Architectural design dynamic process management system based on Internet

Abstract

The invention relates to the technical field of building design, and discloses an Internet-based building design dynamic process management system, which comprises a multi-factor identity verification module, a distributed data storage module, a process dynamic optimization engine, a behavior evaluation and intervention module and an intelligent audit tracking system. According to the building design dynamic process management system based on the Internet, through cooperative work of the process dynamic optimization engine and the behavior evaluation and intervention module, the problem that task optimization of a traditional process management system depends on a preset rule is solved; the flow dynamic optimization engine dynamically adjusts the task priority and resource allocation by using an improved ant colony algorithm based on the real-time task progress and the executor behavior data; meanwhile, the behavior evaluation and intervention module quantitatively evaluates the behavior of the executor according to a weighted scoring model, triggers an automatic intervention process when necessary, and can quickly adapt to sudden changes, so that the resource allocation efficiency is improved, and human errors are reduced.

Description

Technical Field

[0001] The present invention relates to the technical field of architectural design, and particularly to an Internet-based dynamic process management system for architectural design. Background Art

[0002] Dynamic process management in architectural design refers to a systematic method that dynamically optimizes the task allocation, progress tracking, resource scheduling, and collaboration process of architectural design projects through real-time data collection, intelligent algorithm analysis, and automated decision-making means. Different from traditional static process management that relies on fixed rules and manual intervention, its core lies in dynamically adjusting task priorities, resource allocation strategies, and collaboration models according to real-time feedback during project execution (such as task completion, executor behavior characteristics, external environment changes, etc.) to improve design efficiency, reduce human errors, and adapt to complex and changing project requirements.

[0003] However, there are still significant limitations in the prior art in achieving this goal: on the one hand, the task optimization of traditional process management systems relies on preset rules and lacks the ability to deeply mine real-time behavior data and make intelligent responses, resulting in lagging resource allocation and difficulty in coping with sudden changes; on the other hand, architectural design data involves a large number of high-precision BIM models and sensitive information, and existing storage solutions have security risks in data encryption, distributed collaboration, and operation traceability, which are prone to information leakage or version conflicts, seriously restricting the reliability of cross-team collaboration. Therefore, an Internet-based dynamic process management system for architectural design is proposed. Summary of the Invention

[0004] (I) Technical Problems to be Solved

[0005] Aiming at the deficiencies of the prior art, the present invention provides an Internet-based dynamic process management system for architectural design, which has the advantages of intelligent adaptive task scheduling, multi-layer data security protection, behavior-driven automated intervention, and blockchain audit tracking, and solves the problems that there are still significant limitations in the prior art in achieving this goal: on the one hand, the task optimization of traditional process management systems relies on preset rules and lacks the ability to deeply mine real-time behavior data and make intelligent responses, resulting in lagging resource allocation and difficulty in coping with sudden changes; on the other hand, architectural design data involves a large number of high-precision BIM models and sensitive information, and existing storage solutions have security risks in data encryption, distributed collaboration, and operation traceability, which are prone to information leakage or version conflicts, seriously restricting the reliability of cross-team collaboration.

[0006] (II) Technical Solutions

[0007] To achieve the above-mentioned purposes of intelligent adaptive task scheduling, multi-layer data security protection, behavior-driven automated intervention, and blockchain audit tracking, the present invention provides the following technical solutions: An Internet-based dynamic process management system for architectural design, including a multi-factor authentication module, a distributed data storage module, a process dynamic optimization engine, a behavior evaluation and intervention module, and an intelligent audit tracking system;

[0008] The multi-factor authentication module is used to verify the user's identity through a combination of biometric recognition, dynamic tokens, and hardware keys, where the biometric features include double cross-verification of fingerprint, iris, and voiceprint;

[0009] The distributed data storage module stores architectural design drawings and documents in blockchain nodes by hash sharding, and realizes data encryption and computability through homomorphic encryption technology;

[0010] The process dynamic optimization engine dynamically adjusts task priorities and resource allocations based on real-time task progress and executor behavior data through an improved ant colony algorithm;

[0011] The behavior evaluation and intervention module quantitatively evaluates the executor's behavior through a preset weighted scoring model (weight coefficients: task completion rate 40%, collaboration efficiency 30%, compliance 30%), and triggers an automated intervention process when the score is lower than the threshold;

[0012] The intelligent audit tracking system records all operation logs and generates an immutable blockchain deposit, supporting version backtracking and conflict arbitration based on timestamps.

