A method and management system for establishing a multi-feature digital twin model for smart construction sites

Abstract

The present invention relates to a method for establishing and managing a multi-feature digital twin model of a smart construction site, and belongs to the field of applied information technology. The implementation method is as follows: digital information of personnel, machines, materials, processes and environments is collected through a variety of sensors, and the information is transmitted to a digital twin processing module through a distributed network. The established digital twin model will be transmitted to a central processing unit, and the processor will use a chain similarity algorithm to compare the newly established digital twin model with the original digital twin model in the model library. After the comparison, the corresponding model in the model library will be updated and optimized, or a new digital twin model will be established and stored in the library. The present invention uses digital twin technology to achieve the establishment, update, optimization, prediction and evaluation of multi-feature models of smart construction sites, and can provide technical support for the dynamic management of smart construction sites.

Description

Technical Field

[0001] The present invention relates to the field of smart construction site management, and more specifically, to a method for establishing and managing a multi-feature digital twin model of a smart construction site. Background Art

[0002] With the continuous development of the economy and society, the traditional construction site management model has been greatly challenged in recent years, and the smart construction site management system has increasingly become the first choice for construction site management. At present, the relevant research on the smart construction site management system has made great progress, and the digital twin technology has also developed rapidly in recent years. This technology can effectively reflect various business processes by collecting various raw data and performing data fusion processing. At the same time, it can realize the comprehensive presentation, accurate expression and dynamic monitoring of the status and behavior of physical entities in the digital world. The use of digital twin technology can comprehensively collect and present multiple features of personnel, machines, materials, processes, and the environment. The present invention realizes dynamic management, improves the timeliness and accuracy of the digital twin model, and provides a new solution for smart construction site management. Summary of the Invention

[0003] The purpose of this invention is to provide a more effective method for establishing a multi-feature digital twin model of a smart construction site and a management system, so as to overcome the shortcomings of existing smart construction site management and improve the efficiency and accuracy of smart construction site management.

[0004] To achieve the above object, the present invention provides the following technical solutions:

[0005] A method and management system for establishing a multi-feature digital twin model for a smart construction site, including:

[0006] The multi-feature data acquisition module is used to obtain data on personnel, machines, materials, processes, and the environment on the construction site, and can upload the data in real time. Personnel includes construction workers, engineers, managers, etc. Machines include all large, medium, and small operating equipment on the construction site. Materials include all materials used on the construction site. Processes refer to the construction techniques used on the construction site. The environment includes the geographical environment, climate environment, and cultural environment.

[0007] Data transmission module, used to transmit the collected data to the digital twin module through the IMT-2020 high-speed distributed network;

[0008] The digital twin processing module is used to perform digital twin modeling on the various feature data received, and to establish a digital twin model that is highly similar to the features of multiple objects on the actual construction site;

[0009] The central processing unit module is used for model information storage and comparison, and determines the similarity of feature models by comparing multi-feature digital twin models;

[0010] In the above solution, the characteristic information data includes: shape, size, speed, temperature, wind force, pressure, weight and humidity;

[0011] In the above solution, the specific method of comparing the CPU model information is as follows:

[0012] The processor receives the feature model data transmitted by the digital twin processing module and performs a similarity comparison of the data chain through a chain comparison algorithm. During the comparison, it is first determined that the compared feature model belongs to an object among personnel, machines, materials and environment, and then it is determined that the model belongs to a feature under the object. After that, the twin data of the relevant model is retrieved from the sub-similarity chain library. By comparing the new model feature data with the model feature data in the sub-library, the longest common subsequence of the data chain is determined to obtain specific similarity data. The similarity data is represented by DL=(d1, d2, d3, d4, d5), where d1 represents the feature attribution, which is represented by a letter, d2 represents the lower-level attribution, which is represented by two digits, d3 represents the sub-similarity chain of a certain feature, which is also represented by two digits, d4 represents the similarity between a feature data model in the received digital twin model and the sub-similarity chain, which is also represented by a number. The size of the number represents the similarity, and d5 indicates whether the sub-similarity chain will be updated or replaced, which is represented by the number 0 or 1. If it is 0, it means that the update process will not be performed, and if it is 1, the update process will be performed.

