Intelligent logistics project bidding and tendering method and system based on meta universe

Abstract

The invention relates to the technical field of virtual space, in particular to an intelligent logistics project bidding and tendering method and system based on meta-universe, and the method achieves the multi-dimensional presentation of project information through the technologies of three-dimensional scene modeling and the like through the technical advantages of immersive experience, high interactivity, infinite expansibility, digital body, high sociability and high economic value of meta-universe. The immersive display interaction effect is realized, and the user perception is enhanced; creating a virtual image and an interaction link to ensure information symmetry and fairness; various bidding data are processed by means of a multi-modal fusion neural network algorithm, the bid evaluation efficiency and accuracy are improved, and a better, more efficient and just solution is provided for intelligent logistics project bidding and tendering.

Description

Technical Field

[0001] The present invention relates to the field of virtual space technology, and in particular to a method and system for bidding for intelligent logistics projects based on the metaverse. Background Art

[0002] With the continuous advancement of information technology, the gradual rise of online bidding technology has brought new opportunities for bidding. By building an online bidding system, it is helpful to improve the efficiency, transparency and fairness of traditional bidding and reduce the cost of bidding. In the existing online bidding system, through the Internet, database, blockchain, electronic signature and seal, intelligent bidding and other technologies, the electronic management and operation of the bidding process are realized, which can improve the efficiency, transparency and fairness of bidding to a certain extent, reduce the cost of bidding, and also enhance the security of information.

[0003] At present, the online bidding system can only display project information in a flat manner through a web interface. Existing solutions mostly rely on traditional two-dimensional interfaces to display bidding information, such as documents, pictures, etc. The solution display lacks an immersive effect, making it difficult for users to fully perceive the actual situation of complex logistics projects.

[0004] Existing technologies make it difficult for users to deeply integrate into the bidding process. Users often just passively receive information and submit documents. The interactivity is not strong, resulting in a lack of user participation, information asymmetry, and failure to ensure fairness.

[0005] The intelligent bid evaluation algorithm of the existing platform mainly processes structured numerical data in electronic bid documents, such as bid price, technical parameters and other indicator parameters. For multimodal unstructured data such as solution videos, renderings, 3D models and other multimodal data in the bid documents, it currently relies mainly on manual work, resulting in low bid evaluation efficiency. Summary of the invention

[0006] In view of this, the purpose of the present invention is to provide a smart logistics project bidding method and system based on the metaverse to overcome the problems existing in the current prior art.

[0007] To achieve the above objectives, the present invention adopts the following technical solutions: On the one hand, the present application provides a bidding method for a smart logistics project based on the Metaverse, comprising: Obtain user information, and create a Metaverse account and a corresponding virtual image in the Metaverse platform according to the user information; wherein the users include: tenderers, bidders, bid evaluation committee members and notaries; Obtaining the operation instructions of the tenderer, creating a virtual bidding three-dimensional space in the Metaverse platform according to the operation instructions, and uploading the digital bidding documents prepared by the tenderer; Publishing a bidding notice in the Metaverse platform according to the digital bidding document; Obtaining the registration instruction of the bidder after receiving the tender notice, registering for bidding in the Metaverse platform according to the registration instruction, and uploading the digital bidding document prepared by the bidder; When the time set in the tender notice arrives, the tenderer, the bid evaluation committee members and the notary are invited to enter the virtual bidding three-dimensional space; Obtain the evaluation results of all digital bidding documents by the evaluation committee members in combination with the intelligent evaluation model pre-created in the Metaverse Intelligent Bidding Evaluation Service of the Metaverse Platform; According to the bid evaluation results, a successful bidder is determined among the bidders, and the tenderer and the successful bidder complete a virtual contract signing in a virtual three-dimensional bidding space.

[0008] Furthermore, the method described above, wherein obtaining the operation instruction of the tenderer, creating a virtual bidding three-dimensional space in the Metaverse platform according to the operation instruction, and uploading the digital bidding document prepared by the tenderer, includes: Obtaining the operation instructions of the tenderer; According to the operation instructions, create an enterprise-specific bidding document display space, virtual bidding opening space, virtual bidding evaluation space and virtual signing space on the Metaverse platform developed based on the Unity 3D engine, and provide an entrance; Obtaining the digital bidding document prepared by the tenderer, and uploading the digital bidding document to the bidding document display space for display; Obtain the bidding qualification standards and evaluation criteria determined by the tenderer, and use the SHA-256 data encryption algorithm to store the bidding qualification standards and evaluation criteria on the blockchain node of the Metaverse platform.

[0009] Furthermore, in the method described above, the digital bidding document includes: electronic bidding document data, video and rendering data, and three-dimensional model multimodal data.

[0010] Furthermore, the method described above, when the time set in the tender notice is reached, inviting the tenderer, the bidder, the bid evaluation committee member and the notary to enter the virtual bidding three-dimensional space includes: After the time set in the tender notice arrives, the tenderer, the bidder and the notary are invited to enter the virtual bid opening space; Invite the bid evaluation committee members to enter the virtual bid evaluation space.

[0011] Furthermore, in the above method, creating the intelligent bid evaluation model includes: Constructing a multimodal data preprocessing and feature extraction module to perform data preprocessing and feature extraction on the electronic bidding document data, video and rendering data, and three-dimensional model multimodal data of the digital bidding document; Build a multi-modal fusion neural network architecture with multiple branches to complete the construction of an intelligent bidding evaluation model; The intelligent bidding evaluation model is trained, and the trained intelligent bidding evaluation model is deployed to the Metaverse platform.

