Research On Safe Operation Intelligent Perception Methods And Collaborative Governance Strategies In Transportation Infrastructure

As a national vital strategic resource,transportation infrastructure provides important support and a strong guarantee for promoting economic development and sustaining social prosperity.After winning the decisive achievement of completing a moderately prosperous society in all respects,China has entered a new development stage along with the process of moving toward the ambitious goal of achieving socialist modernization and building a socialist modern power,where comprehensively implementing the new development concept and accelerating the construction of a new development pattern has put forward higher requirements for the development of transportation infrastructure.In addition,the world is undergoing major changes unseen in a century,and a new round of scientific and technological revolution and industrial change is developing deeply,providing a broader space for the development of transportation infrastructure.Facilitate the deep integration of new-generation information technologies,such as big data and artificial intelligence,etc.,and new governance concepts with transportation infrastructure,promote data resources,technical advantages,and governance concepts empower transportation infrastructure,and accelerate the integration development of transportation infrastructure with the advances of the times,giving rise to a new mode for the operation management of transportation infrastructure resonating with the development of the times has become the core issue.Those problems,challenges,and requirements faced by the transport infrastructure in the development process of the new era are increasingly appearing the contradiction of imbalance and mismatch between the safety operation ser vice level and safety guarantee capability of the transportation infrastructure,pressing the urgent need to improve quality and efficiency,transformation and upgrad ing in promoting the development of transportation infrastructure.To deal with those complex problems,challenges,and requirements of transportation infrastructure in the new era,systematic and effective methods and means are needed.Therefore,this thesis aims at the governance methods and implementation paths for the quality and efficiency improvement,transformation and upgrading of transportation infrastructure in the new era,focusing on the research of intelligent perception methods and collaborative governance strategies in transportation infrastructure,with the following systematic research works.This thesis first analyzes the essence of traditional management into intelligent management and collaborative governance,respectively.Thereafter,it defines the concepts of intelligent perception and collaborative governance of transportation infrastructure from the perspective of problem response,and sorts out the features of the research object under study from the concept.From the theoretical perspective as the breakthrough of the problem,this thesis draws on operation management theory,complex system theory,big data theory,information fusion theory,stakeholder theory,synergy theory,and game theory as the theoretical basis to construct a theoretical framework of intelligent perception and collaborative governance for safe operation in transportation infrastructure.In this framework,intelligent perception method s,the intelligent assessment method,and the collaborative governance strategy are linked and integrated to clarify their respective ways of penetration and th e progressive relationship among them to reach the research line of this thesis.Aiming at low complexity scenes of transportation infrastructure,an image intelligence perception method is developed based on its feature analysis.The core of the method is the introduction of Transformer,which works with Convolutional Neural Network(CNN),coupl ed with the loss function and evaluation metrics for extraction and learning of image features.Supported by a framework constructed based on operation management theory,complex system theory,and big data theory,the integrated application from data collection to result is realized.The road transportation infrastructure is adopted as an example to verify the performance of the method for detecting risk factors,demonstrating its advantages of low cost and high efficiency.Targeting high complexity scenes of transportation infrastructure,a panoramic scene point cloud segmentation method is proposed based on its feature analysis.The method contains coupled algorithms for data degradation,data normalization,feature fusion,and an encoder-decoder deep learning model.The advantages of the method are the ability to process large-scale point cloud data into a regular 3D structure for data decomposition,fusion of point multidimensional features as input,and coupled pipeline parallelism for aggregation of different model layers of feature maps to facilitate extraction and learning of multiple features for the deep model.The proposed method is based on tunnel transportation infrastructure as an example to verify that the method is able to achieve panoramic segmentation with superior performance,indicating that the method has great potential to be applied in practice for defect detection with high efficiency and low cost.On the basis of analyzing the issues and elements involved in realizing intelligent assessment for safety in transportation infrastructure and deeply understanding the differences between human-causal thinking and machine learning,this thesis innovatively designs a new framework for automated and intelligent risk assessment.In accordance with the cause-and-effect relationships between different modules of the assessment process,the intelligent perception method is introduced into the logical chain of intelligent risk assessment of transportation infrastructure to construct an intelligent assessment method.By clarifying the relationship of each module,the functions of each module are elaborated,and the realization path of each module and the whole assessment process is analyzed in depth.From a practical perspective,the road transportation infrastructure is exemplified to veri fy the feasibility and practicality of the method by introducing UAVs for data collection,constructing data scale conversion coefficients to calculate the physical size,and establishing evaluation criteria as the basis for assessment.This application process involves examining the validity and advancement of the method to show its superiority.For handling the maintenance tasks determined by the assessment results of the transportation infrastructure,a new method of collaborative governance strategy making for multi-stakeholders is constructed.On the basis of clarifying the responsibilities and functions of each stakeholder,this thesis focuses on the correlation relationship and collaborati ve mechanism to address the conflict of interests among multi-stakeholders.Specifically,a bi-level interaction model based on game theory is proposed to analyze the interaction and collaboration relationship among the multi-stakeholders so as to reach a satisfactory outcome for all stakeholders by obtaining an equilibrium solution,eliminating conflicts of interest,and creating a synergistic atmosphere to improve the performance of transportation infrastructure and user satisfaction.The feasibility of the method is verified based on a maintenance case of road transportation infrastructure.Compared with the traditional method,the proposed method produces better results and is able to effectively manage the conflict of interest among multistakeholders to realize improvement in revenue,performance,and user satisfaction.The proposed method is capable of achieving equal and decentralized interaction among multi-stakeholders,breaking the traditional pyramidal management,which can serve as a practical tool for strategies with high guidance value.This thesis caters to the new era of national development,and deeply integrates the new-generation information technologies and collaborative governance theory with the development of transportation infrastructure,upon which to construct deep learning algorithm-based intelligent perception methods,intelligent assessment method,and a bi-level model of collaborative governance strategy making.The proposed methods enable to liberate labor,improve productivity,reduce operational costs,and boost the risk prevention capability and decision-making efficiency of transportation infrastructure.Overall,the research results of this thesis are beneficial to the theoretical system of transportation infrastructure and enrich the governance methods and implementation paths of transportation infrastructure,providing valuable references and meaningful drawings for the quality and efficiency improvement,and transformation and upgrading of transportation infrastructure.