A Digital Twin Model of High-Pile Docks’ Structures based on Neo4j Knowledge Graph

High-pile docks plays a vital role in modern civil engineering field. The operational efficiency and structural safety of these docks are paramount. However, structural damage to high-pile docks can lead to economic and human losses. To effectively manage structural health monitoring information for high-pile docks, this research utilizes knowledge graph technology to construct a visual knowledge repository for high-pile docks, named as the High-Pile Dock Knowledge Graph (DKG-Neo4j). For the construction of DKG-Neo4j, a plethora of literature, news, and standards related to health monitoring in the field of high-pile docks were gathered. Then, the collected information was organized hierarchically and categorized into tables, with CSV files exported using Python. Finally, this aggregated information was integrated into the Neo4j graph database to establish a comprehensive knowledge graph. The proposed High-Pile Dock Knowledge Graph (DKG-Neo4j) not only reflects the current state of the docks but also captures the connections between issues and maintenance strategies, empowering experts to engage in structural health monitoring and risk analysis. The DKG-Neo4j model boasts structural flexibility and scalability, contributing to smart cities and providing new approaches for the sustainable development of high-pile docks and the digital management of civil engineering.


Introduction
Nowadays, High-pile docks are being applied in many fields, including marine engineering, shipping, and petrol industries due to their support capacity, stability, and load-bearing capability.As of early 2022, there were approximately 2,400 high-pile docks in China.The Port Authority of New York and New Jersey in America has many high-pile docks, and there are over 30 high-pile docks in the North Sea oil fields in England.[1] South Korea's Hyundai Heavy Industries has made significant investments in constructing high-pile docks, with many high-pile docks at its Hyundai shipyards.[2] However, over 80% of high-pile docks in China being affected and the safety of many port docks in China remains compromised.For example, in 2016, the "Kai Rui 7" overloaded and collided with the Shantou Port, causing severe damage to the high-pile dock itself and various degrees of cracking in the surrounding beams.[3] On January 2, 2005, the vessel "Dong Hong 8" touched berth 2 while berthing at the Seven-Mile Wharf of Wenzhou Port, resulting in a direct economic loss of 2.8 million yuan due to ship residul speed, wind, and current effects.Furthermore, regarding structural damage data for high-pile docks, traditional data systems have many inconveniences.[4] As technologies advance, researchers utilize automation, 3D printing, and digital detection technologies to replace traditional

Fatigue analysis
Step2.Establishing nodes, relationships and hierarchies according to each keyword in Neo4j and creating hierarchical or parallel relationships between different keywords.Step3.Outputting keywords and their relationships to create the high-pile dock digital twin model in Neo4j.The Analytic Hierarchy Process (AHP) is used to measure the ranks of indicators in each layer with respect to the indicators in the previous layer.[6] Computing the maximum eigenvalue of the judgment matrix and then using equation ( 1) to calculate the consistency index: In the formula, n represents the order of the judgment matrix.Obtain the average random consistency index RI of the judgment matrix and calculate the consistency ratio using equation (2): = (2) Step4.Selecting aimed label by clicking on the labels on the right side of the Neo4j interface.Only the selected label will remain, while the other parts of the knowledge graph appear in grey.

Neo4j Code of DKG-Neo4j
As illustrated in the Figure 1 there are a total of 200 nodes and relationships, and using Neo4j makes it easy read them.This research used different color to distinguish different relationships and types, as depicted in figure 2 and figure 3.This means "Durability" is a node, presenting its label name is "n23" and its label name is "Concrete Wear Resistance".

Querying nodes and relationshipsIn Neo4j
, MATCH is used to query nodes and relationships.Figure 6 shows the example of querying all nodes:MATCH (n) RETURN (n) .This statement will return all nodes according to the codes

Application of DKG-Neo4j and cases studies
DKG-Neo4j has different applications and can be used in different fields.Learning specific cases can help people understand its usage well.

DKG-Neo4j innovations
First, the DKG-Neo4j system fills the gap in systematic databases for high-pile docks.Second, using DKG-Neo4j system.Third, the DKG-Neo4j system provides information to users.Fourth, it can give suggestions based on existing similar cases.Fifth, the DKG-Neo4j uses a property graph model to store data, with each data point represented as an independent node connected by data relationships.

DKG-Neo4j Model feasibility analysis
Currently, advanced technologies like 3D modelling and digital control are commonly used in piled foundation construction.[8] Maintenance work is also adopting new technologies like robot inspections and anti-corrosion coatings, improving work efficiency and stability.[9] As piled foundation projects continue to grow, future challenges will increase, and data will become scattered. [10]

Cases analysis
For example, when a typhoon threatens a port, for ships, the following measures should be taken before a typhoon: First, inspect deck equipment, and ensure that ventilation ducts and air pipes on open decks are sealed.Second, close deck entrances and exits, passageways, and portholes.Third, close all side doors and watertight doors except emergency doors.Fourth, inspect the main engine, auxiliary machinery, boilers, and other equipment to ensure they are in good condition.

Conclusion
The proposed Digital Twin Model (DKG-Neo4j) is a system that monitors, diagnoses, and optimizes real piled foundation systems through visualizing the database.Based on physical modelling and datadriven methods, this model can continuously model, providing valuable information for system maintenance and operation.More ports are adopting advanced information technologies such as digital twins, 5G, the Internet of Things, big data, cloud computing, and digital sensing to build "digital ports," creating regional "digital" information hubs that facilitate extensive connectivity among resources, participants, and logistics in port logistics.[10]

Research Contributions:
The DKG-Neo4j in digital port construction can significantly improve port operational efficiency, reduce logistics costs, and provide real-time feedback to port managers, helping them monitor and adjust port operations.[11] With the continuous development and popularization of Internet, big data, and artificial intelligence technologies, the application prospects of the Piled Foundation Digital Twin Model are even broader, being used in more fields.[12] 4.2 The potential of Neo4j in Digital Twin Models First, Neo4j is good at monitoring many aspects from civil engineering areas to radio monitoring.[13] By monitoring and processing various data such as ocean conditions and ship statuses, combing with machine learning, the model can learn and optimize itself, enhancing its ability to predict and optimize piled foundation systems.[14] It supports the integration of data resources from various existing information systems, providing a real-time three-dimensional representation of all elements in the port, such as port gates, docks, yards, vehicles, and vessels.Through analysis of keywords in various areas, such as production, operation, transportation, logistics, security, and more, the model can provide comprehensive perception of all elements in the port, assisting people in understanding the port's operational situation.Second, the model can achieve digital management and optimization and provide data to other related fields, such as logistics, shipping, and port management, contributing more to the digitization and intelligence of the industry.[15]

Figure 1 .
Figure 1.The knowledge graph of defined keywords using Neo4j.

Figure 2 .
Figure 2. Part of the node labels of the knowledge graph.

Figure 3 .
Figure 3. Part of the relationship types of the knowledge graph.

Figure 4 .
Figure 4.The knowledge graph of defined keywords using Neo4j.

Figure 5
Figure 5 shows an example of relationship: MATCH p=()-[r:included]->() RETURN p LIMIT 25 This represents the relationship between r and P is "included".There are arrows point from Label002 to Type 002.LIMIT 25 means the maximum number of relationships is 25.

Figure 5 .
Figure 5.The knowledge graph of defined keywords using Neo4j.