[0012] Furthermore, in the above method, the electronic bidding document data, video and rendering data, and three-dimensional model multimodal data of the digital bidding document are subjected to data preprocessing and feature extraction, including: Extracting structured data of the electronic tender document data; Cleaning the structured data and converting the data format; Analyze the text description of the electronic tender document data using natural language processing technology, extract key features, and generate an electronic tender document data vector; Processing the video data using image recognition and video analysis techniques to extract key frames from the video data; Performing image feature extraction on the key frame and effect picture data to generate the video and effect picture data vectors; Performing geometric feature analysis on the multimodal data of the three-dimensional model, calculating the volume, surface area and center of gravity of the model, and determining the topological structure characteristics of the model; The multimodal data vector of the three-dimensional model is generated according to the volume, surface area, center of gravity position and topological structure characteristics of the model.

[0013] Furthermore, the method described above, wherein a multimodal fusion neural network architecture including multiple branches is constructed to complete the construction of an intelligent bid evaluation model, comprises: Constructing branch neural networks for electronic bidding document data, branch neural networks for video image data, and branch neural networks for 3D model data respectively; Using a feature splicing fusion method, the outputs of the electronic tender data branch neural network, the video image data branch neural network and the three-dimensional model data branch neural network are fused at a high level of the network to generate a fused feature vector; According to the fused feature vector and the weights and biases of the fully connected layer, the intelligent bid evaluation model is constructed.

[0014] Furthermore, in the above method, the training of the intelligent bid evaluation model and the deployment of the trained intelligent bid evaluation model to the Metaverse platform include: Get historical bidding data; Dividing the historical bidding data into a training set, a validation set, and a test set; The intelligent bid evaluation model is trained by the training set, and the intelligent bid evaluation model is verified by the verification set; When training the intelligent bidding evaluation model, a loss function is defined for the intelligent bidding evaluation model, and a stochastic gradient descent optimization algorithm is used. Updating the parameters of the intelligent bid evaluation model, and using regularization technology to limit the size of the parameters of the intelligent bid evaluation model; Introducing a cross-modal interaction module into the network structure of the intelligent bidding evaluation model, realizing information exchange between different modal data through an attention mechanism, so that the intelligent bidding evaluation model can learn the association and complementary information between different modal data; The intelligent bid evaluation model after verification is tested through the test set, and after passing the test, the intelligent bid evaluation model is deployed to the Metaverse platform.

[0015] On the other hand, the present application provides a Metaverse-based intelligent logistics project bidding system, including: an infrastructure layer, a data layer, a Metaverse scenario layer, a Metaverse service layer, and a user interface layer; The infrastructure layer includes data collection facilities, network infrastructure, computing resources, storage devices and interactive devices; The data collection facilities include human motion capture somatosensory equipment, language recognition somatosensory equipment, expression recognition somatosensory equipment and bidding document data collection equipment; The network infrastructure provides the network connection required for system operation; The computing resources include server clusters and cloud computing platforms, which provide processing capabilities for various complex computing tasks of the system; The storage device is used to store bidding data; The interactive device is used to provide human-computer interaction hardware between the user and the Metaverse bidding system; The data layer is responsible for data storage and processing, including a data storage unit and a data processing unit; The metaverse scene layer is used to generate virtual characters and virtual scenes; The Metaverse service layer is used to provide users with virtual reality services, bidding management services, intelligent bid evaluation services, user management services and data security services; The user interface layer provides different user access methods, including XR interface, Web interface and mobile interface.

[0016] The beneficial effects of the present invention are: This application has constructed a bidding method and system for intelligent logistics projects based on the Metaverse through the Metaverse's immersive experience, strong interactivity, unlimited scalability, digital avatars, strong sociality and high economic value technical advantages. It uses technologies such as three-dimensional scene modeling to present project information in multiple dimensions, achieve immersive display interaction effects, and enhance user perception; create virtual images and interactive links to ensure information symmetry and fairness; use multimodal fusion neural network algorithms to process various types of bidding data, improve bid evaluation efficiency and accuracy, and bring better quality, efficient and fair solutions to the bidding of intelligent logistics projects. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0018] Figure 1 It is a flow chart provided by an embodiment of a bidding method for a smart logistics project based on the Metaverse of the present invention; Figure 2 It is a structural diagram provided by an embodiment of a bidding system for intelligent logistics projects based on the Metaverse of the present invention. DETAILED DESCRIPTION

[0019] To make the purpose, technical solution and advantages of the present invention clearer, the technical solution of the present invention will be described in detail below. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments of the present invention, all other implementation methods obtained by ordinary technicians in this field without creative work belong to the scope of protection of the present invention.

[0020] Figure 1 This is a flowchart provided by an embodiment of a bidding method for a smart logistics project based on the Metaverse of the present invention. Figure 1 , this embodiment may include the following steps: S1. Obtain user information and create a Metaverse account and a corresponding virtual image in the Metaverse platform based on the user information; the users include: tenderers, bidders, bid evaluation committee members and notaries; It should be noted that before conducting bidding activities, the tenderer, bidder, evaluation committee member, and notary public can use VR, PC, mobile devices and other terminal devices to use their ID cards to create a corporate or personal account in the Metaverse platform and create a personal real-name authenticated and adjustable virtual image.

[0021] S2. Obtain the tenderer's operating instructions, create a virtual bidding three-dimensional space in the Metaverse platform according to the operating instructions, and upload the digital bidding documents prepared by the tenderer; S3. Publish the bidding notice in the Metaverse platform according to the digital bidding documents; It should be noted that the tenderer publishes the tender notice for the smart logistics project on the Metaverse platform. The notice should cover key contents such as basic project information, registration guide, bidding deadline, tender document display space, virtual bid opening space, etc. After receiving the notice through Metaverse terminal devices such as mobile phones and computers, suppliers of smart logistics projects can register online, obtain detailed project tender documents, and enter the tender document display space through VR devices to learn about project details.

[0022] S4. Obtain the registration instructions of the bidder after receiving the tender notice, register for bidding in the Metaverse platform according to the registration instructions, and upload the digital bidding documents prepared by the bidder; It should be noted that after receiving the bidding announcement through the push message, search function and other channels of the Metaverse platform, the bidder should carefully read the content of the announcement to understand the basic situation and requirements of the project. If eligible for bidding, follow the guidelines in the announcement to register online and fill in the relevant company information and contact information.

