Big data is an important foundation in social economy, science and technology, life and other fields, which also becomes a strategic emerging industry and has a crucial impact on the development of enterprises. As a new business model, its development is greatly limited due to the huge amount of data and difficult management. At present, there are many problems in power trading enterprises, such as backward management and low efficiency. The development of big data and blockchain technology would provide new management models for power trading enterprises and eliminate data inconsistency. It can improve data quality and help improve work efficiency, so as to reduce operating costs. Therefore, this paper introduced big data and blockchain based on fuzzy algorithm into the research of digital transformation of enterprise management. Blockchain technology provided technical support for enterprise data management. By starting from the concept of big data and blockchain, this paper would study and analyze how to promote the digital transformation of enterprise management. The research results showed that big data and blockchain based on fuzzy algorithm could promote the digital transformation of power enterprise management and improve the digital transformation process of power enterprises. This was about 11% higher than the digitalization process of traditional enterprises, and the satisfaction score was about 14.7% higher. Through data governance, the speed of digital transformation of power enterprise management was improved.

With the rapid expansion of high-speed railway network, the real-time monitoring of trackside equipment becomes particularly important. To detect trackside equipment information more accurately, a YOLO-R algorithm grounded on the improved You Only Look Once v3 (YOLOv3) algorithm is proposed, and the trackside equipment identification and detection model is constructed. By introducing feature pyramid network and adaptive Bessel curve network, the new model can effectively identify and locate different types of trackside equipment such as switch machine, derailer, and shaft counter. The experiment findings denote that the new model is superior to the existing technology in all aspects of on-orbit equipment recognition and detection, the computer resource occupancy rate is only 22%, the image recognition accuracy rate is more than 98%, and the processing speed is up to 200 images/second. This research not only raises the automation level of trackside equipment monitoring, but also provides a powerful technology for railway safety operation.

Generative artificial intelligence represented by ChatGPT has attracted wide attention in the field of education because of its powerful generative ability, both personalized learning, understanding the learner’s motivation, and providing personalized tutoring and feedback for education. With the advent of the Education 2.0 era, smart classroom has become a strategic choice for the construction of education modernization, and is widely used in higher education and vocational education. Generative AI enlightens students’ engineering thinking, computational thinking, design thinking and systems thinking, which not only helps students to master their professional courses, understand what they have learned, and improve their academic performance, but also assists teachers in updating their course content, keeping abreast of students’ learning trends, improving their teaching efficiency, and simplifying their work. However, generative AI is faced with expertise gaps and uncertainty about the existence of generated content in its application, as well as ethical issues, and this study proposes that the needs and values of education should be respected, with the aim of efficient and convenient services, and that data-driven and ethical ethics should be emphasized in future development. Smart classroom and enlightened thinking with the application of generative AI is a new way of thinking about educational change, which can help teachers and students to effectively carry out multiple interactions, enable teachers to better understand students, play the role of human beings in education, and truly allow technology to be used for teaching and promote classroom teaching reform.

Ultra-low energy buildings for building energy efficiency development, compared with traditional buildings have obvious advantages. This paper simulates ultra-low-energy residential buildings in severely cold regions through Software PHES, and calculates the energy-saving results of ultra-lowenergy residential buildings. The carbon emission factor method is analyzed, and the carbon emission factor is calculated at different stages in the life cycle of the building. Select ultra-low-energy residential buildings in cold regions for modeling, input meteorological parameters, indoor environmental parameters and internal disturbance settings, building envelope, and combine with heat recovery system to simulate the operation of ultra-low-energy residential buildings in cold regions. Analyze the indoor and outdoor temperature and humidity values of traditional houses and compare them with those of ultra-low-energy-consumption houses to verify the advantages of ultralow-energy-consumption residential buildings. Calculate the energy-saving efficiency of ultra-lowenergy residential buildings. Using the 9# residential building of Ruihu·Yunshanfu in Datong as a practical verification case, this ultra-low energy residential building has a total life-cycle carbon emission of 171.078 tCO₂/a, with a unit area carbon emission of 16.415 kgCO₂/m²·a. Compared to the energy-saving design standards implemented in 2016, the carbon emission intensity is reduced by 60.02%, fully confirming the carbon reduction benefits of ultra-low energy residential buildings in severe cold regions.

In recent years, the development of study activities is in full swing. In order to study the eco-education effect in national park study activities, this paper introduces Bayesian network and constructs an ecoeducation effect assessment model based on Bayesian inference. In the comparison of the absolute error of the assessment value with other assessment models, the assessment accuracy of the Bayesian inference assessment model in this paper is obtained. After constructing the ecological education effect assessment index system and completing the assignment, the level of ecological education that should be achieved in the national park study activities is obtained through Bayesian inference diagnosis. Finally, according to the results of education effect assessment, the probability of each indicator being in various states is obtained by simulation using Monte Carlo method. The mean absolute error of the Bayesian assessment model is 0.26 points, which is smaller than other comparative assessment models and has the highest assessment accuracy. The model’s ecosystem principles, anthropogenic intervention impacts, ecological disasters and ecological protection measures should be guaranteed to reach 75.6, 64.8, 67.9 and 69.4. The ecological operation rules (59.4→79.8), climate change (50.6→70.2), biodiversity reduction (52.2→69.8), and pollution prevention and control (56.4→78.3) have the highest accuracy for the ecosystem principle, anthropogenic intervention impacts, ecological disasters and ecological protection measures, respectively. , anthropogenic intervention effects, ecological disasters and ecological conservation measures, and ecological education effects had the greatest impact. The overall score of ecological education effect was 84.1, and the scores of ecosystem principle, human intervention impact, ecological disaster and ecological protection measures were 83.8, 85.2, 83.0 and 84.2.

For enterprises, development is ultimately reflected in the task completion performance of employees, and in order for employees to create higher task performance, it is necessary to consider not only their education and knowledge level, but also their emotional management ability. This study first collects data related to employees’ emotion management ability and task completion efficiency improvement through questionnaires, and then analyzes the statistical data by using the potential impact identification model designed based on Bayesian neural network model to obtain the potential impact probability of each dimension of emotion management ability on task completion efficiency improvement. The analysis of the forward and reverse inference probabilities of the Bayesian network model indicated that the most important potential influence factor leading to the improvement of task completion efficiency was the emotion expression ability, with a forward and reverse inference probability of 36.2% and 59.4%, respectively, followed by the emotion regulation ability and emotion acceptance ability. The results of this study reveal the important potential influence of emotion management ability on task completion efficiency enhancement, and the formulation of task completion efficiency enhancement strategies based on the perspective of emotion management ability can effectively enhance employee task performance, which in turn promotes the overall development and competitive advantage of enterprises.