Information security acquisition system and method for smart park

Abstract

The invention discloses an information security acquisition system and method for a smart park, and the system comprises distributed sensor nodes which are in communication connection with a gateway device, and a central server which is connected with the gateway device through a network and is responsible for receiving, storing, analyzing and managing data from a gateway; the safety protection module is used for carrying out joint training and learning on the multi-modal data based on a recurrent neural network to form a comprehensive information acquisition model; in combination with a fuzzy extractor and a block chain technology, the training process of the information acquisition model is optimized, and the generalization ability and stability of the information acquisition model are improved; wherein the sensor node also has a self-diagnosis function, and can automatically detect a self fault and send a maintenance request to the gateway. According to the method, the park management is greatly optimized through a digital means, the cost is greatly reduced, the decision-making quality is greatly improved, and finally the core competitiveness is constructed through the efficiency.

Description

Technical Field

[0001] The present invention relates to the technical field of smart parks, and particularly to an information security collection system and method for smart parks. Background Art

[0002] In recent years, the construction of smart parks has become a representative information-based facility and a new trend in the development of global parks. However, most of the current research on smart parks focuses on how to achieve "intelligence", while its "security" has not received sufficient attention, and the security risks that may be generated by the most common "people" in smart parks have not been focused on.

[0003] There is currently a lack of management for users in smart parks, and the security requirements are still relatively high. There is an urgent need to uniformly manage the data of all intelligent devices in the park and make full use of the collected data to authenticate and control users. At the same time, the mobile logs of users should be recorded to facilitate the tracing of mobile trajectories afterwards, and truly use the park's intelligent devices to provide data support for park security management. The smart park platform only realizes the collection of some system business data within the park. During the process of data collection and data aggregation, emphasis is placed on obtaining data while neglecting data quality. At the same time, the data that the park system can directly collect is very limited.

[0004] In view of the deficiencies of traditional information collection, the present invention provides an information security collection system and method for smart parks. Summary of the Invention

[0005] The purpose of this part is to outline some aspects of the embodiments of the present invention and briefly introduce some preferred embodiments. Simplifications or omissions may be made in this part, as well as in the abstract and title of the present application, to avoid obscuring the purpose of this part, the abstract, and the title. However, such simplifications or omissions shall not be used to limit the scope of the present invention.

[0006] To solve the above technical problems, the present invention provides the following technical solutions: An information security collection system for a smart park, characterized by comprising:

[0007] Distributed sensor nodes for real-time monitoring and collection of physical environment data, device operation status information, and network traffic within the smart park;

[0008] At least one gateway device communicatively connected to the multiple sensor nodes for aggregating the data of each node and performing preliminary processing;

[0009] A central server connected to the gateway device through a network, responsible for receiving, storing, analyzing, and managing the data from the gateway;

[0010] A security protection module, integrated in the central server or set independently, provides intrusion detection, abnormal behavior recognition, and threat response mechanisms;

[0011] A user terminal interface that allows authorized users to access the system and obtain reports or alerts;

[0012] Jointly train and learn multi-modal data based on a recurrent neural network to form a comprehensive information collection model;

[0013] Combine a fuzzy extractor and a blockchain to optimize the training process of the information collection model and improve the generalization ability and stability of the information collection model;

[0014] Wherein the sensor node also has a self-diagnosis function, which can automatically detect its own faults and send maintenance requests to the gateway.

[0015] As a preferred solution of the information security collection system for a smart park according to the present invention, the gateway device further includes an encryption unit for encrypting the data in transmission to ensure the security of the data.

[0016] As a preferred solution of the information security collection system for a smart park according to the present invention, the central server further includes a machine learning algorithm library for continuously optimizing the data analysis model and improving the accuracy and efficiency of the system.

[0017] As a preferred solution of the information security collection system for a smart park according to the present invention, the security protection module adopts a multi-layer defense architecture, combined with a firewall, an intrusion prevention system, and anti-virus software to form a comprehensive security barrier.

[0018] As a preferred solution of the information security collection system for a smart park according to the present invention, the user terminal interface supports multiple access methods, including but not limited to a Web browser, a mobile application, and a desktop client.

[0019] As a preferred solution of the information security collection system for a smart park according to the present invention, it further includes a cloud service platform for expanding storage capacity, enhancing computing power, and realizing remote monitoring.

[0020] As a preferred solution of the information security collection method for a smart park according to the present invention, it includes the following steps:

[0021] Step 1: Real-time collect various types of information in the park through distributed sensor nodes;

[0022] Step 2: Upload the collected data to the central server via the gateway device;

[0023] Step 3: Analyze the data on the central server using preset rules or machine learning models;

[0024] Step 4: Trigger corresponding security measures or issue warnings to users based on the analysis results.

