Statistical regional visitor data query method

Abstract

The invention discloses a statistical regional visitor data query method, which relates to the technical field of information processing, and collects, fuses and processes visitor biological characteristics and environment interaction data by deploying a multi-mode biological characteristic acquisition device and an Internet of Things sensor. A deep neural network model is trained in a distributed mode through the federated learning and quantum encryption technology, and data security is ensured on the premise that privacy is protected; quantum computing is introduced to accelerate query matching, a detailed analysis report is generated in combination with deep learning semantic comprehension, a model is updated online according to user feedback, and a cloud server optimizes global parameters regularly. Efficient, safe and accurate visitor data query and analysis are realized, and powerful support is provided for regional management.

Description

Technical Field

[0001] The present invention relates to the field of information processing technology, in particular to a method for querying visitor data in a statistical area. Background Art

[0002] With the continuous advancement of information technology, regional visitor profile query technology has also continued to develop. Initially, it relied primarily on manual registration and simple database queries. Later, biometric recognition technologies such as fingerprint and facial recognition, as well as IoT sensors used to collect visitor behavior data, were gradually introduced, improving data processing and query efficiency. In recent years, artificial intelligence technologies such as deep learning and federated learning have begun to be applied in this field. By building complex neural network models, they can more accurately identify visitor characteristics and behavior patterns. However, existing technologies still have many shortcomings in terms of data security, privacy protection, query efficiency, and accuracy, making it difficult to meet the growing demand for complex regional visitor profile queries.

[0003] Existing technologies often face data security risks. Model parameters can be easily intercepted and cracked during transmission, leading to visitor privacy leaks. Furthermore, traditional query methods struggle to efficiently process large-scale, multi-source, heterogeneous data, resulting in slow query speeds and an inability to accurately mine the deep connections underlying visitor behavior. Furthermore, the models suffer from poor adaptability, making it difficult to quickly adapt to changes in visitor behavior patterns. This inability to continuously optimize query performance prevents efficient and secure visitor data querying and in-depth analysis. Summary of the Invention

[0004] In view of the above existing problems, the present invention is proposed.

[0005] Therefore, the present invention provides a method for querying visitor information in a statistical area to solve the problems of high data security risks, insufficient privacy protection, low query efficiency and poor accuracy in the existing method for querying visitor information in a statistical area, as well as how to achieve efficient and secure visitor information query and in-depth analysis.

[0006] In order to solve the above technical problems, the present invention provides the following technical solutions:

[0007] In a first aspect, the present invention provides a method for querying visitor data in a statistical area, characterized by comprising the following steps:

[0008] S1: Deploy multimodal biometric acquisition equipment to collect visitor biometric data. Simultaneously, use IoT sensors to obtain visitor environmental interaction data. All data is time-stamped and transmitted to edge computing nodes. The edge computing nodes clean and standardize the data, and generate a unique biometric identifier for each visitor. The processed multi-source heterogeneous biometric data is then correlated and integrated with the environmental interaction data to form a fused dataset.

[0009] S2: Use the fused dataset to initialize the deep neural network model and adopt a federated learning mechanism. Only the encrypted model parameter gradients are sent to the cloud coordination server through a quantum encryption algorithm. The cloud server decrypts and aggregates the data, updates the global model parameters, and sends them back to each edge node for quantum encryption to update the local model. After multiple rounds of iterative training, the local federated learning model is obtained.

[0010] S3: After the user submits a query request, the query conditions are quantized and encoded into a qubit query vector. The parallel processing capabilities of quantum computing are used to perform parallel matching calculations with the quantized visitor data stored in the local federated learning model. Quantum gate circuits and entanglement properties are used to quickly locate highly relevant visitor data, and the matching results are output to the analysis step.

[0011] S4: Combined with a deep learning model, the matching results are semantically understood and associated with mining to generate an analysis report containing detailed information such as the visitor's biometrics, behavioral trajectory, and consumption habits. This report is output to the query terminal for display to the user, and user feedback data is fed back to the model update step.

[0012] S5: Use user feedback data to update the local federated learning model online at the edge computing node to optimize model query performance. The feedback data is quantum encrypted and sent to the cloud coordination server to update the global model parameters.

[0013] S6: The cloud coordination server regularly integrates feedback from each edge node to optimize global model parameters, continuously improve the model's generalization ability and adaptability, and ensure the efficient and stable operation of the entire visitor information query system.

[0014] As a preferred solution of the method for querying visitor information in a statistical area described in the present invention, the multimodal biometric feature acquisition device adopts adaptive sampling technology to dynamically adjust the sampling frequency according to visitor traffic to ensure the real-time and accuracy of data acquisition; the edge computing node uses an anomaly detection algorithm based on deep learning during the data cleaning process to automatically identify and correct abnormal data.

