In this paper, the non-chiral boundary of the mixed kdv-mkdv equation is transformed into a chiral boundary by the construction of auxiliary functions, and a new linear difference format is constructed for the chiral boundary problem. Based on the traditional difference format, explicit and implicit differences are used alternately to construct a class of explicit-implicit (E-I) and implicit-explicit (I-E) alternating difference formats, and the unconditional stability of the numerical solutions is proved by taking advantage of the symmetric discrete numerical advantage of this class of alternating difference formats. The exact solution of the kdv-mkdv equation and its dynamical behavior are explored in the calculations using the semi-fixed separation of variables method combined with the phase diagram method for planar dynamical systems. Various types of exact solutions of the equations are obtained under special parametric conditions, and the existence problem of isolated wave solutions of the kdvmkdv equations is analyzed in conjunction with the exact solutions of the equations. Numerical examples verify the accuracy and feasibility of the constructed differential format, indicating the existence of isolated wave solutions for the KdV-mKdV equation.

The construction of university English teaching resources is an inevitable requirement to adapt to the development of the times and educational reform. Based on the concept of knowledge and classification, this paper puts forward the theory of Rough set, and applies the idea of partition to the data simplification based on Rough set. Based on the applicability of the partition strategy, the partition idea is added in the process of attribute simplification to achieve the purpose of reducing the complexity of the data simplification algorithm about Rough set. After deriving the decision table, the attribute approximation algorithm based on the attribute order and the partition method is given, i.e., the efficient knowledge approximation method based on the partition method for Rough set. Analyze the performance of Rough set efficient knowledge reduction method based on partitioning method in multiple datasets. To build a knowledge acquisition system platform for university English teaching resources using the efficient knowledge reduction method based on the Rough set of the partition method. In the Heart dataset, the classification accuracies of DIDS method, IV-FS-FRS method, and this paper’s method are 0.5936, 0.5536, and 0.6689, respectively, and this paper’s method outperforms the classification accuracies of DIDS method, IV-FS-FRS method 0.0753, and 0.1153, respectively. The knowledge acquisition system platform of university English teaching resources constructed by using this algorithm has operational advantages in instance analysis.

Based on the common problems of the original fuzzy testing technique and the needs of RESTful API fuzzy testing, this paper proposes a white-box fuzzy testing method of REST API based on graph resource nodes for RESTful API software interface testing by using EvoMaster as a basic tool. The effectiveness of the fuzzy testing technique in this paper is analyzed. 21 apps with millions of downloads obtain more than 65,000 web request data and more than 8.5GB HAR files, and an average of 2,966 web request data is collected for each app. The REST interface filtering method of this paper’s fuzzy testing approach effectively and accurately targets interface objects for fuzzy testing. The number of generated requests of the REST API white-box fuzzing test method based on graph resource nodes in this paper is much lower than that of other tools, and the efficiency of vulnerability discovery is much higher than that of other tools. The test method in this paper improves the number of lines of code covered in six hours by an average of 53.86% over other tools. The test method in this paper can identify more vulnerabilities and can cover all the vulnerabilities found.

Inadequate writing skills can prevent learners from improving their writing performance and interfere with their subsequent writing performance in authentic scenarios. The article’s research focuses on the effects of metacognitive regulation on students’ authentic writing performance in a web-based constructivist learning environment, which relies on constructivist learning environments to better present the authentic writing problems learners face in their studies and lives. In this paper, we adopt the method of randomized group sampling to conduct a single-group pre-test and post-test experiment on 40 students in a public high school. It also chooses students’ writing learning achievement as the dependent variable, and students’ metacognitive regulation level and writing selfefficacy as the independent variable and mediator variable, respectively, and explores the degree of influence of metacognitive regulation level on students’ writing learning achievement through multiple linear regression. The results showed that there was no significant difference between pretest 1 and pretest 2, while posttest 1 and posttest 2 were much higher than pretest 1 and pretest 2. There was a significant positive effect of students’ level of metacognitive regulation on students’ learning achievement in writing (0.459), and there was a significant mediating effect of students’ writing selfefficacy between students’ level of metacognitive regulation and students’ learning achievement in writing. Relying on the web-based constructivist learning environment can significantly enhance students’ metacognitive regulation level and provide a new teaching path to promote students’ writing learning achievement.

In the context of artificial intelligence technology, the current academic research on the relationship between exosome molecular screening and the role of gene probes is relatively weak. Accordingly, this paper formulates a modeling study of exosome molecule screening and its relationship with gene probes under the framework of multi-objective genetic algorithm. The multi-objective genetic algorithm is applied to realize the screening of secretory body molecules, and after the completion of the screening work, the mechanism of the role between exosome molecules and gene probes is investigated by constructing a regression model, and the above theoretical knowledge is applied to empirically analyze the research scheme of this paper. The regression coefficients of exosome molecules and gene probes showed significant correlation at 0.05 level, indicating that the mechanism of action between the two is monotonically increasing, which well reveals the influence of exosome molecules on gene probes.