[0025] As a preferred solution of the information security acquisition method for smart campuses described in the present invention, the S1 includes visual sensors, sound sensors, and dynamic sensors.

[0026] As a preferred solution of the information security acquisition method for smart campuses described in the present invention, the S2 includes neural network algorithms, similarity extraction algorithms, and analytic hierarchy process to improve the information acquisition efficiency. The neural network algorithm includes the ant colony algorithm. By improving the ant colony algorithm and dynamically setting the evaporation factor, the algorithm performance is enhanced. By improving the calculation metrics of the similarity extraction algorithm and the support vector machine algorithm, data with high similarity is extracted, and then it is determined whether the data is extracted based on the similarity and transmitted to the central server.

[0027] Advantages of the present invention: By adopting distributed sensor nodes, the present invention collects campus information, uses a recurrent neural network to form a neutral information acquisition model for the collected information data, and combines neural network algorithms with fuzzy extractors and blockchain technology, neural network algorithms, similarity extraction algorithms, and analytic hierarchy process to improve the information acquisition efficiency, improve the accuracy, real-time performance, and stability of information security acquisition, reduce the error rate of information acquisition, and ensure the security of smart campuses. Description of the Drawings

[0028] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following will briefly introduce the drawings required for the description of the embodiments. Obviously, the following drawings are only some embodiments of the present invention. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0029] Among them:

[0030] Figure 1 is the system block diagram of the information security acquisition system and method for smart campuses provided by an embodiment of the present invention;

[0031] Figure 2 is the flow chart of the acquisition method steps of the information security acquisition system and method for smart campuses provided by an embodiment of the present invention. Detailed Embodiments

[0032] To make the above objects, features, and advantages of the present invention more obvious and understandable, the following will provide a detailed description of the specific embodiments of the present invention with reference to the drawings in the specification.

[0033] In the following description, many specific details are set forth in order to provide a thorough understanding of the present invention. However, the present invention may be practiced in other ways different from those described herein. Those skilled in the art can make similar extensions without departing from the spirit of the present invention. Therefore, the present invention is not limited by the specific embodiments disclosed below.

[0034] Embodiment 1

[0035] Referring to Figure 1 , the first embodiment of the present invention provides an information security acquisition system for an intelligent park, which is characterized by including:

[0036] Distributed sensor nodes for real-time monitoring and acquisition of physical environment data, device operation status information, and network traffic within the intelligent park;

[0037] At least one gateway device communicatively connected to the multiple sensor nodes for aggregating data from each node and performing preliminary processing;

[0038] A central server connected to the gateway device through a network, responsible for receiving, storing, analyzing, and managing data from the gateway;

[0039] A security protection module integrated in the central server or independently provided, providing intrusion detection, abnormal behavior recognition, and threat response mechanisms;

[0040] A user terminal interface allowing authorized users to access the system and obtain reports or alerts;

[0041] Joint training and learning of multi-modal data based on a recurrent neural network to form a comprehensive information acquisition model;

[0042] Combining a fuzzy extractor and blockchain technology to optimize the training process of the information acquisition model and improve the generalization ability and stability of the information acquisition model;

[0043] Wherein the sensor node also has a self-diagnosis function and can automatically detect its own faults and send maintenance requests to the gateway.