[0015] As a preferred solution to the method for querying visitor information in a statistical area described in the present invention, the deep neural network model adopts a multi-task learning architecture to simultaneously learn visitor biometric classification and behavior pattern prediction tasks, sharing the underlying feature extraction layer to improve the generalization ability of the model; the quantum encryption algorithm combines quantum key distribution and quantum state encryption to ensure the transmission security of model parameter gradients.

[0016] As a preferred solution of the method for querying visitor information of a statistical area described in the present invention, the quantum bit query vector is generated by a quantum entangled pair, and each query condition corresponds to an entangled quantum bit pair, thereby achieving efficient and accurate parallel matching; the quantum computing uses a quantum random walk algorithm to accelerate the feature comparison process.

[0017] As a preferred solution of the method for querying visitor information in a statistical area described in the present invention, the deep learning model is combined with a generative adversarial network (GAN) to generate virtual samples of visitor behavior patterns, which are used to expand the training data set and improve the robustness of the model; the semantic understanding and association mining adopt a graph neural network (GNN) to construct a visitor behavior knowledge graph.

[0018] As a preferred solution of the method for querying visitor information in a statistical area described in the present invention, the user feedback data is processed using differential privacy technology to protect user privacy while providing effective information for model updates; the online update uses an online transfer learning algorithm to quickly adapt to new visitor behavior patterns.

[0019] As a preferred solution of the method for querying visitor information in a statistical area described in the present invention, the cloud-based coordination server adopts a distributed collaborative optimization algorithm, integrates the model parameter update information fed back by each edge node, and dynamically adjusts the global model parameters; the model optimization process introduces a reinforcement learning mechanism, uses model query accuracy and response speed as reward signals, and continuously optimizes model performance.

[0020] In a second aspect, the present invention provides a method system for querying visitor data in a statistical area, comprising:

[0021] Data collection module: Deploy multimodal biometric collection equipment to collect visitor biometric data, and use IoT sensors to obtain visitors' environmental interaction data. All data are time-stamped and synchronized before being transmitted to the edge computing node.

[0022] Data fusion module: Cleans and standardizes data at the edge computing node, generates a unique biometric identifier for each visitor, and associates and integrates the processed multi-source heterogeneous biometric data with environmental interaction data to form a fused data set.

[0023] Model training module: Use the fused data set to initialize the deep neural network model, adopt the federated learning mechanism, and only send the encrypted model parameter gradients to the cloud coordination server through the quantum encryption algorithm. The cloud server decrypts and aggregates the data, updates the global model parameters, and quantum encrypts them again and sends them back to each edge node to update the local model. After multiple rounds of iterative training, the local federated learning model is obtained.

[0024] Query matching module: After the user submits a query request, the query conditions are quantized and encoded into a quantum bit query vector. The parallel processing capabilities of quantum computing are used to perform parallel matching calculations with the visitor data stored in the quantized form in the local federated learning model. The highly relevant visitor data is quickly located through quantum gate circuits and entanglement characteristics, and the matching results are output to the analysis step.

[0025] Report generation module: Combines deep learning models to perform semantic understanding and association mining on matching results, generates analysis reports containing detailed information such as visitor biometrics, behavioral trajectories, consumption habits, etc., outputs them to the query terminal for display to users, and transmits user feedback data back to the model update step.

[0026] Feedback update module: Use user feedback data to update the local federated learning model online at the edge computing node, optimize model query performance, and send the feedback data to the cloud coordination server after quantum encryption to update the global model parameters.

[0027] Model optimization module: The cloud coordination server regularly integrates feedback from each edge node to optimize global model parameters, continuously improve the model's generalization ability and adaptability, and ensure the efficient and stable operation of the entire visitor information query system.

[0028] In a third aspect, the present invention provides a computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: when the computer program is executed by the processor, any step of the method for querying visitor information of a statistical area as described in the first aspect of the present invention is implemented.

[0029] In a fourth aspect, the present invention provides a computer-readable storage medium having a computer program stored thereon, wherein: when the computer program is executed by a processor, any step of the method for querying visitor information of a statistical area as described in the first aspect of the present invention is implemented.