[0013] Preferably, the multi-factor authentication module includes:

[0014] A biometric fusion verification sub-module: performs feature fusion on the feature maps of fingerprint, iris, and voiceprint through a convolutional neural network (CNN), and outputs a verification result with a confidence level ≥ 99.9%;

[0015] A dynamic token generator: generates a 6-digit digital verification code based on time synchronization and the one-time password (OTP) algorithm, with a validity period of 60 seconds;

[0016] A hardware key authentication sub-module: performs hardware binding authentication through the FIDO2 protocol of the USB security key, and the key is stored in the secure element (SE).

[0017] Preferably, the distributed data storage module includes:

[0018] A hash sharding algorithm: divides the file content into N segments, each segment generates a hash value through the SHA-3 algorithm, and is stored in at least three geographically dispersed blockchain nodes;

[0019] Homomorphic Encryption Engine: Adopts the BGV homomorphic encryption scheme based on lattice cryptography, allowing addition and multiplication operations to be performed on encrypted data, enabling version comparison and difference analysis without decryption.

[0020] Preferably, the process dynamic optimization engine includes:

[0021] Task Progress Prediction Model: Performs time series prediction on historical task data through an LSTM neural network, and outputs the task completion probability distribution;

[0022] Resource Allocation Sub-module: Based on the improved ant colony algorithm (ACO), dynamically adjusts the task allocation path through the pheromone update rule and heuristic factors (task urgency, executor load).

[0023] Preferably, the behavior evaluation and intervention module includes:

[0024] Behavior Data Collection Unit: Real-time captures the operation logs, collaboration records, and design document submission frequencies of the executor;

[0025] Weighted Scoring Model: Through the formula:

[0026]

[0027] Quantitatively evaluates the executor's behavior;

[0028] Automated Intervention Trigger: When the score is lower than 70, automatically pushes optimization suggestions to the executor or triggers the manual review process.

[0029] Preferably, the intelligent audit tracking system includes:

[0030] Blockchain Evidence Preservation Sub-module: Adopts a consortium chain architecture, and each block contains: timestamp, operation type, operator identity hash, operation data hash, and previous block hash;

[0031] Version Backtracking Engine: Restores the file version and modification records at any point in time by comparing the hashes of different blocks.

[0032] Preferably, it further includes:

[0033] Permission Hierarchical Management Module: Assigns differentiated operation permissions according to user roles (design engineer, project manager, auditor), and permission changes require two-factor authentication confirmation by at least two administrators;

[0034] Abnormal Behavior Warning Module: Triggers real-time alarms for abnormal behaviors such as frequent modification of core drawings during non-working hours by real-time monitoring of operation frequencies, data access patterns, and IP address changes.

[0035] Preferably, the system supports cross-platform API interfaces, including:

[0036] BIM model docking interface: Seamlessly integrated with building information modeling (BIM) software through the IFC standard protocol;

[0037] Third-party collaboration tool integration interface: Supports real-time data synchronization with collaboration platforms such as Slack and Microsoft Teams.