[0044] Specifically, multiple gateway devices are provided, further including an encryption unit for encrypting the data in transmission to ensure data security. The central server also includes a machine learning algorithm library for continuously optimizing the data analysis model to improve the accuracy and efficiency of the system. The user terminal interface supports multiple access methods, including but not limited to web browsers, mobile applications, and desktop clients. The user interface terminal is the interface for users to interact with the computer system; currently, due to the low attention of the park to mobile personnel, there is no perfect solution for controlling the entry and exit of personnel. During special periods, the park can only control the entry and exit of personnel through traditional manual management methods, without achieving the so-called intelligence. To ensure the safe operation of the park, it is urgent to manage the data of various intelligent devices in the park, including card readers, infrared detectors, cameras, etc., on a unified platform, and make full use of the linkage effect of the devices to authenticate and manage personnel. Common methods for identity authentication include password-based authentication, smart card-based authentication, biometric-based authentication, etc. However, since these solutions only use a single factor for authentication, problems such as forgotten passwords, easy replication of fingerprints, lost cards, and errors in face recognition are likely to occur, posing security risks. Although the use of multiple factors increases security, it reduces the user experience, and the authentication time is prolonged. In most of the solutions that use a combination of multiple factors for authentication, "password" is used as one of the factors. However, in the smart park, users may not have time to enter passwords when passing through the access control, such as entering the park gate during the peak working hours. Considering that the traffic efficiency in the park is relatively important, the present invention selects to use two factors, namely RFID cards and biometric features, to complete the user identity authentication, which can not only improve security but also provide a very good user experience. In terms of biometric features, face information is selected first to avoid contact with devices and is not as easy to replicate as fingerprints. At the same time, face recognition technology has been relatively mature and can be combined with RFID information to quickly complete authentication. When performing access control on users, there are problems such as policy conflicts caused by multi-party joint authorization, inability to detect implicit conflicts, failure to detect redundant policies, and low efficiency when traversing all policies during detection. After considering the problems existing in the current identity authentication and access control technologies, a two-factor authentication protocol based on face information and RFID information is proposed to achieve safe and reliable identity authentication. This protocol does not simply superimpose the two authentication methods, but designs the specific authentication process. At the same time, compared with existing multi-factor solutions, it can resist more security risks and requires lower computational overhead on user terminals with weaker computing power;

[0045] The present invention improves the fuzzy extractor, and improves the algorithm performance by means of a string with random distribution, stable recoverability and confidentiality. The fuzzy extractor includes a generation algorithm Ben(·) and a regeneration algorithm Rep(·). The fuzzy extractor uses a noise random source with a certain error to generate a required uniform, random and accurately reproducible string;

[0046] Ben(Bio)=(τ,σ)

[0047] The generation algorithm requires the input of a biometric feature Bio (i.e., a single sampling of the noise random source), and the output is a recoverable biometric string σ and an auxiliary string τ;

[0048] Rep(Bio',τ)=σ'

[0049] The regeneration algorithm requires the input of a biometric feature Bio′ (i.e., another sampling of the noise random source) and the auxiliary string τ. If the distances between the two samplings Bio′ and Bio are close enough, the biometric string σ′ = σ can be successfully output.

[0050] Attribute-based access control technology can effectively make decisions and authorizations on user requests according to the relevant attributes of the subject, object and environment, so as to achieve more efficient and secure management and control. The access control model consists of three entities: the subject, the object and the environment. Among them, the subject is the party that requests access to the object, the object is the object that can satisfy the subject's access request, and the environment generally refers to the objective state of the outside world when this request occurs.

[0051] Furthermore, multiple sensor nodes are fused through blockchain technology, enabling them to share the same ledger data and adopting a unified method to ensure the security and credibility of the data. Blockchain itself is a chain structure. In the blockchain, all data is stored in blocks, sorted according to the time sequence, and each block consists of a block header and a block body. When a user makes an access request, identity authentication is first required to confirm their legal identity. The present invention authenticates the user identity based on two-factor identity authentication. First, authentication is performed in the initialization stage of identity authentication, which mainly includes a system initialization algorithm, a device key initialization algorithm, and a user key initialization algorithm.

[0052] 1. System initialization algorithm: This algorithm is run by the central server. The output is the system public parameter mpk and the master private key msk. The central server publishes mpk and keeps msk confidential. The specific operation process is as follows:

[0053] (1) Select an elliptic curve E on a finite field p (a, b), satisfying

[0054] 4a3 +27b ≠ 0 (mod p),

[0055] where p is a large prime number, and a point P on the elliptic curve is selected as the base point;

[0056] (2) Randomly select SK as the system private key, and calculate the system public key PK = SK·P;

[0057] (3) Select a hash function

[0058] (4) Initialize the generation algorithm Gen(·) and the regeneration algorithm Rep(·) of the fuzzy extractor;

[0059] (5) Publicize the system parameters mmpk = {E P (a, b), p, P, h, PK, Gen(·), Rep(·)} and save msk = {SK}.

[0060] 2. Device key initialization algorithm: This algorithm is run by the central server, and the output is the authentication device identity information and the public-private key pair. The central server assigns an identity ID to the authentication device j , selects the device private key mk j , and the public key pk j = sk j P, and sends ID j and mk j to the authentication device for storage.