[0030] The beneficial effects of the present invention are as follows: through the innovative combination of advanced technologies such as multimodal biometric collection, federated learning, quantum encryption, quantum computing and deep learning, the security, accuracy and efficiency of regional visitor information queries are effectively improved. First, the fusion collection of multimodal biometric data and the pre-processing of edge computing nodes ensure the diversity and quality of the data, providing a basis for subsequent accurate queries. Secondly, the application of the federated learning mechanism and quantum encryption technology not only protects data privacy, but also realizes efficient distributed training of the model, avoiding the risk of data leakage. Thirdly, the introduction of quantum computing greatly accelerates the query matching process and improves query efficiency. Finally, the semantic understanding and association mining capabilities of the deep learning model can generate detailed visitor information analysis reports, while using user feedback data to continuously optimize the model, enhancing the adaptability and practicality of the system. Overall, the present invention achieves efficient query and in-depth analysis of regional visitor information under the premise of ensuring data security and privacy protection, significantly improves user experience and system performance, and provides strong support for regional management and service optimization. BRIEF DESCRIPTION OF THE DRAWINGS

[0031] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. 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 any creative work.

[0032] Figure 1 This is a flow chart of the method for querying visitor information in a statistical area in Example 1. DETAILED DESCRIPTION

[0033] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the specific embodiments of the present invention are described in detail below with reference to the accompanying drawings.

[0034] In the following description, many specific details are set forth to facilitate a full understanding of the present invention. However, the present invention may also be implemented in other ways different from those described herein. Those skilled in the art may make similar generalizations without violating the connotation of the present invention. Therefore, the present invention is not limited to the specific embodiments disclosed below.

[0035] Secondly, the term "one embodiment" or "embodiment" herein refers to a specific feature, structure, or characteristic that may be included in at least one implementation of the present invention. The phrase "in one embodiment" appearing in various places throughout this specification does not necessarily refer to the same embodiment, nor does it refer to a separate or selective embodiment that is mutually exclusive of other embodiments.

[0036] Example 1, reference Figure 1, which is the first embodiment of the present invention, provides a method for querying visitor data in a statistical area, characterized by comprising the following steps:

[0037] S1: Deploy multimodal biometric acquisition equipment to collect visitor biometric data. Simultaneously, use IoT sensors to obtain visitor environmental interaction data. All data is time-stamped and transmitted to edge computing nodes. The edge computing nodes clean and standardize the data, and generate a unique biometric identifier for each visitor. The processed multi-source heterogeneous biometric data is then correlated and integrated with the environmental interaction data to form a fused dataset.

[0038] S2: Use the fused dataset to initialize the deep neural network model and adopt a federated learning mechanism. Only the encrypted model parameter gradients are sent to the cloud coordination server through a quantum encryption algorithm. The cloud server decrypts and aggregates the data, updates the global model parameters, and sends them back to each edge node for quantum encryption to update the local model. After multiple rounds of iterative training, the local federated learning model is obtained.

[0039] S3: After the user submits a query request, the query conditions are quantized and encoded into a qubit query vector. The parallel processing capabilities of quantum computing are used to perform parallel matching calculations with the quantized visitor data stored in the local federated learning model. Quantum gate circuits and entanglement properties are used to quickly locate highly relevant visitor data, and the matching results are output to the analysis step.

[0040] S4: Combined with a deep learning model, the matching results are semantically understood and associated with mining to generate an analysis report containing detailed information such as the visitor's biometrics, behavioral trajectory, and consumption habits. This report is output to the query terminal for display to the user, and user feedback data is fed back to the model update step.

[0041] S5: Use user feedback data to update the local federated learning model online at the edge computing node to optimize model query performance. The feedback data is quantum encrypted and sent to the cloud coordination server to update the global model parameters.

[0042] S6: The cloud coordination server regularly integrates feedback from each edge node to optimize global model parameters, continuously improve the model's generalization ability and adaptability, and ensure the efficient and stable operation of the entire visitor information query system.

[0043] It should be noted that multimodal biometric collection equipment, including high-precision fingerprint scanners, facial recognition cameras, and iris recognition devices, is first deployed at each entrance, key node, and relevant facility within the area. IoT sensors are also used to collect visitor interaction data, such as user activity on smart guide devices and duration of stay in interactive display areas. This collected data is time-stamped and transmitted to edge computing nodes. At the edge computing nodes, the data is cleaned to remove noise and erroneous information, such as by denoising fingerprint images and performing face detection and alignment on facial recognition video frames. Data in different formats is then standardized and represented in a unified format, such as converting fingerprint feature vectors, facial feature vectors, and iris feature vectors into fixed-length floating-point vectors. A unique biometric identifier is then created for each visitor, and the processed biometric data is linked to the environmental interaction data to form a multi-source, heterogeneous biometric fusion dataset.

[0044] This step realizes the comprehensive collection and preprocessing of visitors' multimodal biometric data and environmental interaction data, providing a rich and high-quality data foundation for subsequent visitor information query and analysis.

[0045] Through multimodal biometric collection, the accuracy and uniqueness of visitor identification are improved; data cleaning and standardization ensure the quality and consistency of the data, which is conducive to the accuracy of subsequent model training and queries; a unique biometric identifier is generated for each visitor, which facilitates the management and association of visitor data and improves the efficiency and accuracy of data processing.