[0038] (III) Beneficial effects

[0039] Compared with the prior art, the present invention provides an Internet-based dynamic process management system for building design, having the following beneficial effects:

[0040] 1. For the Internet-based dynamic process management system for building design, through the collaborative work of the process dynamic optimization engine and the behavior evaluation and intervention module, the problem that the task optimization of the traditional process management system depends on preset rules is solved; the process dynamic optimization engine, based on real-time task progress and executor behavior data, uses an improved ant colony algorithm to dynamically adjust task priorities and resource allocations; meanwhile, the behavior evaluation and intervention module quantitatively evaluates the executor's behavior according to a weighted scoring model and triggers an automated intervention process when necessary; this mechanism ensures the ability to deeply mine real-time behavior data and make intelligent responses, can quickly adapt to sudden changes, thereby improving resource allocation efficiency and reducing human errors.

[0041] 2. For the Internet-based dynamic process management system for building design, the distributed data storage module uses the hash sharding algorithm and homomorphic encryption technology to shard and store building design drawings and documents on blockchain nodes according to hash values, and realizes data encryption and computability; the intelligent audit tracking system records all operation logs and generates an immutable blockchain certificate, supporting version backtracking and conflict arbitration; this not only enhances the security of data encryption and distributed collaboration, but also provides a reliable operation traceability mechanism, effectively preventing information leakage or version conflicts, and greatly improving the reliability of cross-team collaboration. Description of the drawings

[0042] Figure 1 It is a schematic diagram of the system structure of the present invention. Detailed implementation manners

[0043] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the embodiments and drawings of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all of the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without making creative efforts shall fall within the protection scope of the present invention.

[0044] Please refer to Figure 1, An Internet-based dynamic process management system for architectural design, including a multi-factor authentication module, a distributed data storage module, a process dynamic optimization engine, a behavior evaluation and intervention module, and an intelligent audit tracking system;

[0045] The multi-factor authentication module is used to verify the user identity through the combination of biometric recognition, dynamic token, and hardware key, where the biometric includes double cross-verification of fingerprint, iris, and voiceprint;

[0046] The distributed data storage module stores architectural design drawings and documents in blockchain nodes by hash sharding, and realizes data encryption and computability through homomorphic encryption technology;

[0047] The process dynamic optimization engine dynamically adjusts task priorities and resource allocations based on real-time task progress and executor behavior data through an improved ant colony algorithm;

[0048] The behavior evaluation and intervention module quantitatively evaluates the executor's behavior through a preset weighted scoring model (weight coefficients: task completion rate 40%, collaboration efficiency 30%, compliance 30%), and triggers an automated intervention process when the score is lower than the threshold;

[0049] The intelligent audit tracking system records all operation logs and generates an immutable blockchain evidence deposit, supporting version backtracking and conflict arbitration based on timestamps.

[0050] Example 1:

[0051] This example details the technical implementation of the multi-factor authentication module, including three core sub-modules: biometric fusion verification, dynamic token generation, and hardware key authentication.

[0052] (1). Construction and algorithm design:

[0053] The biometric fusion verification sub-module constructs a multi-modal feature extraction model using a deep learning framework (such as TensorFlow 2.10); the fingerprint recognition model is based on the MobileNetV2 architecture, with an input image size of 256×256 pixels. After extracting texture features through the convolutional layer, it outputs a 128-dimensional feature vector; the iris recognition model uses ResNet-50, with an input image size of 128×128 pixels, and outputs a 256-dimensional feature vector; the voiceprint recognition module uses Mel Frequency Cepstral Coefficients (MFCC) to extract audio features. The input audio sampling rate is 16kHz and the duration is 3 seconds, and it outputs a 40-dimensional feature vector; feature fusion adopts a weighted average strategy, and the weights of fingerprint, iris, and voiceprint are 0.4, 0.3, and 0.3 respectively, and finally generates a 256-dimensional fused feature vector; the dynamic token generator is based on the TOTP algorithm, with a time step of 30 seconds and a key length of 20 bytes (Base32 encoding), and the entropy source is provided by / dev / urandom of Linux; the hardware key authentication sub-module uses a YubiKey 5Nano device, and its security unit (ATECC608A chip) stores the ECC-P256 key pair and supports the WebAuthn interface of the FIDO2 protocol.