[0013] From the above solution description, it can be seen that this solution provides a method and management system for establishing a multi-feature digital twin model of a smart construction site. It obtains the feature data of important objects on the construction site through multiple sensors, uses digital twin technology to model the data information, and then performs in-depth comparison through the central processing unit.

[0014] Compared with existing smart construction site management systems, this invention is efficient and accurate. This solution fully utilizes digital twin technology to realize the collection of feature data of various objects in the smart construction site and the dynamic management of the smart construction site. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] In order to more clearly illustrate the method used in the present invention and its use process, the following briefly introduces the drawings required for describing the above solutions.

[0016] Figure 1 This is an embodiment of a multi-feature digital similarity chain for a method for establishing a multi-feature digital twin model of a smart construction site and a management system of the present invention;

[0017] Figure 2 This is a flow chart of the method for establishing and managing a multi-feature digital twin model of a smart construction site according to the present invention;

[0018] Figure 3 This is a schematic diagram of the workflow of the method for establishing a multi-feature digital twin model of a smart construction site and the management system of the present invention. DETAILED DESCRIPTION

[0019] The present invention is a method and management system for establishing a multi-feature digital twin model of a smart construction site. The specific implementation method will be described in detail below with reference to the accompanying drawings. This implementation method is only used to explain the invention scheme and should not be understood as a limitation on the patent of this invention.

[0020] Reference Attachment Figure 1 , Figure 1 This is an implementation example of the multi-feature digital similarity chain of the present invention;

[0021] As shown in the figure, the similarity chain of the case is R 23 54 68 1, where R represents that the object to which the collected feature belongs is a person, 23 represents that the feature is feature No. 23 in the person model, 54 and 68 indicate that the similarity between the collected feature and the feature sub-model No. 54 reaches 68%, and 1 indicates that the similarity chain and its corresponding digital twin model will be updated, indicating that the similarities corresponding to the remaining features identified this time are within a reasonable range, and only the similarity corresponding to this feature is within a reasonable range. If this item is 0, it means that the similarity chain and its corresponding digital twin model will not be updated, indicating that the similarities of all features identified this time are within a reasonable range.

[0022] Reference Attachment Figure 2 , Figure 2 This is a flow chart of the present invention, specifically including:

[0023] The multi-feature data acquisition module is used to collect multi-feature data from personnel, machines, materials, processes, and environments through multiple sensors such as temperature sensors, humidity sensors, infrared sensors, cameras, weight sensors, wind sensors, speed sensors, current and voltage sensors, and transmit them to the digital twin processing module through the network.

[0024] The digital twin processing module is used to receive data collected from multiple sensors and establish a digital twin model for the collected multi-feature data through digital twin technology.

[0025] The central processing unit module includes a model information storage and a model information comparison module. The processor is used to perform similarity comparison on the received digital twin model in real time and update and store it, and at the same time establish a multi-feature digital similarity chain based on the stored model.

[0026] The similarity comparison module is used to execute the specific process of processor model information comparison. It compares the newly received model feature data chain with the feature data chain in the model library through the chain comparison algorithm, and finally obtains the similarity.

[0027] Specifically, there is an essential difference between the multi-feature digital similarity chain and the feature data chain. The former represents the ownership information and similarity value of the model, expressed in 8-bit characters, and directly displays the ownership and similarity information of the feature data; the latter represents a data chain composed of the feature information of a certain feature, which is directly used for feature comparison. Each feature has multiple groups of parallel sub-data chains that can be used for similarity comparison.

