysis. Combined with system dynamics, the system causality diagram is drawn according to the causal feedback relationship between internal and external factors to explore the causal relationship affecting the value penetration of Civic and Political Education, and then explain the mechanism of the role of traditional Chinese culture Civic and Political Education. It was found that all five paths among latent variables passed the significance level test of 0.001, and the teacher’s mission and ideal belief in teaching philosophy had the most significant effect on the value penetration of traditional Chinese culture Civic and political education, with path coefficients of 0.98. In the process of Chinese traditional culture civic education, it is necessary to reflect the unity of humanistic spirit and modern spirit, the unity of professional ethics and values, and to form the style of course civic education and course civic education characteristics with Chinese traditional culture.

At present, machine translation performs better in the general domain translation effect of large-scale bilingual corpus, but the translation effect in specific domains still needs to be improved. In order to optimize the accuracy of machine translation in the domain of English translation of professional terms, this paper proposes a translation model that incorporates syntactic knowledge and terminology. Aiming at the problem of more limited translation domain knowledge in the RNMT and Transformer models based on the self-attention mechanism, an optimization method is proposed. According to the domain characteristics of English translation of professional terms, English syntactic keywords are incorporated into the model training process, the special information contained inside the text of professional terms is learned, and the lexical properties of each word in the dataset are recognized before they are input into the model. Then attempts are made to incorporate the specialized terminology into the model to enrich the parallel corpus required by the model. The experiments confirm the excellent performance of the optimized translation model in this paper on the De→En terminology translation task, which improves 22.67 BLEU values compared to the base model. And the fluctuation of its BLEU value with the change of sentence length is small, which further indicates that the method optimizes the accuracy of the machine translation model in the English translation of professional terms.

Aiming at the needs of reconstructing the structure of calligraphic seal cutting strokes and virtual display, this study designs a GAN technique that integrates three models, namely, “WGAN, DCGAN and CGAN”. The Cycle GAN model is used to obtain the mapping relationship between learning and style migration by utilizing its cyclic consistency loss. Adaptive pre-morphing technique is introduced to process the input image to capture the outline information and morphological features of calligraphic seal carvings, and a Generative Adversarial Network-based Generative Model for Structural Reconstruction of Calligraphic Fonts (CRA-GAN) is proposed. Meanwhile, an online virtual display system is designed to provide users with a good sense of experience in the virtual display of calligraphy. The results show that the CRA-GAN model can better capture the details and global information of the fonts, and its recognition rate of the eight calligraphic fonts ranges from 90.42% to 97.38%, and the MOS rating value of the text image is > 8.5 points, and its recognition results are in line with the observation characteristics of the human eye for calligraphic images. The FID calculation result of the CRA-GAN method ( 204.361) of the CRA-GAN method is much lower than that of other methods, which obviously improves the diversity and visual quality of the generated calligraphic fonts. This paper evaluates the user’s experience of the system from five aspects: narrative experience, emotional experience, sensory experience, cognitive experience and interactive experience, and calculates that the final score of the system is in the range of 80-100, which indicates that the user’s satisfaction is very high after actually experiencing the virtual display system.