[0054] (2). Deployment and integration:

[0055] The biometric model is quantized to INT8 through TensorRT 8.4 and deployed in a Docker container. The inference service interacts with the front end through the gRPC protocol, and the response latency ≤200ms; the dynamic token service is deployed in a Kubernetes cluster (based on AWS EKS), provides an API using the HTTPS interface, and supports JWT token verification; the hardware key authentication module integrates the WebAuthn protocol stack (such as the Authenticator.js library), and the server uses Node.js to implement the FIDO2 authentication endpoint, supporting USB and NFC interfaces; all sub-modules are docked with the identity management system (such as Keycloak) through the OAuth 2.0 protocol to ensure a unified authentication process.

[0056] (3). Operation and verification:

[0057] The user needs to complete the input of biometric features of fingerprint (collected by an optical sensor), iris (collected by a structured light camera), and voiceprint (collected by a directional microphone) in sequence; the system calculates the fused feature vector in real time and performs a cosine similarity comparison with the pre-registered feature library (stored in an encrypted database), and determines to pass when the similarity is ≥0.95; the dynamic token generator polls to generate a new token every 30 seconds, and the server verifies the token validity period (≤30 seconds); when authenticating with a hardware key, after the user inserts the YubiKey, the SE chip generates an ECDSA signature, and the server completes the authentication by verifying the consistency of the signature and the pre-stored public key; after multi-factor authentication passes, the system generates a JWT token and assigns corresponding role permissions.

[0058] Embodiment 2:

[0059] This embodiment describes the implementation details of the hash sharding algorithm of the distributed data storage module and the homomorphic encryption engine.

[0060] (1). Design of the hash sharding algorithm:

[0061] The file sharding adopts a fixed-size strategy, with each slice being 16MB (the last slice is filled with zero bytes); the SHA3-256 algorithm generates a hash value for each slice, and the hash value format is file name_slice number_hash value; the storage node selection is based on the consistent hashing algorithm, and the prefix of the hash value is hashed and mapped with the node ID to ensure that each slice is stored in at least 3 geographically dispersed nodes (such as AWS us-east-1, eu-west-1, ap-southeast-1); the sharding storage backend uses IPFS 0.17.0, and each blockchain node (based on Hyperledger Fabric 2.4) maintains a local IPFS instance and communicates with the sharding manager through the gRPC interface.

[0062] (2). Implementation of the homomorphic encryption engine:

[0063] The homomorphic encryption adopts the BGV scheme (based on the NTRU cryptography library), and the parameter configuration is ring order n = 8192, modulus chain length L = 128, and standard deviation σ = 3.2; when generating the encryption key, the server creates a public-private key pair through a secure random number generator (such as OpenSSL), and the private key is stored in the HSM (Hardware Security Module); in the encryption process, the sharded data is first symmetrically encrypted by AES-256-GCM, and the key is then encrypted by the BGV public key, and finally the ciphertext is stored; when decrypting, the server uses the private key in the HSM to recover the symmetric key and then decrypts the data; the homomorphic calculation supports addition and multiplication operations. The addition operation of the encrypted data is achieved by adding the ciphertexts of the BGV scheme, and the multiplication operation needs to reduce the noise growth through the relinearization of the ciphertext and the key.

[0064] (3). Deployment and Data Synchronization:

[0065] The shard manager is deployed as a Go language microservice, and service discovery is achieved through Consul; blockchain nodes are deployed on AWS EC2 c5.2xlarge instances (8-core CPU, 16GB memory), and each node runs the Peer service of Hyperledger Fabric and the IPFS daemon; asynchronous queue (Kafka 3.3.1) is used for data synchronization. When writing shards, it is first written to the local IPFS, and then other nodes are notified to synchronize through Kafka; when reading, the client requests the shard hash, and the system pulls data from the two nearest nodes and verifies the hash consistency. If they are inconsistent, the arbitration mechanism is triggered to select the majority version.