[0046] S2: Model training based on federated learning and quantum encryption

[0047] Using the fused dataset generated in step S1, deep neural network models are initialized at edge computing nodes in multiple regions. Using a federated learning mechanism, each edge computing node sends only the encrypted model parameter gradients to the cloud coordination server via a quantum encryption algorithm. Specifically, the model parameter gradients are encrypted using the unclonability of quantum states and the absolute security of quantum key distribution. After receiving the encrypted gradients from each edge computing node, the cloud coordination server decrypts and aggregates them, updates the global model parameters, and sends the updated global model parameters back to each edge computing node via quantum encryption for local model updates. After multiple rounds of iterative training, the local model gradually learns the association between visitor behavior patterns and biometric characteristics, resulting in a trained local federated learning model.

[0048] This step aims to build a deep neural network model that can learn the association between visitor behavior patterns and biometric features while protecting data privacy and security.

[0049] The application of the federated learning mechanism enables model training to be carried out without sharing the original data, thus protecting the privacy of visitor data. The introduction of the quantum encryption algorithm further enhances the security of data transmission, preventing the model parameters from being stolen and tampered with during transmission. Through multiple rounds of iterative training, the local model can learn more accurate visitor behavior patterns and biometric associations, thereby improving the model's performance and generalization capabilities.

[0050] S3: Efficient Query Matching Assisted by Quantum Computing

[0051] When a user submits a visitor profile query request through a query terminal, the query terms are first quantized and converted into a query vector in the form of quantum bits. Then, leveraging the parallel processing capabilities of quantum computing, a large-scale parallel matching calculation is performed between the quantized query vector and the quantized visitor data stored in the local federated learning model. During this calculation, quantum gate circuits are used to achieve efficient feature comparison and similarity calculation, and the characteristics of quantum entanglement are used to accelerate the information retrieval process, thereby quickly locating visitor profiles that are highly relevant to the query terms and outputting matching results.

[0052] This step achieves efficient query and matching of visitor information, and can quickly and accurately find visitor information related to the query conditions.

[0053] The introduction of quantum computing has greatly improved the efficiency of query matching, enabling it to process large amounts of data in a short period of time and quickly respond to user query requests. Through quantum coding and quantum entanglement characteristics, the accuracy and reliability of matching have been improved, enabling more accurate location of relevant visitor information.

[0054] S4: Deep learning-driven semantic understanding and analysis report generation

[0055] The matching results from step S3 are then further analyzed using a deep learning model to further understand semantics and mine associations. By analyzing the visitor's biometrics, behavioral trajectories, and other information contained in the matching results, the deep learning model uncovers underlying patterns and associations within the visitor's profile, such as their behavioral preferences and spending tendencies. Based on these findings, a detailed visitor profile analysis report is generated, including the visitor's basic biometric information, behavioral trajectories, and spending habits. This report is then displayed to the user via the query terminal. Simultaneously, user feedback on the query results is fed back to the model update step.

[0056] This step conducts in-depth analysis and interpretation of the matching results, providing users with a one-stop comprehensive data query and analysis report, while collecting user feedback to optimize the model.

[0057] The semantic understanding and association mining capabilities of deep learning models can extract more valuable information from visitor data, providing users with a more comprehensive and in-depth visitor portrait, helping users better understand visitor behavior and needs; the collection of user feedback data provides a basis for subsequent model optimization, which can further improve model performance and query accuracy.

[0058] S5: Online model update based on user feedback

[0059] The local federated learning model is updated online at the edge computing node using the user feedback data returned in step S4. Specifically, the model parameters are adjusted based on the user feedback data to optimize the model's query performance. Simultaneously, the feedback data is quantum-encrypted and sent to the cloud coordination server for updating the global model parameters.

[0060] This step optimizes the local federated learning model in real time based on user feedback, ensuring that the model can adapt to changes in visitor behavior and continuously improve performance.

[0061] Through the online update mechanism, the model can learn the latest changes in visitor behavior patterns in a timely manner, improving the model's adaptability and query performance; quantum encryption ensures the security of feedback data during transmission, preventing the risk of data leakage and malicious exploitation.

[0062] S6: Model parameter optimization on cloud servers

[0063] The cloud-based coordination server regularly integrates model parameter updates from each edge computing node and uses a distributed collaborative optimization algorithm to optimize and adjust global model parameters. During this optimization process, a reinforcement learning mechanism is introduced, using model query accuracy and response speed as reward signals to continuously improve global model parameters, thereby continuously enhancing the model's generalization and adaptability, ensuring the efficient and stable operation of the entire visitor information query system.