[0023] After successful registration, the bidder pays the bid deposit to the tenderer and downloads detailed digital bidding documents for the project through the Metaverse platform. The documents should include detailed project requirements, technical specifications, bidding format, etc. The bidder organizes professionals to conduct in-depth research on the bidding documents and make full preparations for bidding.

[0024] Bidders form a virtual bidding team in the Metaverse, and conduct in-depth analysis and collaborative planning of smart logistics projects remotely through Metaverse terminal devices. With the real-time collaboration function of the Metaverse, team members can communicate and share information efficiently and jointly optimize the bidding plan for smart logistics projects. At the same time, digital bidding documents for smart logistics projects are prepared and uploaded to the designated area of ​​the Metaverse platform through encrypted channels.

[0025] Bidders form a virtual bidding team in the Metaverse according to the requirements of the smart logistics project. Team members can come from different departments and regions, and collaborate through Metaverse terminal devices and virtual identities in the Metaverse. At the same time, each team member has a unique virtual character ID for identity authentication, which cannot be tampered with. When logging in, the system automatically verifies whether you are a team member through technologies such as face recognition and voice recognition. Only after the verification is passed can the virtual image be activated to conduct team collaboration.

[0026] During the in-depth analysis of the project, team members made full use of the data analysis tools and visualization functions provided by the Metaverse platform to fully and deeply understand the background, goals and needs of the project. During the solution planning stage, team members conducted brainstorming and solution discussions through real-time collaboration functions to jointly develop the best bidding solution.

[0027] When preparing digital bidding documents, bidders use multimedia production technology to prepare digital bidding documents including electronic bidding documents, renderings, program animations, 3D models, etc. for logistics projects, so that the bidding documents can be displayed more vividly, comprehensively and intuitively in the metaverse space. After the digital bidding documents are completed, they should be stored in the bidder's local client, and ensure that the file storage path is clear and easy to find for subsequent upload operations.

[0028] S5. When the time set in the tender notice arrives, invite the tenderer, bidders, bid evaluation committee members and notary public to enter the virtual bidding three-dimensional space; S6. Obtain the evaluation results of all digital bidding documents by the evaluation committee in combination with the intelligent evaluation model pre-created in the Metaverse Intelligent Evaluation Service of the Metaverse Platform; It is understandable that a bid evaluation committee is formed to evaluate the bid proposals. During the bid evaluation, the bid evaluation committee members enter the virtual bid evaluation space of the Metaverse, use interactive tools and equipment to carefully view the three-dimensional models, multimedia presentations and other contents in the bid documents, and enable the Metaverse intelligent bid evaluation service. The Metaverse intelligent bid evaluation service conducts a preliminary quantitative evaluation based on the preset bid evaluation criteria and the multimodal fusion deep neural network algorithm. The bid evaluation committee members conduct a qualitative evaluation of the bid proposals based on the results of the algorithm evaluation and professional judgment, and conduct real-time discussions and exchanges in the Metaverse space to ensure the accuracy and fairness of the bid evaluation results.

[0029] The virtual bidding evaluation space in the Metaverse includes virtual meeting rooms, solution display areas, and intelligent bidding evaluation service areas, which support bid evaluation committee members to access the space smoothly and are equipped with corresponding interactive tools and equipment. Bid evaluation committee members can use virtual reality glasses, handles and other devices to carefully view virtual models, multimedia presentations and other content in the bidding documents, such as logistics distribution route planning videos, warehouse layout 3D models, enterprise qualification certification documents and other types of data. Virtual models can be zoomed in, out, rotated and other operations, which facilitate bid evaluation committee members to observe the design and implementation details of the project from different angles. In this space, bid evaluation committee members can discuss and communicate in real time through virtual images, share their views and opinions, and jointly explore the advantages, disadvantages and feasibility of the bidding plan. During the discussion, the system automatically records the speeches and opinions of experts to provide a basis for subsequent bid evaluation reports.

[0030] The Metaverse Intelligent Bid Evaluation Service Area is used for intelligent bid evaluation of bidding documents. It has a built-in multimodal fusion deep neural network algorithm that can be activated by experts through specific language or gesture commands. It is used to analyze multimodal data in digital bidding documents, including electronic bid documents, videos, renderings, and three-dimensional models, and compare them with preset evaluation criteria to quickly give a comprehensive and integrated quantitative evaluation.

[0031] S7. Based on the bid evaluation results, the successful bidder is determined among the bidders, and the tenderer and the successful bidder complete the virtual signing in the virtual bidding three-dimensional space.

[0032] It is understandable that, based on the final evaluation results, the bid document with the highest score will be determined as the winning candidate and will be announced on the bidding announcement space of the tenderer's enterprise Metaverse platform. During the announcement period, if other bidders have objections, they can raise inquiries and complaints through the Metaverse platform, and the tenderer and the bid evaluation committee will review and respond in a timely manner.

[0033] After the bid is determined, the tenderer and the successful bidder hold a contract signing ceremony in a virtual signing space through virtual images. When signing the contract, the smart contract technology based on blockchain is used to sign the contract to ensure that the content and terms of the contract cannot be unilaterally tampered with. Within five working days after the contract is signed, the tenderer will return the bid deposit to the successful bidder and the unsuccessful bidder in one lump sum.

[0034] It is understandable that in view of the shortcomings of existing online bidding technology, this application uses the technical advantages of Metaverse immersive experience, strong interactivity, unlimited scalability, digital avatar, strong sociality, high economic value, etc. to build a bidding method and system for intelligent logistics projects based on Metaverse, aiming to achieve immersive project display, enhance user participation and fairness, and achieve comprehensive intelligent bid evaluation. By constructing a bidding method and system based on Metaverse, using technologies such as three-dimensional scene modeling to present project information in multiple dimensions, achieving immersive display interaction effects and enhancing user perception; creating virtual images and interactive links to ensure information symmetry and fairness; using multimodal fusion neural network algorithms to process various types of bidding data, improve bid evaluation efficiency and accuracy, and bring better quality, efficient and fair solutions to the bidding of intelligent logistics projects.