[0066] Example 3:

[0067] This example details the implementation of a dynamic optimization engine based on LSTM and an improved ant colony algorithm.

[0068] (1). Task Progress Prediction Model:

[0069] The input of the LSTM model is the historical task completion time series (sampled hourly, sequence length 120), the hidden layer dimension is 128, and the output is the normal distribution parameters μ and σ of the task completion probability; the model uses the Adam optimizer (learning rate 0.001, β1 = 0.9, β2 = 0.999), and is trained for 100 rounds on 100,000 historical data (TensorFlow Dataset), and the validation set accuracy ≥ 92%; during inference, the model receives the current task progress (in the range of 0 - 1), and outputs the completion probability distribution for the next 24 hours. If μ ≥ 0.8 and σ ≤ 0.1, then the resource allocation adjustment is triggered.

[0070] (2). Improved Ant Colony Algorithm (ACO):

[0071] The state transition rule of the ACO algorithm uses probability selection:

[0072]

[0073] Among them, τ is the pheromone (initial value 0.1, evaporation coefficient ρ = 0.1), η is the heuristic factor (η = 1 / (task urgency + load factor), load factor = current load / maximum load); the pheromone update rule is:

[0074] τ ij = (1 - ρ)τ ij + Δτ ij

[0075] Δτ is determined by the total revenue of the path completed by the ants, and the revenue formula is:

[0076]

[0077] The algorithm introduces a taboo list (recording the assigned tasks) and an elite strategy (retaining the pheromones of the optimal path); the iteration period is 15 minutes, 100 ants are generated in each iteration, and finally the optimal task assignment path is output.

[0078] (3). Deployment and scheduling:

[0079] The engine is deployed as a Python microservice, and Celery 5.2 framework is used to implement distributed task scheduling; the LSTM model is deployed on the NVIDIA Tesla V100 GPU through TensorRT, and the inference time ≤ 500ms; the ACO algorithm is executed in parallel on the CPU cluster (AWS EC2 c5.4xlarge), and the task status is passed through the Redis queue; the optimization result notifies the task scheduler through Kafka to dynamically adjust the resource pool (such as increasing GPU resources or allocating backup engineers).

[0080] Example 4:

[0081] This example describes the implementation details of the behavior evaluation model and the automated intervention process.

[0082] (1). Behavior data collection and cleaning:

[0083] Data collection is implemented through the ELK stack (Elasticsearch 8.5, Logstash 8.5, Kibana 8.5); operation logs are collected through Nginx access logs and application logs (such as MDC of Spring Boot), and the fields include operation type (such as "modifying drawings"), timestamp, operator ID, and operation object ID; collaboration records capture team chat messages (such as Slack message content) and file sharing events through WebSocket; data cleaning uses Apache Kafka stream processing to filter invalid logs (such as repeated clicks, non-business operations), and standardize the field format.

[0084] (2). Weighted scoring model:

[0085] The scoring formula is:

[0086] Score = 0.4 · Task completion rate + 0.3 · Collaboration efficiency + 0.3 · Compliance

[0087] Among them, the calculation formula for the task completion rate is:

[0088]

[0089] The collaboration efficiency is calculated as:

[0090]

[0091] The compliance is calculated as:

[0092]

[0093] The model is deployed through the Flask API, pulls the latest data from Elasticsearch every hour, calculates the user score, and stores it in the Redis cache.

[0094] (3). Automated intervention trigger mechanism:

[0095] When the user Score < 70, the system executes the following intervention process:

[0096] Low-risk intervention: Push optimization suggestions to the user workbench (such as "Article 3.2 of the design specification does not meet the standards"), and record the type of suggestion.

[0097] Medium-risk intervention: Trigger the manual review process and notify the project manager to approve the intervention request through Slack.