[0028] Reference Attachment Figure 3 , Figure 3 The workflow diagram of the present invention is as follows:

[0029] Table 1

[0030] Object / Feature time Place Testing results staff 7:37 factory gate No abnormalities crane 8:20-8:25 Block C No abnormality, construction in progress Stone 8:42 Construction Site Area 2 The weight is insufficient, please check as soon as possible Mechanical backfill process 8:44 Construction Site Area 3 No abnormality, construction in progress construction site environment 8:00-9:00 construction site Breeze level 3, 7℃, no rain within 1 hour

[0031] Table 1 shows the data records of a certain period of time in the intelligent construction site management system, including when and where employees entered the construction site, as well as data information on multiple characteristics such as employees' facial data, behavioral characteristics, fingerprint characteristics, etc. The working status of the crane in a certain short period of time, various physical information of the stone, process detection of the mechanical backfill process, and implementation detection of the construction site environment can all be clearly displayed in the system.

[0032] The whole process will be described in detail below with reference to the embodiments:

[0033] When a construction site vehicle enters the sensor data collection area, the sound sensor, multiple biometric sensors, cameras, multiple environmental data sensors, and vibration sensors will collect multiple feature data information of the vehicle in sequence. While acquiring the data information, the system has automatically saved the data records shown in Table 1. These feature data information will then be immediately uploaded to the digital twin processing module.

[0034] The digital twin processing module receives the vehicle's multi-feature information and uses digital twin technology to model this information, generating a simulation model and a digital twin model. The multi-feature information includes the vehicle's color, structure, license plate, speed, and model. The digital twin model is then uploaded to the central processing unit (CPU). The CPU uses the model information storage module and the model information comparison module to perform a real-time similarity comparison on the received digital twin models, updating and storing them. The digital twin model is then uploaded to the CPU.

[0035] The central processing unit includes two modules: information storage and information comparison. After receiving the digital twin model, the central processing unit will extract the data link from the received digital twin model in real time, and perform in-depth comparison through the chain comparison algorithm. If the data links of multiple features of the vehicle are compared with the data links in the library, and the similarity of all feature data links involved in the comparison with the data links in the library reaches a reasonable range, the comparison is completed, and the data links in the library will not be updated and optimized. If after the comparison, there is only one feature data link in the feature data links involved in the comparison whose similarity with the data link in the library does not reach a reasonable range, and the remaining feature data links are within a reasonable range, the feature will be stored in the library. If after multiple comparisons, some data links in the multiple parallel relationship data links under this feature have never reached a reasonable range of similarity with the new data link, then the data link will be replaced by the newly stored data link, thereby achieving the purpose of updating and optimizing the data link. While performing the comparison, the processor will establish a multi-feature digital similarity chain based on the stored model.

[0036] The model library corresponding to the multi-feature digital similarity chain is expressed as DL = (d1, d2, d3, d4, d5). Each model feature corresponds to a multi-feature digital similarity chain, which is used to determine the attribution and similarity of the feature. d1 represents that the feature belongs to an object in the construction site. In this example, the construction site vehicle belongs to the machine, and J is used to represent the first character of the similarity chain; d2 represents sub-layer attribution, that is, the feature is a feature of a certain object. In this example, the color, structure, license plate, speed and model of the vehicle are specific features of the construction site vehicle, which can be represented by 01, 02, 03, 04 and 05 respectively; d3 represents the sub-data chain of the parallel relationship corresponding to a specific feature. In this example, the model feature of the car can have dozens of sub-data chains. In a certain acquisition process, one of the sub-data chains will appear, or a sub-data chain not in the library will appear. In one case, if only the model feature among the vehicle features collected at a certain time has not found the corresponding sub-data chain in the library, that is, the similarity between the newly appeared feature and any sub-data chain in the library is within a reasonable range, then the new feature will be stored as a new sub-data chain in the model feature. If the model feature and some other features are not matched to a sub-data chain within a reasonable similarity range in the library, the construction site vehicle is considered non-compliant, and the model number in d3 can be any of 01-99; d4 represents the maximum similarity between the new feature and multiple sub-data chains, that is, the similarity between the new feature and a sub-data chain is the maximum value of the similarities between the new feature and all sub-data chains; d5 indicates whether the sub-data chain will be stored and updated. If it is 1, it means that a certain feature of a vehicle is stored as a sub-data chain. If it is 0, it means that it will not be stored and updated.