[0064] This step aims to further improve the performance of the global model so that it can better adapt to the query requirements of the entire system.

[0065] The distributed collaborative optimization algorithm can make full use of the feedback information of each edge node to more effectively optimize the global model parameters; the introduction of the reinforcement learning mechanism enables the model to dynamically adjust parameters according to the actual query results, further improving the query accuracy and response speed of the model, and enhancing the overall performance and stability of the system.

[0066] Specifically, the multimodal biometric feature collection device adopts adaptive sampling technology to dynamically adjust the sampling frequency according to visitor traffic to ensure the real-time and accuracy of data collection; the edge computing node uses an anomaly detection algorithm based on deep learning during the data cleaning process to automatically identify and correct abnormal data.

[0067] It should be noted that multimodal biometric data collection equipment, including high-precision fingerprint scanners, facial recognition cameras, and iris recognition devices, is deployed at various entrances, key nodes, and related facilities within the area. IoT sensors are also used to collect visitor interaction data, such as user activity on smart guide devices and duration of stay in interactive display areas. The collected biometric data and sensor data are synchronized with timestamps before being transmitted to edge computing nodes. At the edge computing nodes, the data is first cleaned to remove noise and erroneous information. For example, fingerprint images are denoised, and face detection and alignment are performed on facial recognition video frames. Next, data in different formats is standardized and represented in a unified manner. For example, fingerprint feature vectors, facial feature vectors, and iris feature vectors are converted into fixed-length floating-point vectors. A unique biometric identifier is then created for each visitor, and the processed biometric data is linked to the environmental interaction data to form a multi-source, heterogeneous biometric fusion dataset.

[0068] Step S1 achieves comprehensive collection and preprocessing of visitors' multimodal biometric data and environmental interaction data. Deploying a variety of biometric collection devices and IoT sensors enables the acquisition of rich visitor information, ensuring data comprehensiveness and diversity. Data timestamps are synchronized and transmitted to edge computing nodes, ensuring real-time and accuracy. Data cleaning and standardization at the edge computing nodes removes noise and erroneous information, unifies data formats, and improves data quality and consistency, providing a high-quality data foundation for subsequent model training and query analysis. Generating a unique biometric identifier for each visitor facilitates the unified management and association of visitor data, improves data processing efficiency and accuracy, and ensures the accuracy and reliability of visitor profile queries in subsequent steps.

[0069] Specifically, the deep neural network model adopts a multi-task learning architecture to simultaneously learn visitors' biometric classification and behavior pattern prediction tasks, share the underlying feature extraction layer, and improve the model's generalization ability; the quantum encryption algorithm combines quantum key distribution and quantum state encryption to ensure the secure transmission of model parameter gradients.

[0070] It should be noted that the fused dataset formed in step S1 is used to initialize deep neural network models at edge computing nodes in multiple regions. Using a federated learning mechanism, each edge computing node sends only the encrypted model parameter gradients to the cloud coordination server via a quantum encryption algorithm. Specifically, the model parameter gradients are encrypted using the unclonability of quantum states and the absolute security of quantum key distribution. After receiving the encrypted gradients from each edge computing node, the cloud coordination server decrypts and aggregates them, updates the global model parameters, and sends the updated global model parameters back to each edge computing node via quantum encryption for updating the local model. After multiple rounds of iterative training, the local model gradually learns the association between visitor behavior patterns and biometric characteristics, resulting in a trained local federated learning model.

[0071] Step S2 enables model training while protecting data privacy. The application of a federated learning mechanism enables model training without sharing original data. Each edge computing node transmits only encrypted model parameter gradients, effectively protecting the privacy of visitor data. The introduction of quantum encryption algorithms further enhances data transmission security. Leveraging the unclonability of quantum states and the absolute security of quantum key distribution, model parameters are prevented from being stolen or tampered with during transmission. Multiple rounds of iterative training enable the local model to learn the complex correlations between visitor behavior patterns and biometrics, improving model performance and generalization capabilities, and providing efficient and accurate model support for subsequent visitor profile queries.

[0072] Specifically, the quantum bit query vector is generated through quantum entangled pairs, and each query condition corresponds to an entangled quantum bit pair, achieving efficient and accurate parallel matching; the quantum computing uses a quantum random walk algorithm to accelerate the feature comparison process.

[0073] It should be noted that when a user submits a visitor profile query request through a query terminal, the query terms are first quantized and converted into a query vector in the form of quantum bits. Then, leveraging the parallel processing capabilities of quantum computing, a large-scale parallel matching calculation is performed between the quantized query vector and the quantized visitor data stored in the local federated learning model. During this calculation, quantum gate circuits are used to achieve efficient feature comparison and similarity calculations, and the characteristics of quantum entanglement are used to accelerate the information retrieval process, thereby quickly locating visitor profiles highly relevant to the query terms and outputting matching results.