[0035] Preferably, step S2 comprises: Obtain the tenderer's operating instructions; According to the operating instructions, create a company-specific bidding document display space, virtual bidding opening space, virtual bidding evaluation space and virtual signing space on the Metaverse platform developed based on the Unity 3D engine, and provide an entrance; Obtain the digital bidding documents prepared by the tenderer and upload the digital bidding documents to the bidding document display space for display; Obtain the bidding qualification standards and evaluation criteria determined by the tenderer, and use the SHA-256 data encryption algorithm to store the bidding qualification standards and evaluation criteria on the blockchain node of the Metaverse platform.

[0036] It is understandable that the tenderer creates a company-exclusive tender document display space, virtual bid opening space, virtual bid evaluation space, and virtual signing space on the Metaverse platform developed based on the Unity 3D engine, and provides an entrance so that relevant personnel can enter the corresponding space in a specific process to carry out bidding work.

[0037] The tenderer shall form a professional technical team, use 3D modeling, data visualization and virtual reality technology to produce digital tender documents, and upload them to the tender document display space. The digital tender documents can be parsed and displayed in the tender document display space to ensure that the logistics project information can be fully and accurately displayed. The general appearance, structure and layout of the project, such as the surrounding building environment, equipment and facilities, can be presented through 3D modeling and visualization technology, and the project technical indicators (such as beat, flow and other requirements) can be presented in a data visualization way, so that the bidder can design the scheme according to the technical requirements of the project. At the same time, the bidding qualification standards and evaluation criteria are clarified, and this information is stored through blockchain technology. The public chain blockchain type can be used, and the SHA-256 data encryption algorithm can be used to store this information on the blockchain node of the Metaverse platform. No party can unilaterally change this information to ensure that the data cannot be tampered with and is stored securely.

[0038] Preferably, the digital bidding document includes: electronic bidding document data, video and rendering data, and three-dimensional model multimodal data.

[0039] Preferably, step S5 includes: When the time set in the tender notice arrives, invite the tenderer, bidders and notary public to enter the virtual tender opening space; Invite bid evaluation committee members to enter the virtual bid evaluation space.

[0040] It is understandable that at the specified bid opening time, the tenderer and the bidder enter the virtual bid opening venue in the metaverse through VR equipment with virtual images, and the virtual characters can reflect human tone, expression and action, so as to achieve character interaction and enhance user participation. Then, a virtual bid opening ceremony is held in the virtual bid opening venue, and key information such as the bidder's name and bid price are announced in turn. The whole process is open to the notary to ensure the transparency and fairness of the bid opening process.

[0041] The virtual images of the tenderer and bidder should restore human expressions and movements as much as possible, and achieve natural dialogue and communication through voice recognition and synthesis technology, and realize real body movements through motion capture technology.

[0042] The bidder uploads the bid document to the designated bid opening area through the encrypted channel provided by the Metaverse platform, and the bidder opens it to ensure the security and confidentiality of the document. The bidder can view a certain proportion of the three-dimensional model of the proposal in the Metaverse space, and the three-dimensional model can be configured with the information in the bid document. The main technical parameters can be displayed through three-dimensional visualization elements. The renderings and the proposal video are placed in the same area of ​​the three-dimensional model for comparison.

[0043] In the virtual bidding venue, the host introduced the bidding process and participants, and then used a virtual large screen to display and voice broadcast key information such as the bidder's name, bid price, etc., to ensure that all participants could see and hear clearly.

[0044] The notary enters the virtual bidding venue with a virtual image through a special notary channel to supervise and notarize the entire bidding process. The notary can use interactive tools to check any link in the bidding process at any time, such as checking the encryption status of the bidding documents, verifying the identity of the bidder, etc. If any problem is found, the notary can ask questions through voice or text chat functions, and the tenderer and bidder will respond in real time and provide explanations.

[0045] Preferably, creating an intelligent bid evaluation model includes: Construct a multimodal data preprocessing and feature extraction module to perform data preprocessing and feature extraction on the electronic bidding document data, video and rendering data of the digital bidding documents, and the multimodal data of the three-dimensional model; Build a multi-modal fusion neural network architecture with multiple branches to complete the construction of an intelligent bidding evaluation model; Train the intelligent bidding evaluation model and deploy the trained intelligent bidding evaluation model to the Metaverse platform.

[0046] Preferably, the electronic bidding document data, video and rendering data, and three-dimensional model multimodal data of the digital bidding document are subjected to data preprocessing and feature extraction, including: Extract structured data from electronic tender data; Clean structured data and convert data formats; Use natural language processing technology to analyze the text description of electronic bidding data, extract key features, and generate electronic bidding data vectors; Use image recognition and video analysis technology to process video data and extract key frames from video data; Extract image features from key frames and renderings to generate video and renderings data vectors; Analyze the geometric features of the multimodal data of the three-dimensional model, calculate the volume, surface area and center of gravity of the model, and determine the topological structure characteristics of the model; A multimodal data vector of the three-dimensional model is generated according to the volume, surface area, center of gravity position and topological structure characteristics of the model.

[0047] It is understandable that digital bidding documents include multimodal data such as electronic tender data, video and rendering data, and three-dimensional models, and data preprocessing and feature extraction are required for the three types of data respectively.

[0048] ① Electronic bidding data processing For electronic tender documents, extract structured data such as bid price 、Enterprise Qualification (can be quantified by qualification level coding, etc.), technical parameters The data is cleaned to remove erroneous or incomplete data items and converted into a format suitable for algorithm processing, such as numerical or categorical coding. At the same time, natural language processing technology is used to analyze the text description in the electronic bidding document to extract key features, such as keywords of the project implementation plan. (Keyword weights can be determined by methods such as word frequency-inverse document frequency), description of advantages and characteristics, etc., in order to better understand the bidder's intentions and capabilities. After processing, the electronic bidding document data can be represented as a vector .