Traditional landscape design methods have low efficiency, poor subjectivity and insufficient goal optimization. This paper proposes a landscape design optimization and spatial layout method based on artificial intelligence (AI) algorithms to achieve scientific and efficient landscape design through the combination of collected information data and algorithms. The optimization design of landscape facility paths and spatial dimensions is carried out by adopting a heuristic polygonal layout algorithm, establishing a data model based on the database and scene templates, and combining the landscapes in the polygonal space after landscape matching. The optimal sequence of the landscape is obtained by using the scoring function, and then combined with the particle swarm algorithm to realize the optimization of the landscape layout. The Hypervolume index is stable to about 0.815 in 30 generations, which has a good quality of Pareto optimal solution set. In this paper, the algorithm formulates three groups of landscape design optimization and spatial layout planning schemes for different situations, making full use of the land that is utilized for a certain place. The implementation of the sustainable development scenarios improves the local environmental and social benefits significantly, and the average annual growth rate of employment in related industries reaches 3.16%. Satisfaction survey results show that local residents are most satisfied with the green environment and cultural atmosphere after the implementation of the program, respectively 80.03, 79.35, through the smart management to improve the local environmental quality and cultural atmosphere.

As a large tourism province, Hainan Province produces carbon emissions from the tourism industry that should not be underestimated. In view of the problems reflected, this paper designs a research program based on the data envelopment analysis model to evaluate the carbon emission efficiency of Hainan’s tourism industry and optimize the emission reduction path. The tourism industry of 10 regions in Hainan Province is taken as the object of this study, and the DEA-SMB model is designed. Drawing on existing research results, nine carbon emission efficiency evaluation indicators for the tourism industry in Hainan are set, in addition to improving the principle of measuring carbon emissions and energy consumption in the tourism industry. In order to better promote the green, lowcarbon and sustainable development of the tourism industry, it is proposed to adopt the DEA-SMB model to optimize the carbon emissions of the tourism industry in Hainan Province. Finally, the 2005- 2024 Hainan Tourism Statistical Yearbook is taken as the main data source of this study, and combined with related research data, the optimization effect of carbon emission efficiency and emission reduction path of Hainan tourism industry is explored. In the optimization process of tourism carbon emissions in 10 regions of Hainan Province, the model of this paper has a particularly prominent effect on the optimization of tourism emission reduction in region C, and its emission reduction efficiency is increased to 64.21%, which verifies the tourism emission reduction effect of the model of this paper, and also reflects that there is still a huge room for improvement in local tourism emission reduction projects in Hainan Province.

The all-round penetration of artificial intelligence technology has brought about a drastic change in the educational landscape, and the teaching system of colleges and universities relies on artificial intelligence technology to expand its own boundaries, leading to interdisciplinary knowledge fusion between dual colleges and universities. With the support of AI technology, a teaching system design idea of interdisciplinary knowledge integration is proposed, and a teaching innovation system of interdisciplinary knowledge integration between dual colleges and universities is established. Taking the learners’ interdisciplinary knowledge point response situation as an entry point, input modeling is carried out for the learners’ interdisciplinary knowledge points, forgetting coefficient, etc., and the dual colleges’ interdisciplinary knowledge tracking SA-BiGRU model is established by combining BiGRU and the attention mechanism, and simulation verification is carried out to verify its effectiveness. Taking a vocational college in province G as an example, a dual college interdisciplinary teaching comparison experiment was designed in combination with the teaching innovation system, so as to verify the effectiveness of the interdisciplinary knowledge integration teaching innovation system. The results show that the AUC and ACC of the SA-BiGRU model can reach up to 0.837 and 0.841 respectively in interdisciplinary knowledge tracking, and the learners’ interdisciplinary knowledge reserve and ideological literacy level have been improved by 1.36 and 1.82 points respectively compared with that before the experiment. Relying on artificial intelligence technology can promote interdisciplinary knowledge integration, provide a new research direction for the development of interdisciplinary intelligence in BiGR, and lay the foundation for the cultivation of highly skilled and qualified applied talents.

Smart campus relies on IoT technology to realize teaching management, location monitoring, business processing and other teaching and management activities, this paper draws on the characteristics of the development of smart campus, and builds a decision support system for educational management of smart campus by applying the conditions of IoT technology. The IoT multi-sensor is used to collect educational management data, and the Grobes criterion is applied to exclude the data with too large an error, and the consistency test is performed on the collected data. The least squares method and variance calculation are combined to process the multi-sensor data to optimize the data fusion accuracy. Comparison tests were conducted to analyze the fusion accuracy and variance of the observed data under different methods. Distribute questionnaires online and offline to analyze the feasibility of the construction of IoT in smart campus. Collate the ratings of teachers and students on the educational management decision support system of the smart campus, in which the ratings of teachers and students on the educational management decision part of the school are concentrated in the range of 0.7 to 0.8, and the overall rating of the educational management decision support system of the smart campus is 86.453 points.

Energy, as one of the larger contributing industries to greenhouse gas emissions, has an urgent task to reduce emissions, and standardizing the carbon footprint and trading mechanism of the energy market is an important concern for the development of the current energy industry. Under the guidance of the principle of green, low-carbon and sustainable development of the energy market, this paper first uses heterogeneous blockchain and federated reinforcement learning to design a decentralized energy trading mechanism model. It is found that the model fails to realize the intelligent detection and control of carbon footprint, in this regard, on the original model, the carbon footprint origin algorithm is introduced. Combining the above models and algorithms, the current interactive energy market is explored and analyzed. Consumer user 5 has the largest net benefit, with a specific value of 15.05 million yuan, and comprehensive energy supplier 3 has the largest net benefit, with a value of 37,467,000 yuan, indicating that this paper’s model implements the principle of green, low carbon and sustainable development of energy while meeting the energy needs of consumers and suppliers, maximizing the interests of each other in the process of energy trading, which proves that this paper’s research has excellent practical application value.