[0098] High-risk intervention: If the Score remains below 60 for 3 consecutive hours, the system automatically freezes some user permissions (such as prohibiting the modification of core drawings), and sends an alarm to the security center. The intervention record is stored in the blockchain evidence storage module, supporting post-event auditing.

[0099] Example 5:

[0100] This example describes the implementation technology of the blockchain evidence storage and version traceability engine.

[0101] (1). Blockchain evidence storage sub-module:

[0102] Adopt the consortium chain architecture (Hyperledger Fabric 2.4), and the block structure includes:

[0103] Block header: Timestamp (millisecond level), previous block hash, Merkle root hash (generated from all transaction hashes in the block body).

[0104] Block body: Each transaction record includes the operation type (such as "modifying the drawing"), the operator identity hash (SHA-256), the operation data hash (SHA-256), and the target resource ID.

[0105] The consensus mechanism adopts an improved PBFT algorithm with the number of nodes N = 5 and the consensus passing threshold of 3 / 4 (at least 4 nodes need to confirm); the nodes are deployed on Alibaba Cloud ECS instances (c5.2xlarge), and cross-node data synchronization is achieved through IPFS; transaction broadcasting is implemented through gRPC, and the block generation interval is 10 seconds.

[0106] Version backtracking engine:

[0107] The version tree is implemented in a Git-like structure, and each file version corresponds to a block hash. The version difference analysis uses the libgit2 library. By comparing the Merkle tree paths of two versions, the added / modified / deleted file blocks are calculated; the backtracking process is as follows:

[0108] Step 1. The user specifies the target timestamp, and the system traverses the blockchain to find the nearest block hash;

[0109] Step 2. Retrieve the corresponding shard data according to the hash and obtain the file content through IPFS;

[0110] Step 3. Generate a version report, including modification records (such as "the height of the beam on the 12th floor is changed from 300mm to 350mm") and the identity information of the operator.

[0111] The engine is deployed as a Go language microservice and provides the following interfaces through REST API:

[0112] / v1 / audit / history?timestamp=1678981234: Obtain the version snapshot at the specified time point;

[0113] / v1 / audit / diff?from=hash18to=hash2: Compare the differences between two versions.

[0114] Example VI:

[0115] This example details the implementation of the permission control system and the anomaly detection model.

[0116] (1). Hierarchical permission management:

[0117] The system adopts the RBAC model and defines the following roles:

[0118] Design engineer: Can create / edit tasks, but has no permission to delete other people's data or modify permission configurations;

[0119] Project manager: Can adjust task priorities, allocate resources, but cannot access audit logs;

[0120] Auditor: Can view all operation logs, but has no modification permission.

[0121] Permission changes require a two-factor authentication process:

[0122] Step 1. The applicant (must be a "super administrator") submits a change request, including the target user ID, new role, and reason for the change;

[0123] Step 2. At least two other "super administrators" sign a digital signature after passing biometric authentication;

[0124] Step 3. After the signature is verified through a multi-signature contract, the permission change is written to the blockchain for evidence storage.

[0125] The role permission mapping is stored in a MySQL cluster (master-slave architecture), and data access is implemented through Hibernate ORM, with a query latency ≤ 50ms.

[0126] (2). Abnormal behavior warning:

[0127] The anomaly detection model uses the Isolation Forest algorithm (scikit-learn 1.2.2), and the input features include:

[0128] Operation frequency: The number of operations per hour (threshold upper limit 200 times);

[0129] Data access pattern: The frequency of accessing files outside the scope of duties (e.g., a design engineer accessing financial data);

[0130] IP address change: The number of logins from a non-office network (e.g., not 192.168.1.0 / 24);

[0131] The model training data includes 100,000 normal behavior logs and 10,000 abnormal samples, and the threshold is set to an abnormal score > 0.8; it is deployed in a Spark Streaming cluster (consuming log streams through Kafka), calculating features in real-time and triggering alarms; the alarms are visualized through Prometheus + Grafana and sent as text messages to the security team through the Twilio API.