[0074] Through step S3, efficient querying and matching of visitor profiles is achieved. The introduction of quantum computing enables the processing of large amounts of data in a short period of time, significantly improving query matching efficiency and enabling rapid response to user query requests. The application of quantum coding and quantum entanglement makes feature matching and similarity calculation more efficient and accurate, enabling more precise location of relevant visitor profiles, improving query accuracy and reliability, and providing powerful computing support for the user-friendly engine.

[0075] Specifically, the deep learning model is combined with a generative adversarial network (GAN) to generate virtual samples of visitor behavior patterns, which are used to expand the training data set and improve the robustness of the model; the semantic understanding and association mining adopts a graph neural network (GNN) to construct a visitor behavior knowledge graph.

[0076] It should be noted that the matching results output in step S3 are further semantically understood and associations mined using a deep learning model. By analyzing visitor biometrics, behavioral trajectories, and other information contained in the matching results, the deep learning model uncovers underlying patterns and associations within the visitor profile, such as their behavioral preferences and spending tendencies. Based on these mining results, a detailed visitor profile analysis report is generated, including the visitor's basic biometric information, behavioral trajectories, and spending habits. This report is then displayed to the query terminal for the user. Simultaneously, user feedback on the query results is fed back to the model update step.

[0077] Step S4 enables in-depth analysis and interpretation of the matching results. The deep learning model's semantic understanding and association mining capabilities can extract more valuable information from visitor profiles, providing users with a more comprehensive and in-depth visitor profile. This helps users better understand visitor behavior and needs, providing strong support for decision-making. The collection of user feedback data provides a basis for subsequent model optimization, further improving model performance and query accuracy, forming a closed-loop system of continuous optimization.

[0078] Specifically, the user feedback data is processed using differential privacy technology to protect user privacy while providing effective information for model updates; the online update uses an online transfer learning algorithm to quickly adapt to new visitor behavior patterns.

[0079] It should be noted that the user feedback data returned in step S4 is used to perform online updates to the local federated learning model at the edge computing node. Specifically, the model parameters are adjusted based on the user feedback data to optimize the model's query performance. Simultaneously, the feedback data is quantum-encrypted and sent to the cloud coordination server for updating the global model parameters.

[0080] Step S5 achieves real-time model optimization and performance improvement. By timely adjusting model parameters based on user feedback, the model can quickly adapt to changes in visitor behavior patterns, improving its adaptability and query performance. Quantum encryption ensures the security of feedback data during transmission, preventing the risk of data leakage and malicious exploitation, and further protecting data privacy.

[0081] Specifically, the cloud coordination server adopts a distributed collaborative optimization algorithm to integrate the model parameter update information fed back by each edge node and dynamically adjust the global model parameters; the model optimization process introduces a reinforcement learning mechanism, using model query accuracy and response speed as reward signals to continuously optimize model performance.

[0082] It should be noted that the cloud-based coordination server regularly integrates model parameter updates fed back by each edge computing node and uses a distributed collaborative optimization algorithm to optimize and adjust global model parameters. During this optimization process, a reinforcement learning mechanism is introduced, using model query accuracy and response speed as reward signals to continuously improve global model parameters, thereby continuously enhancing the model's generalization and adaptability, ensuring the efficient and stable operation of the entire visitor information query system.

[0083] Through step S6, the global model is continuously optimized and its performance improved. The distributed collaborative optimization algorithm fully utilizes feedback from each edge node to more effectively optimize global model parameters. The introduction of a reinforcement learning mechanism enables the model to dynamically adjust parameters based on actual query results, further improving the model's query accuracy and response speed, enhancing the system's overall performance and stability, and ensuring the long-term, efficient, and stable operation of the entire visitor profile query system.

[0084] This embodiment also provides a method system for querying visitor data in a statistical area, including:

[0085] Data collection module: Deploy multimodal biometric collection equipment to collect visitor biometric data, and use IoT sensors to obtain visitors' environmental interaction data. All data are time-stamped and synchronized before being transmitted to the edge computing node.

[0086] Data fusion module: Cleans and standardizes data at the edge computing node, generates a unique biometric identifier for each visitor, and associates and integrates the processed multi-source heterogeneous biometric data with environmental interaction data to form a fused data set.

[0087] Model training module: Use the fused data set to initialize the deep neural network model, adopt the federated learning mechanism, and only send the encrypted model parameter gradients to the cloud coordination server through the quantum encryption algorithm. The cloud server decrypts and aggregates the data, updates the global model parameters, and quantum encrypts them again and sends them back to each edge node to update the local model. After multiple rounds of iterative training, the local federated learning model is obtained.