[0049] ②Video image data processing For Video And the effect diagram , using image recognition and video analysis techniques to extract key frames from the video , extract image features from key frames and renderings. By calculating the color histogram of the image , (in Represents the color channel Histogram distribution), and edge detection to obtain edge features , ( Represents a certain quantitative value of edge features), and obtains the feature vector of the image. The video and rendering data are processed and represented as vectors and (here and are the color histogram and edge features corresponding to the effect image).

[0050] ③3D model data processing For 3D models , perform geometric feature analysis and calculate the volume of the model , surface area , Center of gravity (can be expressed as three-dimensional coordinates ), as well as the topological structure characteristics of the model, such as the number of faces , Number of edges These features can reflect the complexity of the model and the characteristics of the spatial layout, and are used to analyze the conformity of the model with project requirements and the adaptability of performance indicators. After processing, the 3D model data can be represented as a vector .

[0051] Preferably, a multi-modal fusion neural network architecture containing multiple branches is constructed to complete the construction of an intelligent bid evaluation model, including: Constructing branch neural networks for electronic bidding document data, branch neural networks for video image data, and branch neural networks for 3D model data respectively; The outputs of the electronic bidding data branch neural network, the video image data branch neural network and the 3D model data branch neural network are fused at the high level of the network by using the feature splicing fusion method to generate a fusion feature vector. The intelligent bidding evaluation model is constructed based on the fused feature vector and the weights and bias of the fully connected layer.

[0052] It is understandable that a multimodal fusion neural network architecture is constructed with multiple branches, each branch specializing in processing data of one modality.

[0053] ① Construction of branch neural network for electronic bidding data For the electronic tender data branch, build a fully connected network Structure, the number of input layer nodes is based on the electronic bidding document feature vector The dimension of is determined and set to The number of hidden layers is set to , the number of nodes in each hidden layer decreases in sequence (such as , ,…, ), the activation function selects the ReLU function , the number of nodes in the output layer is determined according to the subsequent fusion requirements. This branch extracts features and performs preliminary processing on the electronic bidding document data. The formula is expressed as: , where represents the output of the hidden layer ( ), , and are the weight and bias of the hidden layer respectively.

[0054] ② Construction of branch neural network for video image data For the video and rendering data branches, a convolutional neural network (CNN) structure is used. Taking the processing of video key frame data as an example, the input image data (Assuming the key frame image size is , is the height , is the width, is the number of channels), the convolution layer uses multiple convolution kernels of different sizes (such as ) performs convolution operation, and the number of convolution layers is , the pooling layer adopts the maximum pooling operation, and the pooling kernel size is This branch extracts image features through convolution and pooling operations, and the formula is expressed as: ,

[0055] in, Indicates The output of the convolutional layer ( ), , For the The convolution kernel of the convolution layer, is the bias, For the The output of the pooling layer, is the activation function (such as ReLU function). Finally, the extracted image features are mapped to the same dimensional space as other modal features through the fully connected layer. Suppose the number of nodes in the fully connected layer is , the output is Similarly, the effect diagram output can be obtained as

[0056] ③ Construction of branch neural network for 3D model data For the 3D model data branch, a graph neural network (GNN) can be used to analyze its geometric and topological structure. Suppose the input 3D model feature vector After multi-layer message passing and update operations of GNN, the processed features are obtained , the GNN update formula is:

[0057] in, time step, is an aggregate function, A function to obtain neighbor node information. A function that updates the node status.

[0058] ④Neural network high-level fusion The feature splicing fusion method is used to fuse the outputs of each branch at the high level of the network. The spliced ​​feature vector , and then conduct a comprehensive evaluation through the subsequent fully connected layer, and finally output the score or evaluation result of the bidding plan The calculation process of the fully connected layer can be expressed as:

[0059] in is the fused feature vector, and are the weights and biases of the fully connected layer, is the activation function.

[0060] Preferably, the intelligent bid evaluation model is trained and deployed to the Metaverse platform, including: Get historical bidding data; Divide the historical bidding data into training set, validation set and test set; The intelligent bid evaluation model is trained through the training set, and the intelligent bid evaluation model is verified through the verification set; When training the intelligent bidding evaluation model, define the loss function for the intelligent bidding evaluation model and use the stochastic gradient descent optimization algorithm Update the parameters of the intelligent bid evaluation model and use regularization technology to limit the size of the parameters of the intelligent bid evaluation model; Introducing a cross-modal interaction module into the network structure of the intelligent bidding evaluation model, the information exchange between different modal data is realized through the attention mechanism, so that the intelligent bidding evaluation model can learn the association and complementary information between different modal data; The verified intelligent bidding evaluation model is tested through the test set, and after passing the test, the intelligent bidding evaluation model is deployed to the Metaverse platform.

[0061] It is understandable that a large amount of historical bidding data is used to train the constructed neural network. The data set is divided into training set, validation set and test set, and the data can be allocated in a ratio of 70%:15%:15%. When training the model, define the loss function , such as mean square error ,in is the sample size, is the evaluation score predicted by the model, is the true label or value.

[0062] Use stochastic gradient descent optimization algorithm Update the model parameters, the update formula is , in is the learning rate, Represents model parameters (including weights and biases of each branch network). Set appropriate hyperparameters such as learning rate and number of iterations according to data size and model complexity.

[0063] To prevent the model from overfitting, regularization technology is used to limit the size of model parameters and prevent the model from being too complex. The Dropout method is used to randomly discard some neurons with a certain probability during the training process to increase the robustness of the model. The performance of the model on the validation set is monitored by the early stopping method. When the performance no longer improves, the training is stopped to prevent the model from overfitting the training data. In addition, at the end of each round of training, the evaluation indicators of the model on the validation set, such as accuracy and recall, are calculated. If the performance of the validation set does not improve for multiple consecutive rounds, the training is stopped. According to the evaluation indicators on the validation set, the model hyperparameters are adjusted, and the optimal hyperparameter combination is found through random search to improve the generalization ability of the model.

