Pengjun Wu

Dr. Pengjun Wu

orcid

Research Focus & Metrics

Academic Career Length

3Years (2024-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
2 March 2024 in Scientific Reports
Verified
Journal Article
Open access
PDF file
PWPengjun Wu
Pengjun Wu
ZZZhanzhi Zhang
Zhanzhi Zhang
XPXueyi Peng
Xueyi Peng
RWRan Wang
Ran Wang
In the realm of urban planning, the integration of deep learning technologies has emerged as a transformative force, promising to revolutionize the way cities are designed, managed, and optimized. This research embarks on a multifaceted exploration that combines the power of deep learning with Bayesian regularization techniques to enhance the performance and reliability of neural networks tailored for urban planning applications. Deep learning, characterized by its ability to extract complex patterns from vast urban datasets, has the potential to offer unprecedented insights into urban dynamics, transportation networks, and environmental sustainability. However, the complexity of these models often leads to challenges such as overfitting and limited interpretability. To address these issues, Bayesian regularization methods are employed to imbue neural networks with a principled framework that enhances generalization while quantifying predictive uncertainty. This research unfolds with the practical implementation of Bayesian regularization within neural networks, focusing on applications ranging from traffic prediction, urban infrastructure, data privacy, safety and security. By integrating Bayesian regularization, the aim is to, not only improve model performance in terms of accuracy and reliability but also to provide planners and decision-makers with probabilistic insights into the outcomes of various urban interventions. In tandem with quantitative assessments, graphical analysis is wielded as a crucial tool to visualize the inner workings of deep learning models in the context of urban planning. Through graphical representations, network visualizations, and decision boundary analysis, we uncover how Bayesian regularization influences neural network architecture and enhances interpretability.

16 November 2024 in Displays
Verified
Journal Article
Closed access
PWPengjun Wu
Pengjun Wu
YLYao Liu
Yao Liu
HCHuijie Chen
Huijie Chen
XLXiaowen Li
Xiaowen Li
HWHui Wang
Hui Wang

31 October 2024 in PLOS ONE
Verified
Journal Article
Open access
PDF file
YLYao Liu
Yao Liu
PWPengjun Wu
Pengjun Wu
XLXiaowen Li
Xiaowen Li
WMWei Mo
Wei Mo
This paper takes the example of industrial architectural heritage in Dalian to explore design scheme generation methods based on generative artificial intelligence (AIGC). The study compares the design effects of three different tools using the Analytic Hierarchy Process (AHP). It first establishes the key indicator weights for the renovation of industrial architectural heritage, with the criterion layer weights as follows: building renovation 0.230, environmental landscape 0.223, economic benefits 0.190, and socio-cultural value 0.356. Among the goal layer weights, the highest weight is for the improvement of living quality at 0.129, followed by resident satisfaction at 0.096, and educational and display functions at 0.088, while the lowest is for renovation costs at only 0.035. The design schemes are generated using Stable Diffusion, Mid Journey, and Adobe Firefly tools, and evaluated using a weighted scoring method. The results show that Stable Diffusion excels in overall image control, Mid Journey demonstrates strong artistic effects, while Adobe Firefly stands out in generation efficiency and ease of use. In the overall score, Stable Diffusion leads the other two tools with scores of 6.1 and 6.3, respectively. Compared to traditional design processes, these tools significantly shorten the design workflow and cycle, improving design quality and efficiency while also providing rich creative inspiration. Overall, although current generative artificial intelligence tools still have limitations in understanding human emotions and cultural differences, with continuous technological iteration, this method is expected to play a larger role in the design field, offering more innovative solutions for the renovation of industrial architectural heritage.

25 April 2025 in Land
Verified
Journal Article
Open access
PDF file
PWPengjun Wu
Pengjun Wu
DXDandan Xu
Dandan Xu
NCNannan Cui
Nannan Cui
XLXiaowen Li
Xiaowen Li
YLYAO LIU
YAO LIU
With the acceleration of urbanization, the accessibility and equity of urban green spaces have become crucial issues in urban planning and public health. In the context of the 15-min living circle, whether residents can conveniently reach parks within a walkable or bikeable time frame directly impacts their quality of life and social well-being. Traditional park accessibility evaluation methods, such as the G2SFCA, effectively measure accessibility but fail to fully account for the diversity of travel modes and the impact of regional disparities on equity. This study employs the TB-G2SFCA method, integrating the concept of the 15-min living circle, to analyze the equity of park accessibility in the Dianchi Lake ring area of Kunming under different travel modes. The findings reveal significant disparities in park accessibility for walking and cycling, particularly in suburban communities distant from the city center, where many areas cannot reach a park within 15 min. Although accessibility improves under driving and public transit modes, resources remain concentrated in well-connected areas, leaving peripheral regions with insufficient access. Compared to the traditional G2SFCA method, the TB-G2SFCA approach more accurately reflects spatial differences and equity issues across travel modes. This study suggests that future urban park planning should optimize resource allocation, improve transportation networks, and enhance park accessibility in peripheral areas—especially for walking and cycling—to achieve a more equitable and sustainable distribution of urban green spaces.

15 September 2025 in Ain Shams Engineering Journal
Verified
Journal Article
Open access
YLYao Liu
Yao Liu
BXBeiyuan Xu
Beiyuan Xu
JFJiarong Feng
Jiarong Feng
PWPengjun Wu
Pengjun Wu

17 January 2025 in Scientific Reports
Verified
Journal Article
Open access
PDF file
PWPengjun Wu
Pengjun Wu
XLXiaowen Li
Xiaowen Li
YFYaFang Feng
YaFang Feng
YLYAO LIU
YAO LIU
As the need for balancing cultural heritage preservation with economic development becomes increasingly urgent, the sustainable development of ancient villages has emerged as a critical issue. This study focuses on the ancient villages of Tengchong, Yunnan, to examine the profound impact of art intervention in public spaces on their development. Through literature review, QGIS analysis, questionnaire surveys, and hierarchical analysis, the study reveals that art intervention plays a pivotal role in enhancing the environmental appearance (0.307), social and cultural values (0.270), village transformation (0.242), and economic benefits (0.181) of these villages. Art intervention not only significantly improves infrastructure and optimizes transportation but also effectively protects natural landscapes and environmental sanitation, fostering sustainable economic growth. By reasonably controlling costs, art intervention provides an innovative and feasible pathway for balancing cultural preservation with economic development, highlighting its unique value in promoting the sustainable development of ancient villages.