[0088] Query matching module: After the user submits a query request, the query conditions are quantized and encoded into a quantum bit query vector. The parallel processing capabilities of quantum computing are used to perform parallel matching calculations with the visitor data stored in the quantized form in the local federated learning model. The highly relevant visitor data is quickly located through quantum gate circuits and entanglement characteristics, and the matching results are output to the analysis step.

[0089] Report generation module: Combines deep learning models to perform semantic understanding and association mining on matching results, generates analysis reports containing detailed information such as visitor biometrics, behavioral trajectories, consumption habits, etc., outputs them to the query terminal for display to users, and transmits user feedback data back to the model update step.

[0090] Feedback update module: Use user feedback data to update the local federated learning model online at the edge computing node, optimize model query performance, and send the feedback data to the cloud coordination server after quantum encryption to update the global model parameters.

[0091] Model optimization module: The cloud coordination server regularly integrates feedback from each edge node to optimize global model parameters, continuously improve the model's generalization ability and adaptability, and ensure the efficient and stable operation of the entire visitor information query system.

[0092] This embodiment also provides a computer device suitable for the method of querying visitor information in a statistical area, including: a memory and a processor; the memory is used to store computer-executable instructions, and the processor is used to execute the computer-executable instructions to implement the method of querying visitor information in a statistical area proposed in the above embodiment.

[0093] The computer device may be a terminal, comprising a processor, a memory, a communication interface, a display screen and an input device connected via a system bus. The processor of the computer device is used to provide computing and control capabilities. The memory of the computer device comprises a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and a computer program. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The communication interface of the computer device is used to communicate with an external terminal in a wired or wireless manner, and the wireless manner may be achieved through WIFI, an operator network, NFC (near field communication) or other technologies. The display screen of the computer device may be a liquid crystal display or an electronic ink display screen, and the input device of the computer device may be a touch layer covering the display screen, or a button, trackball or touchpad provided on the housing of the computer device, or an external keyboard, touchpad or mouse.

[0094] This embodiment also provides a storage medium having a computer program stored thereon, which, when executed by a processor, implements the method for implementing a visitor information query in a statistical area as proposed in the above embodiment; the storage medium can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as static random access memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic memory, flash memory, magnetic disk or optical disk.

[0095] In summary, the present invention effectively improves the security, accuracy, and efficiency of regional visitor profile queries by innovatively combining advanced technologies such as multimodal biometric data collection, federated learning, quantum encryption, quantum computing, and deep learning. First, the fusion collection of multimodal biometric data and pre-processing by edge computing nodes ensures data diversity and quality, providing a foundation for subsequent precise queries. Second, the application of federated learning mechanisms and quantum encryption technology not only protects data privacy but also enables efficient distributed training of models, avoiding the risk of data leakage. Third, the introduction of quantum computing significantly accelerates the query matching process and improves query efficiency. Finally, the semantic understanding and association mining capabilities of the deep learning model can generate detailed visitor profile analysis reports, while continuously optimizing the model using user feedback data, enhancing the adaptability and practicality of the system. Overall, the present invention achieves efficient query and in-depth analysis of regional visitor profiles while ensuring data security and privacy protection, significantly improving user experience and system performance, and providing strong support for regional management and service optimization.

[0096] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit the present invention. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical solutions of the present invention, which should all be included in the scope of the claims of the present invention.

Claims

1. A method for querying visitor data in a statistical area, characterized in that: The following steps are involved: S1: Deploy multimodal biometric acquisition equipment to collect visitor biometric data. Simultaneously, use IoT sensors to obtain visitor environmental interaction data. All data is time-stamped and transmitted to edge computing nodes. The edge computing nodes clean and standardize the data, and generate a unique biometric identifier for each visitor. The processed multi-source heterogeneous biometric data is then correlated and integrated with the environmental interaction data to form a fused dataset. S2: Use the fused dataset to initialize the deep neural network model and adopt a federated learning mechanism. Only the encrypted model parameter gradients are sent to the cloud coordination server through a quantum encryption algorithm. The cloud server decrypts and aggregates the data, updates the global model parameters, and sends them back to each edge node for quantum encryption to update the local model. After multiple rounds of iterative training, the local federated learning model is obtained. S3: After the user submits a query request, the query conditions are quantized and encoded into a qubit query vector. The parallel processing capabilities of quantum computing are used to perform parallel matching calculations with the quantized visitor data stored in the local federated learning model. Quantum gate circuits and entanglement properties are used to quickly locate highly relevant visitor data, and the matching results are output to the analysis step. S4: Combined with a deep learning model, the matching results are semantically understood and associated with mining to generate an analysis report containing detailed information such as the visitor's biometrics, behavioral trajectory, and consumption habits. This report is output to the query terminal for display to the user, and user feedback data is fed back to the model update step. S5: Use user feedback data to update the local federated learning model online at the edge computing node to optimize model query performance. The feedback data is quantum encrypted and sent to the cloud coordination server to update the global model parameters. S6: The cloud coordination server regularly integrates feedback from each edge node to optimize global model parameters, continuously improve the model's generalization ability and adaptability, and ensure the efficient and stable operation of the entire visitor information query system.