[0064] Introducing a cross-modal interaction module into the network structure, the information exchange between different modal data is realized through the attention mechanism, so that the model can learn the association and complementary information between different modal data. For example, establishing an information interaction channel between the electronic bidding data branch and the video data branch, we have: , in Indicates the characteristics of electronic bidding documents With video features The attention weights between is the attention function, which can be obtained by calculating the similarity between the two), and then the video features are weighted and summed to obtain the fusion features. , and compare it with the characteristics of electronic bidding documents For splicing or further fusion processing.

[0065] After the algorithm model training is completed, it can be deployed to the cloud platform or local server.

[0066] In the actual bidding evaluation process, when the bid evaluation committee selects a digital bid and starts the intelligent bid evaluation service, the model can automatically process and analyze the multimodal data such as the electronic bid, video, renderings and 3D models in the bidding documents, and output 3D visualized comprehensive evaluation results and related analysis reports. Experts can quickly locate the advantages and disadvantages of the bidding plan based on the quantitative evaluation results of the model, conduct qualitative evaluation, and give the final evaluation results.

[0067] Figure 2 This is a schematic diagram of the structure of an embodiment of a bidding system for intelligent logistics projects based on Metaverse provided by the present invention. Figure 2 , this embodiment may include: an infrastructure layer, a data layer, a metaverse scenario layer, a metaverse service layer, and a user interface layer; The infrastructure layer includes data collection facilities, network infrastructure, computing resources, storage devices, and interactive devices; Data collection facilities include human motion capture somatosensory equipment, language recognition somatosensory equipment, expression recognition somatosensory equipment and bidding document data collection equipment; Network infrastructure provides the network connections required for system operation; Computing resources include server clusters and cloud computing platforms, which provide processing capabilities for various complex computing tasks of the system; The storage device is used to store the bidding data; The interactive device is used to provide human-computer interaction hardware between users and the Metaverse bidding system; The data layer is responsible for data storage and processing, including data storage units and data processing units; The metaverse scene layer is used to generate virtual characters and virtual scenes; The Metaverse service layer is used to provide users with virtual reality services, bidding management services, intelligent bid evaluation services, user management services, and data security services; The user interface layer provides different user access methods, including XR interface, web interface and mobile interface.

[0068] It can be understood that the infrastructure layer is the basic supporting part of the entire system, including data collection facilities, network infrastructure, computing resources and storage devices.

[0069] Data collection facilities include body motion capture, language recognition, expression recognition and other somatosensory equipment, as well as bidding document data collection equipment.

[0070] Network infrastructure provides the network connections required for system operation, ensuring smooth transmission of data between different devices and users. High-bandwidth and low-latency networks are essential for real-time interactions in the Metaverse.

[0071] Computing resources include server clusters and cloud computing platforms, which provide processing power for various complex computing tasks of the system. Among them, the server cluster is the computing core of the system, including high-performance physical servers and flexible cloud servers, which together provide support for complex computing tasks of the Metaverse. The cloud computing platform adopts a distributed computing architecture, which distributes computing tasks to different nodes for parallel processing to avoid single point failures, while speeding up computing speed to meet the real-time and high throughput requirements of the Metaverse bidding system.

[0072] Storage devices are used to store large amounts of bidding data, including tender documents, bid documents, user information, etc., to ensure safe and reliable storage of data and enable rapid retrieval and access.

[0073] Interactive devices are used to provide human-computer interaction hardware between users and the Metaverse bidding system, including VR, AR, MR devices, PC devices and mobile devices.

[0074] The data layer is responsible for data storage and processing, including data storage units and data processing units. The data storage unit uses database management technology to classify and store structured and unstructured data, and uses data encryption technology to ensure data security and privacy. The data processing unit cleans, converts and analyzes the data generated during the bidding process, mines valuable information, and provides support for intelligent bid evaluation and decision-making.

[0075] The metaverse scene layer provides immersive experience, efficient interaction, accurate decision-making support and intuitive data display for bidding participants through virtual characters, virtual scenes, intelligent algorithms and visualization elements, thereby improving the efficiency and quality of the bidding process.

[0076] Virtual characters include tenderers, bidders, bid evaluation committee members, notaries, etc. Everyone can customize a unique virtual image according to their preferences and personality, which increases the fun and immersion of participation. Virtual characters can communicate and interact in the virtual environment through voice, expression and action, making it convenient for all parties to express their views and discuss solutions.

[0077] The virtual scene highly restores the actual bidding environment, including the surrounding landforms, building distribution, and project virtual models of the virtual environment such as bid opening and bid evaluation, which helps participants better understand the project background and requirements. Objects in the virtual scene can interact with virtual characters. For example, the bidder's character can operate the virtual model to display the details of the plan, and the bid evaluation committee can conduct on-site inspections and give feedback. The system can also dynamically adjust the scene based on these feedbacks, continuously optimize the experience of participants, and improve their satisfaction with the bidding process.

[0078] Intelligent algorithms include bid evaluation algorithms, which evaluate the pros and cons and feasibility of different bidding proposals through comprehensive and in-depth analysis of bidding proposal data, providing objective and accurate decision-making basis for bid evaluation committee members.

[0079] Data visualization elements can transform complex bidding content into an intuitive and easy-to-understand form, including various elements such as charts, graphics, three-dimensional models that are intuitively presented in the metaverse virtual space, allowing participants to grasp the key content of the bidding plan more quickly and improve the efficiency of information transmission.

[0080] The Metaverse service layer is used to provide users with various key services, including virtual reality services, bidding management services, intelligent bid evaluation services, user management services, data security services, etc.