[0037] The similarity module provides feedback to the central processing unit module based on the results of the information comparison of the data link, and updates, optimizes and stores the model features in the processor.

[0038] The data collected by the system through image sensors, speed sensors, etc. will be recorded by the system, and information including time and location will also be recorded to better manage the construction site.

[0039] The model library corresponding to the multi-feature digital similarity chain in the central processing unit is DL = (d1, d2, d3, d4, d5), and the specific formula used is as follows:

[0040]

[0041] It can be seen from the formula that the character value of d(i) varies with i, that is, the value corresponding to each character of the multi-feature digital similarity chain will vary depending on the collected and recognized data.

[0042] The model library corresponding to the feature data chain in the central processor is represented by TL, TL=(t1,t2,…,t m ), the newly acquired characteristic digital chain is represented by TL', TL'=(t1', t2', ..., t n '), respectively, using features_t = {people, machine, material, technology, environment: [ft1, ft2, ..., ft m ]}, features_t'={people, machine, material, technology, environment: [ft1, ft2,..., ft n ]}, and m≥n, the corresponding chain comparison algorithm formula is as follows:

[0043]

[0044] The formula shows the specific process of similarity comparison, that is, the longest common subsequence of two similar chains is an empty set, indicating that the two digital chains have no overlapping parts, and the corresponding similarity is 0. The longest common subsequence of the two chains is not an empty set, indicating that the two digital chains have overlapping parts. The specific overlapping length can be obtained according to the formula, and the ratio of the overlapping length to the original chain length can be expressed as the similarity of the two chains.

[0045] The algorithm is only used by the processor to compare the digital twin model. The digital chain is a feature description of the digital twin model, and the similarity chain is used to represent the comparison process and results. When the digital chain is updated, optimized and stored, the digital twin model is also updated, optimized and stored accordingly.

[0046] It should be noted that the method for establishing a multi-feature digital twin model of a smart construction site and the dynamic management system of the present invention are a system that combines software and hardware, and applies multiple important technologies such as digital twins, the Internet of Things, artificial intelligence, and GIS to ensure the authenticity and reliability of data collection; the system can reflect the data that needs to be monitored on the smart construction site in real time, providing a dynamic digital construction site management method.

[0047] In the embodiments provided by the present invention, the modules or components are connected via a network, and some of the infrastructure can be understood as physical connections. The functions of the relevant modules or certain sensors under the modules can be selectively used according to the needs of the actual use scenario.

[0048] The contents shown in the drawings used in this application are for illustrative purposes only and should not be considered as limitations of this patent. Instead, they are intended to be consistent with the broadest scope consistent with the steps and module relationships described herein. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of this invention shall be included within the scope of protection of the claims of this invention.