[0061] 3. User key initialization algorithm: This algorithm is run by the central server. The input is the user's identity information and face information, and the output is the user's RFID tag information. The specific running process is as follows:

[0062] (1) Enter the user's face information Bio i , run the function Ben(Bio i ) = (τ i , σ i ), obtain the biometric string σ i and the recovery string τ i , calculate the hash value k(σ i );

[0063] (2) Set the access validity period T i for the user, and use the central server private key SK for signature: calculate k it = k(ID i , T i ) Select a random number s i , calculate s i P = (x i , y i), calculate H sign =(k it +SKx i ) / s i , the signed message is sig i ={s i P, H sign};

[0064] (3) Input the user identity information ID i , select the user's private key mk i , and the public key pk i =sk i P;

[0065] (4) Store {ID i , mk i , k(σ i ), T i , sig i , mpk} in the user's RFID card, and the central server stores {ID i , pk i , τ i}.

[0066] When performing user identity authentication, two parts of information need to be integrated, namely face information and RFID tag information. The participants in the authentication phase include a camera, an RFID tag, an RFID reader, and an authentication device.

[0067] 1. The camera obtains the face information Bio′ i , sends it to the authentication device, and waits for subsequent two-factor authentication together with the information obtained from the RFID tag.

[0068] 2. After the reader senses the presence of an RFID tag around, it generates a random number s R and the authentication request information request, and sends them to the RFID tag.

[0069] 3. After receiving the authentication request, the tag calculates uses the user's private key mk i to sign the timestamp t i at the current moment and the information k(σ i ) it stores: Calculate k r =k(t i , k(σ i ))), generate a random number s r , calculate s r P=(x r , y i ), calculate d r =(k r +mk i x r) / s r , the signed message is sig r = {s r P, d r}; Send {S i , t i , k(σ i ), sig r , T i, sig i} to the card reader.

[0070] 4. After the card reader receives the message, it encrypts the random number using the device public key pk j : Generate a random number r b , calculate r b P, calculate c R = s R + r b pk j , and the ciphertext is c i = {r b P, c R}, and send it to the authentication device together with the information sent by the tag.

[0071] 5. After the authentication device receives the message, it performs the following operations:

[0072] (1) Verify whether t now satisfies t i - t now < Δt according to the current time t i ;

[0073] (2) Calculate s j using the private key mk R = c R - mk j r b P, and obtain the user identity ID i ;

[0074] (3) Verify whether the user identity is within the validity period T i and verify the correctness of the validity period by verifying sig i : Calculate k' it = k(ID i , T i ), and verify whether k' it P / T sign + x i PK / T sign is equal to r i P. If they are equal, continue;

[0075] (4) Verify sig r : Calculate k'r = k(t i , k(σ i )) and use the user's public key pk i to verify k' r P / s r + x r pk i / s r is equal to r r P. If they are equal, continue;

[0076] (5) Request the user's recovery string τ from the central server i , calculate Re p(Bio' i , τ i ) = σ' i , and determine whether k(σ' i ) is equal to;

[0077] If all of the above verifications pass, the user's identity is considered verified and enter the access control phase. If any step of the verification fails, the identity authentication fails.

[0078] After passing the identity verification, the data is transmitted to the background control terminal in combination with the security protection module. The security protection module adopts a multi-layer defense architecture, combined with a firewall, an intrusion prevention system (IPS), and anti-virus software to form a comprehensive security barrier. Through the cloud service platform, for expanding storage capacity, enhancing computing power, and realizing remote monitoring.

[0079] Embodiment 2

[0080] In the second embodiment of the present invention, different from the previous embodiment, this embodiment provides a collection method to solve and optimize the information security collection method of the smart park, including:

[0081] S1: Collect various types of information in the park in real time through distributed sensor nodes;

[0082] S2: Upload the collected data to the central server via the gateway device;

[0083] S3: Analyze the data on the central server using preset rules or machine learning models;

[0084] S4: Trigger corresponding security measures or issue warnings to users based on the analysis results.

[0085] Specifically, the sensors in S1 include visual sensors, sound sensors, and dynamic sensors. Among them, in an alternative embodiment, the distributed sensor nodes include visual data, sound data, and dynamic data; it also includes the process of updating system configurations or adjusting parameters to adapt to newly emerging security threats.

[0086] As an example, sound data is a sound signal collected by a sound sensor and is used to monitor different sounds in the park, such as people talking and abnormal sounds, to help monitor whether the park is in normal condition.

[0087] As an example, visual data is image or video data collected by a camera visual sensor and is used to observe the actual operation of the park, such as monitoring whether there are any abnormal conditions in the park.

[0088] As an example, dynamic data is images or videos collected by camera sensors and used to monitor changes in objects or people within the park, such as the entry and exit of people, the opening of security doors, and dynamic changes within park buildings, helping to monitor whether the park status is normal.