2. The method for querying visitor data of a statistical area according to claim 1, wherein: The multimodal biometric feature collection device adopts adaptive sampling technology to dynamically adjust the sampling frequency according to visitor traffic to ensure the real-time and accuracy of data collection; the edge computing node uses an anomaly detection algorithm based on deep learning during the data cleaning process to automatically identify and correct abnormal data.

3. The method for querying visitor data of a statistical area according to claim 2, characterized in that: The deep neural network model adopts a multi-task learning architecture to simultaneously learn visitor biometric classification and behavior pattern prediction tasks, sharing the underlying feature extraction layer to improve the model's generalization ability; the quantum encryption algorithm combines quantum key distribution and quantum state encryption to ensure the transmission security of model parameter gradients.

4. The method for querying visitor data of a statistical area according to claim 3, wherein: The quantum bit query vector is generated by quantum entangled pairs, and each query condition corresponds to an entangled quantum bit pair, achieving efficient and accurate parallel matching; the quantum computing uses a quantum random walk algorithm to accelerate the feature comparison process.

5. The method for querying visitor data of a statistical area according to claim 4, characterized in that: The deep learning model is combined with a generative adversarial network (GAN) to generate virtual samples of visitor behavior patterns, which are used to expand the training data set and improve the robustness of the model; the semantic understanding and association mining adopts a graph neural network (GNN) to construct a visitor behavior knowledge graph.

6. The method for querying visitor data of a statistical area according to claim 5, characterized in that: The user feedback data is processed using differential privacy technology to protect user privacy while providing effective information for model updates; the online update uses an online transfer learning algorithm to quickly adapt to new visitor behavior patterns.

7. The method for querying visitor data of a statistical area according to claim 6, wherein: The cloud coordination server adopts a distributed collaborative optimization algorithm to integrate the model parameter update information fed back by each edge node and dynamically adjust the global model parameters; the model optimization process introduces a reinforcement learning mechanism, using model query accuracy and response speed as reward signals to continuously optimize model performance.

8. A method and system for querying visitor data in a statistical area, based on the method for querying visitor data in a statistical area according to any one of claims 1 to 7, characterized in that: include, Data collection module: Deploy multimodal biometric collection equipment to collect visitor biometric data, and use IoT sensors to obtain visitors' environmental interaction data. All data are time-stamped and synchronized before being transmitted to the edge computing node. Data fusion module: Cleans and standardizes data at the edge computing node, generates a unique biometric identifier for each visitor, and associates and integrates the processed multi-source heterogeneous biometric data with environmental interaction data to form a fused data set. Model training module: Use the fused data set to initialize the deep neural network model, adopt the federated learning mechanism, and only send the encrypted model parameter gradients to the cloud coordination server through the quantum encryption algorithm. The cloud server decrypts and aggregates the data, updates the global model parameters, and quantum encrypts them again and sends them back to each edge node to update the local model. After multiple rounds of iterative training, the local federated learning model is obtained. Query matching module: After the user submits a query request, the query conditions are quantized and encoded into a quantum bit query vector. The parallel processing capabilities of quantum computing are used to perform parallel matching calculations with the visitor data stored in the quantized form in the local federated learning model. The highly relevant visitor data is quickly located through quantum gate circuits and entanglement characteristics, and the matching results are output to the analysis step. Report generation module: Combines deep learning models to perform semantic understanding and association mining on matching results, generates analysis reports containing detailed information such as visitor biometrics, behavioral trajectories, consumption habits, etc., outputs them to the query terminal for display to users, and transmits user feedback data back to the model update step. Feedback update module: Use user feedback data to update the local federated learning model online at the edge computing node, optimize model query performance, and send the feedback data to the cloud coordination server after quantum encryption to update the global model parameters. Model optimization module: The cloud coordination server regularly integrates feedback from each edge node to optimize global model parameters, continuously improve the model's generalization ability and adaptability, and ensure the efficient and stable operation of the entire visitor information query system.

9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, the steps of the method for querying visitor information in a statistical area according to any one of claims 1 to 7 are implemented.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the method for querying visitor information in a statistical area according to any one of claims 1 to 7 are implemented.

Images