Claims

1. A method for establishing a multi-feature digital twin model of a smart construction site, comprising the following steps: Multi-feature data collection: sensors are used to collect multi-feature data on personnel, machines, materials, processes, and environments; Digital twin model establishment: Use digital twin technology to establish a digital twin model for the collected multi-feature data; Model information processing: The established digital twin model is transmitted to the central processing unit, which includes a model information storage module and a model information comparison module. The central processing unit compares the similarity of the model information in the received digital twin model in real time and updates and stores it. At the same time, it establishes a multi-feature digital similarity chain based on the stored model information; Specifically, let the multi-feature digital similarity chain be DL=[d(1), d(2), d(3), d(4), d(5)], where d(1) represents the category to which a certain feature data belongs in the multi-feature data, and the categories are: personnel, machine, material, process or environment, represented by the letters R, J, W, G, H respectively; d(2) represents the identifier of the sub-feature data, that is, the feature number of the multi-feature data in the corresponding category; d(3) represents the feature data variant chain, which is used to describe the variant form of the same feature data generated as the model information is updated. They have the same belonging, the same basic attributes, and similar trends or distributions, but are independent of each other; d(4) represents the feature data similarity, which refers to the similarity between the new feature data chain converted from the newly received multi-feature data and the feature data variant chain of d(3), which is calculated by the chain comparison algorithm; d(5) represents the update status identifier, which marks whether the variant in the feature data variant chain has been updated; The corresponding formula is as follows: The chain comparison algorithm is used to process the multi-feature digital similarity chain mentioned above: the feature data variant chain in d(3) is represented by TL, TL = {people, machine, material, technology, environment: [ft1, ft2, ..., ftm]}, m is the length of the feature data variant chain, where ft1, ft2, ..., ftm represent the specific values ​​of the multi-feature data in the feature data variant chain, and the newly acquired multi-feature data is represented by the new feature data variant chain TL', TL' = {people, machine, material, technology, environment: [ft'1, ft'2, ..., ft'n]}, n is the length of the new feature data variant chain, where ft'1, ft'2, ..., ft'n represent the specific values ​​of the newly acquired multi-feature data in the new feature data variant chain, and m ≥ n. The corresponding chain comparison algorithm formula is as follows: Wherein LCS is a dynamic programming algorithm, T is similarity, LCS(tj) represents the longest common subsequence of two feature data variant chains, L(LCS(tj)) represents the length of the subsequence, L(LCS(tj)) / m represents the ratio of the length of the longest common subsequence of two feature data variant chains to the length of TL, and the value range is (0%, 100%); d(2) in the multi-feature digital similarity chain is set as the identifier of the sub-feature data and is not adjusted or updated with the algorithm. After each comparison, the feature data variant chain in d(3) is judged whether to be updated and stored according to the update status identifier of d(5) as the TL for the next similarity calculation; the similarity T between TL and TL' calculated by LCS is stored in d(4) for judging whether the feature data variant chain in d(3) needs to be updated. At the same time, d(5) represents the identifier of whether the feature data variant chain in d(3) has been updated.

2. The method for establishing a multi-feature digital twin model of a smart construction site according to claim 1, characterized in that: The method arranges sensors at locations where personnel, machines, materials, processes and environments on the construction site need to be monitored or supervised, and uses different types of acquisition devices to collect multi-feature data, wherein the sensors include biometric acquisition, image acquisition, sound acquisition, vibration acquisition, current and voltage working parameter acquisition and environmental data acquisition devices.

3. The method for establishing a multi-feature digital twin model of a smart construction site according to claim 1, characterized in that: The method receives data collected by sensors through a digital twin processing module and converts the data into dynamic model information.

4. A multi-feature digital twin model management system for a smart construction site, which is implemented by the method for establishing a multi-feature digital twin model of a smart construction site as described in claim 1, and includes multiple modules: a multi-feature data acquisition module, a digital twin model establishment module and a central processing unit; the system collects multi-feature data of personnel, machines, materials, processes and environment on the construction site through sensors, and uses the digital twin model establishment module to convert it into dynamic model information; the central processing unit uses a chain comparison algorithm to perform real-time dynamic adjustment on the multi-feature digital similarity chain to realize the storage, comparison and update of model information.

5. The management system according to claim 4, characterized in that: The central processing unit includes a model information storage module and a model information comparison module, which can compare the similarity between TL and TL' through a chain comparison algorithm. The chain comparison algorithm uses the LCS dynamic programming algorithm to recursively solve the similarity.

6. The management system according to claim 4, characterized in that: The chain comparison algorithm determines the similarity by comparing the new feature data variant chain with the corresponding feature data variant chain in the model information; when the maximum similarity value between the new feature data variant chain and the feature data variant chain is higher than a preset threshold, the new feature data variant chain is stored.

7. The management system according to claim 4, characterized in that: The chain comparison algorithm dynamically adjusts the preset threshold of similarity based on the actual environment.

Images