[0132] Example 7:

[0133] This example describes the technical implementation of BIM model docking and integration with third-party collaboration tools.

[0134] (1). BIM model docking interface:

[0135] The interface follows the IFC 4.3 standard and uses an OpenBIM toolkit (such as IfcOpenShell 0.7.0) to parse model elements; the process is as follows:

[0136] Step 1. The user uploads the IFC file, and the system extracts the component properties (such as the material of the wall and the cross-sectional dimensions of the beam) through IfcOpenShell;

[0137] Step 2. The component properties are mapped to the internal JSON format of the system, including geometric data (such as coordinates and dimensions) and non-geometric data (such as material specifications);

[0138] Step 3. The mapped data is stored in the distributed storage module, triggering the process optimization engine to generate associated tasks (such as "review the beam cross-section design").

[0139] The interface is deployed as a RESTful API (based on Spring Boot 3.0), supporting API callbacks of Revit and ArchiCAD, and authorizing access to the BIM model library through OAuth 2.0.

[0140] In summary, this Internet-based dynamic process management system for architectural design solves the problem that the task optimization of traditional process management systems depends on preset rules through the collaborative work of the process dynamic optimization engine and the behavior evaluation and intervention module; the process dynamic optimization engine dynamically adjusts the task priority and resource allocation based on real-time task progress and executor behavior data using an improved ant colony algorithm; at the same time, the behavior evaluation and intervention module quantitatively evaluates the executor's behavior according to the weighted scoring model and triggers an automated intervention process when necessary; this mechanism ensures the deep mining and intelligent response ability to real-time behavior data, can quickly adapt to sudden changes, thereby improving the resource allocation efficiency and reducing human errors.

[0141] Moreover, in this Internet-based dynamic process management system for architectural design, the distributed data storage module uses the hash sharding algorithm and homomorphic encryption technology to shard and store architectural design drawings and documents on blockchain nodes according to the hash value, and realizes data encryption and computability; the intelligent audit tracking system records all operation logs and generates an immutable blockchain certificate, supporting version backtracking and conflict arbitration; this not only enhances the security of data encryption and distributed collaboration, but also provides a reliable operation traceability mechanism, effectively preventing information leakage or version conflicts, greatly improving the reliability of cross-team collaboration, and solving the significant limitations that still exist in the prior art when achieving this goal: on the one hand, the task optimization of traditional process management systems depends on preset rules, lacking the deep mining and intelligent response ability to real-time behavior data, resulting in lagging resource allocation and difficulty in coping with sudden changes; on the other hand, architectural design data involves a large number of high-precision BIM models and sensitive information, and existing storage solutions have security risks in data encryption, distributed collaboration, and operation traceability, easily causing information leakage or version conflicts, seriously restricting the reliability of cross-team collaboration.

[0142] All relevant modules involved in this system are hardware system modules or functional modules that combine computer software programs or protocols in the prior art with hardware. The computer software programs or protocols themselves involved in this functional module are all well-known technologies to those skilled in the art and are not the improvements of this system. The improvement of this system lies in the interaction relationship or connection relationship between modules, that is, the overall structure of the system is improved to solve the corresponding technical problems to be solved by this system.