[0081] Virtual reality services are used to build virtual bidding venues and exhibition spaces in the metaverse, using XR and 3D modeling technologies to provide immersive interactive experiences. Users can use motion capture, language recognition, gesture recognition and other technologies to interact with virtual images in virtual bidding venues, making it easier for bidders, bid evaluation committee members, notaries and others to communicate and collaborate holographically during the bidding process.

[0082] The bidding management service covers all aspects of the bidding process for the tenderer and the bidder, such as issuing bidding notices, downloading bidding documents, submitting bidding documents, opening bids, evaluating bids and determining bids, etc. It also includes functions such as project management and contract management to ensure the smooth progress of bidding projects.

[0083] The intelligent bid evaluation service uses a multimodal fusion neural network algorithm to automatically analyze bid documents, extract key information, and score and rank according to the bid evaluation criteria. At the same time, it uses data visualization technology to provide visual bid evaluation and display the results in the form of holographic charts and reports, which facilitates the decision-making of tenderers and bid evaluation committee members.

[0084] The user management service is responsible for user registration, login, identity authentication and permission management, ensuring that only legitimate users can access the corresponding functions and data of the system. Different permissions are assigned according to user roles to provide personalized user interfaces and services.

[0085] Data security services are used to encrypt user data to ensure security during transmission and storage. At the same time, data access behavior and data operations are audited and monitored to promptly detect and handle security incidents.

[0086] The user interface layer provides different user access methods, including XR interface, Web interface, and mobile interface.

[0087] The XR interface provides users with an immersive virtual bidding environment through virtual reality, augmented reality or mixed reality devices. It allows users to easily operate various elements of the virtual scene in an intuitive and natural interactive way, and interact with other user roles through virtual characters in the virtual bidding venue and exhibition space, with a high sense of reality.

[0088] The web interface provides traditional browser access methods. Users can access various functions and services of the system by logging into the web page, ensuring the applicability and convenience of the system.

[0089] Mobile interface Through the mobile application of smartphones or tablets, users can access the system anytime and anywhere, which is convenient and convenient, allowing users to work on the move and view bidding process information in real time.

[0090] There is data interaction among the above layers. The infrastructure layer collects user data and transmits it to the data layer for storage through the network infrastructure. After processing, it is used by the relevant modules of the service layer. At the same time, the application layer calls the service layer function and interacts with the user interface layer to display the results. In addition, there is a mapping relationship between the system architecture and the business process. In the early stage of bidding, the infrastructure layer helps the tenderer to build the space, the data layer stores information, the service layer assists in inputting settings, and the user interface layer provides an operation interface; in the bidding stage, the bidder enters the system through the user interface layer, collaborates with the team with the help of the infrastructure layer equipment, and uploads the bidding documents and conducts intelligent bid evaluation with the support of the service layer. The data layer is responsible for storing and managing related data.

[0091] It can be understood that the same or similar parts of the above embodiments can be referenced to each other, and the contents not described in detail in some embodiments can refer to the same or similar contents in other embodiments.

[0092] It should be noted that, in the description of the present invention, the terms "first", "second", etc. are only used for descriptive purposes and cannot be understood as indicating or implying relative importance. In addition, in the description of the present invention, unless otherwise specified, the meaning of "plurality" refers to at least two.

[0093] Any process or method description in a flowchart or otherwise described herein may be understood to represent a module, segment or portion of code that includes one or more executable instructions for implementing the steps of a specific logical function or process, and the scope of the preferred embodiments of the present invention includes alternative implementations in which functions may not be performed in the order shown or discussed, including performing functions in a substantially simultaneous manner or in the reverse order depending on the functions involved, which should be understood by those skilled in the art to which the embodiments of the present invention belong.

[0094] It should be understood that the various parts of the present invention can be implemented by hardware, software, firmware or a combination thereof. In the above-mentioned embodiments, multiple steps or methods can be implemented by software or firmware stored in a memory and executed by a suitable instruction execution system. For example, if implemented by hardware, as in another embodiment, it can be implemented by any one of the following technologies known in the art or their combination: a discrete logic circuit having a logic gate circuit for implementing a logic function for a data signal, a dedicated integrated circuit having a suitable combination of logic gate circuits, a programmable gate array (PGA), a field programmable gate array (FPGA), etc.

[0095] A person skilled in the art may understand that all or part of the steps in the method for implementing the above-mentioned embodiment may be completed by instructing related hardware through a program, and the program may be stored in a computer-readable storage medium, which, when executed, includes one or a combination of the steps of the method embodiment.

[0096] In addition, each functional unit in each embodiment of the present invention may be integrated into a processing module, or each unit may exist physically separately, or two or more units may be integrated into one module. The above-mentioned integrated module may be implemented in the form of hardware or in the form of a software functional module. If the integrated module is implemented in the form of a software functional module and sold or used as an independent product, it may also be stored in a computer-readable storage medium.

[0097] The storage medium mentioned above can be a read-only memory, a magnetic disk or an optical disk, etc.

[0098] In the description of this specification, the description with reference to the terms "one embodiment", "some embodiments", "examples", "specific examples", or "some examples" means that the specific features, structures, materials or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present invention. In this specification, the schematic representation of the above terms does not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described may be combined in any one or more embodiments or examples in a suitable manner.

[0099] Although the embodiments of the present invention have been shown and described above, it is to be understood that the above embodiments are exemplary and are not to be construed as limitations of the present invention. A person skilled in the art may change, modify, replace and vary the above embodiments within the scope of the present invention.

Claims

1. A bidding method for intelligent logistics projects based on Metaverse, characterized in that: include: Obtain user information, and create a Metaverse account and a corresponding virtual image in the Metaverse platform according to the user information; wherein the users include: tenderers, bidders, bid evaluation committee members and notaries; Obtaining the operation instructions of the tenderer, creating a virtual bidding three-dimensional space in the Metaverse platform according to the operation instructions, and uploading the digital bidding documents prepared by the tenderer; Publishing a bidding notice in the Metaverse platform according to the digital bidding document; Obtaining the registration instruction of the bidder after receiving the tender notice, registering for bidding in the Metaverse platform according to the registration instruction, and uploading the digital bidding document prepared by the bidder; When the time set in the tender notice arrives, the tenderer, the bid evaluation committee members and the notary are invited to enter the virtual bidding three-dimensional space; Obtain the evaluation results of all digital bidding documents by the evaluation committee members in combination with the intelligent evaluation model pre-created in the Metaverse Intelligent Bidding Evaluation Service of the Metaverse Platform; According to the bid evaluation results, a successful bidder is determined among the bidders, and the tenderer and the successful bidder complete a virtual contract signing in a virtual three-dimensional bidding space.

2. The method according to claim 1, characterized in that The obtaining of the operation instruction of the tenderer, creating a virtual bidding three-dimensional space in the Metaverse platform according to the operation instruction, and uploading the digital bidding document prepared by the tenderer, includes: Obtaining the operation instructions of the tenderer; According to the operation instructions, create an enterprise-specific bidding document display space, virtual bidding opening space, virtual bidding evaluation space and virtual signing space on the Metaverse platform developed based on the Unity 3D engine, and provide an entrance; Obtaining the digital bidding document prepared by the tenderer, and uploading the digital bidding document to the bidding document display space for display; Obtain the bidding qualification standards and evaluation criteria determined by the tenderer, and use the SHA-256 data encryption algorithm to store the bidding qualification standards and evaluation criteria on the blockchain node of the Metaverse platform.

3. The method according to claim 2, characterized in that The digital bidding documents include: electronic bidding document data, video and rendering data, and three-dimensional model multimodal data.

4. The method according to claim 3, characterized in that: When the time set in the tender notice is reached, inviting the tenderer, the bidder, the bid evaluation committee member and the notary to enter the virtual bidding three-dimensional space includes: After the time set in the tender notice arrives, the tenderer, the bidder and the notary are invited to enter the virtual bid opening space; Invite the bid evaluation committee members to enter the virtual bid evaluation space.

5. The method according to claim 4, characterized in that Creating the intelligent bid evaluation model includes: Constructing a multimodal data preprocessing and feature extraction module to perform data preprocessing and feature extraction on the electronic bidding document data, video and rendering data, and three-dimensional model multimodal data of the digital bidding document; Build a multi-modal fusion neural network architecture with multiple branches to complete the construction of an intelligent bidding evaluation model; The intelligent bidding evaluation model is trained, and the trained intelligent bidding evaluation model is deployed to the Metaverse platform.

6. The method according to claim 5, characterized in that The electronic bidding document data, video and rendering data, and three-dimensional model multimodal data of the digital bidding document are subjected to data preprocessing and feature extraction, including: Extracting structured data of the electronic tender document data; Cleaning the structured data and converting the data format; Analyze the text description of the electronic tender document data using natural language processing technology, extract key features, and generate an electronic tender document data vector; Processing the video data using image recognition and video analysis techniques to extract key frames from the video data; Performing image feature extraction on the key frame and effect picture data to generate the video and effect picture data vectors; Performing geometric feature analysis on the multimodal data of the three-dimensional model, calculating the volume, surface area and center of gravity of the model, and determining the topological structure characteristics of the model; The multimodal data vector of the three-dimensional model is generated according to the volume, surface area, center of gravity position and topological structure characteristics of the model.

7. The method according to claim 6, characterized in that The method of constructing a multi-modal fusion neural network architecture containing multiple branches to complete the construction of an intelligent bidding evaluation model includes: Constructing branch neural networks for electronic bidding document data, branch neural networks for video image data, and branch neural networks for 3D model data respectively; Using a feature splicing fusion method, the outputs of the electronic tender data branch neural network, the video image data branch neural network and the three-dimensional model data branch neural network are fused at a high level of the network to generate a fused feature vector; According to the fused feature vector and the weights and biases of the fully connected layer, the intelligent bid evaluation model is constructed.

8. The method according to claim 7, characterized in that The step of training the intelligent bid evaluation model and deploying the trained intelligent bid evaluation model to the Metaverse platform includes: Get historical bidding data; Dividing the historical bidding data into a training set, a validation set, and a test set; The intelligent bid evaluation model is trained by the training set, and the intelligent bid evaluation model is verified by the verification set; When training the intelligent bidding evaluation model, a loss function is defined for the intelligent bidding evaluation model, and a stochastic gradient descent optimization algorithm is used. Updating the parameters of the intelligent bid evaluation model, and using regularization technology to limit the size of the parameters of the intelligent bid evaluation model; A cross-modal interaction module is introduced into the network structure of the intelligent bidding evaluation model, and information exchange between different modal data is realized through the attention mechanism, so that the intelligent bidding evaluation model can learn the association and complementary information between different modal data; The intelligent bid evaluation model after verification is tested through the test set, and after passing the test, the intelligent bid evaluation model is deployed to the Metaverse platform.

9. A Metaverse-based intelligent logistics project bidding system, characterized in that: include: Infrastructure layer, data layer, metaverse scenario layer, metaverse service layer, and user interface layer; The infrastructure layer includes data collection facilities, network infrastructure, computing resources, storage devices and interactive devices; The data collection facilities include human motion capture somatosensory equipment, language recognition somatosensory equipment, expression recognition somatosensory equipment and bidding document data collection equipment; The network infrastructure provides the network connection required for system operation; The computing resources include server clusters and cloud computing platforms, which provide processing capabilities for various complex computing tasks of the system; The storage device is used to store bidding data; The interactive device is used to provide human-computer interaction hardware between the user and the Metaverse bidding system; The data layer is responsible for data storage and processing, including a data storage unit and a data processing unit; The metaverse scene layer is used to generate virtual characters and virtual scenes; The Metaverse service layer is used to provide users with virtual reality services, bidding management services, intelligent bid evaluation services, user management services and data security services; The user interface layer provides different user access methods, including XR interface, Web interface and mobile interface.

Images