[0143] Although the embodiments of the present invention have been shown and described, those of ordinary skill in the art can understand that various changes, modifications, substitutions, and variations can be made to these embodiments without departing from the principles and spirit of the present invention. The scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. An Internet-based dynamic process management system for architectural design, characterized in that, It includes a multi-factor authentication module, a distributed data storage module, a process dynamic optimization engine, a behavior evaluation and intervention module, and an intelligent audit tracking system; The multi-factor authentication module is used to verify the user identity through the combination of biometric recognition, dynamic token, and hardware key, where the biometric includes double cross-verification of fingerprint, iris, and voiceprint; The distributed data storage module stores building design drawings and documents in blockchain nodes by hash sharding, and realizes data encryption and computability through homomorphic encryption technology; The process dynamic optimization engine dynamically adjusts task priorities and resource allocations based on real-time task progress and executor behavior data through an improved ant colony algorithm; The behavior evaluation and intervention module quantitatively evaluates the executor's behavior through a preset weighted scoring model (weight coefficients: task completion rate 40%, collaboration efficiency 30%, compliance 30%), and triggers an automated intervention process when the score is lower than the threshold; The intelligent audit tracking system records all operation logs and generates an immutable blockchain evidence deposit, supporting version backtracking and conflict arbitration based on timestamps.

2. The dynamic process management system for architectural design based on the Internet according to claim 1, characterized in that, The multi-factor authentication module includes: Biometric fusion verification sub-module: Feature fusion of fingerprint, iris, and voiceprint feature maps is performed through a convolutional neural network (CNN), and a verification result with a confidence level ≥ 99.9% is output; Dynamic token generator: Generates a 6-digit digital verification code based on time synchronization and one-time password (OTP) algorithm, with a validity period of 60 seconds; Hardware key authentication sub-module: Performs hardware binding authentication through the FIDO2 protocol of the USB security key, and the key is stored in the secure element (SE).

3. An Internet-based dynamic process management system for architectural design according to claim 1, characterized in that, The distributed data storage module includes: Hash sharding algorithm: Divides the file content into N segments, each segment generates a hash value through the SHA-3 algorithm, and is stored in at least three geographically dispersed blockchain nodes; Homomorphic encryption engine: Adopts the BGV homomorphic encryption scheme based on lattice cryptography, allowing addition and multiplication operations to be performed on encrypted data, and version comparison and difference analysis can be completed without decryption.

4. An Internet-based dynamic process management system for architectural design according to claim 1, characterized in that The process dynamic optimization engine includes: Task progress prediction model: Performs time series prediction on historical task data through an LSTM neural network, and outputs the task completion probability distribution; Resource allocation sub-module: Based on the improved ant colony algorithm (ACO), dynamically adjusts the task allocation path through the pheromone update rule and heuristic factors (task urgency, executor load).

5. The dynamic process management system for architectural design based on the Internet according to claim 1, wherein The behavior evaluation and intervention module includes: Behavior data collection unit: Real-time captures the executor's operation logs, collaboration records, and design document submission frequencies; Weighted scoring model: Through the formula: Quantitatively evaluates the executor's behavior; Automated intervention trigger: When the score is lower than 70 points, automatically pushes optimization suggestions to the executor or triggers a manual review process.

6. The dynamic process management system for building design based on the Internet according to claim 1, characterized in that, The intelligent audit tracking system includes: Blockchain evidence deposit sub-module: Adopts a consortium chain architecture, and each block contains: timestamp, operation type, operator identity hash, operation data hash, and previous block hash; Version Backtracking Engine: By comparing the hash values of different blocks, restore the file versions and modification records at any point in time.

7. A dynamic process management system for architectural design based on the Internet according to claim 1, characterized in that, It also includes: Permission Hierarchical Management Module: Assign differentiated operation permissions according to user roles (design engineer, project manager, auditor), and permission changes require confirmation by two-factor authentication of at least two administrators; Abnormal Behavior Warning Module: Trigger real-time alarms for abnormal behaviors such as frequent modification of core drawings during non-working hours by monitoring operation frequency, data access patterns, and IP address changes in real time.

8. An Internet-based dynamic process management system for architectural design according to claim 1, characterized in that, The system supports cross-platform API interfaces, including: BIM Model Docking Interface: Seamlessly integrate with building information model (BIM) software through the IFC standard protocol; Third-Party Collaboration Tool Integration Interface: Support real-time data synchronization with Slack and Microsoft Teams collaboration platforms.

Images