[0089] In summary, by comprehensively considering the data of vision, sound, dynamics, and other environmental parameters (such as temperature, humidity, pressure, current, etc.), various types of information in the park can be collected in real time through distributed sensor nodes, so as to comprehensively monitor the status of the park and detect abnormal situations in time.

[0090] The S2 includes a neural network algorithm, a familiarity extraction algorithm and a hierarchical analysis method to improve the efficiency of information collection. The neural network algorithm includes an ant colony algorithm. The present invention improves the ant colony algorithm by dynamically setting the volatility factor, extracts data with high familiarity through the calculation indicators of the familiarity extraction algorithm and the vector support machine algorithm, and then determines whether the data is extracted based on the familiarity and transmits it to the central server. The algorithm steps are as follows:

[0091] (1) Social hierarchy: The social hierarchy of ants can be constructed by calculating the fitness of each individual in the population. First, all individuals in the population are sorted by fitness, and the three ants with the highest fitness are selected and labeled as α ant, β ant, and δ ant. The rest of the ants are labeled as ω ants.

[0092] (2) Surrounding prey: The ant colony will constantly update its position during the search for prey, thus forming a circle to find prey. Its mathematical model can be expressed as:

[0093]

[0094] Where D represents the distance between the ant colony and the prey; “·” represents the element-by-element product (Hadamard product); A and C are the coordination coefficient vectors; t and t+1 represent the current iteration and the next iteration respectively; X represents the position vector of the ant colony; X p Represents the position vector of the prey. The calculation formula is as follows: D(t+1) represents the next iteration, C is the synergy coefficient vector, X pis the position vector of the prey, and X(t) represents the position vector of the ant colony in the current iteration. The calculation formulas of A and C are:

[0095]

[0096] Where a represents a linearly decreasing number from 2 to 0; r1 and r2 are random vectors from the interval [0,1].

[0097] (3) Hunting: The ant colony identifies prey with the guidance of ant α, ant β, and ant δ. However, due to the complexity of the problem solution and the unclear spatial characteristics, the solution obtained may not be accurate enough. In order to more accurately locate the optimal solution, it is assumed that these three ants have strong hunting and positioning capabilities. Then, through multiple iterations, the three ants with the best fitness are retained one by one, and the location of the optimal solution is gradually determined based on their positions. The corresponding formula is as follows:

[0098]

[0099] Where, X α , X β , X δ The position vectors corresponding to α ant, β ant, and δ ant, respectively, D α , D β , D δ They represent the distances between the current ant and the three best ants. Then we can further analyze the positions of the candidate ants:

[0100]

[0101] The position of the candidate ant can be determined as:

[0102]

[0103] In the formula, X(t+1) represents the position vector of the candidate ant. Combined with mathematical analysis, it can be seen that when |A| is greater than 1, it means that the ants are relatively scattered and are expanding their search range for prey; when |A| is less than 1, it means that the ants are beginning to concentrate and are looking for prey by searching a specific range.

[0104] (4) Application: The behavior pattern of this ant colony algorithm is similar to the behavior of ant colonies in reality. Ants α, β, and δ can be regarded as leaders, who guide other ants to find prey through their foraging experience and insight. Other candidate ants constantly adjust their positions under the guidance of the leaders to form a circle in order to better capture prey.

[0105] Therefore, the optimization algorithm based on this theory has certain application advantages. It can perform global search in the search space, and dynamically adjust the search scope and direction, so as to better find the optimal solution.

[0106] Support Vector Machine (SVM) can efficiently handle high-dimensional data, that is, the case where the number of features is much larger than the number of samples. The final decision function is only determined by a few support vectors, which means that the model is insensitive to the changes of most points in the training set, thus improving the stability of the model. Although SVM was initially designed for binary classification tasks, it can be easily extended to multi-class classification problems through various strategies, such as one-versus-rest, one-versus-one combination mode, or constructing a decision tree structure. Due to its principle of maximum margin and the concept of soft margin (allowing a certain degree of misclassification), SVM can avoid the problem of overfitting to a certain extent. It can improve the classification effect of image recognition and speech recognition. Compared with traditional statistical methods, SVM does not strictly rely on assumptions such as probability measures or the law of large numbers, so it can be applied under a wider range of conditions. SVM can be used to construct a model called "one-class SVM" to identify new instances that do not belong to known classes, and is often used in scenarios such as fraud detection and intrusion detection. By analyzing the importance of support vectors, it can help understand which features are most critical for classification, and thus assist the feature engineering process. In addition to classification, SVM can also be used to predict continuous variables, that is, for regression analysis. The regression formula of Support Vector Machine (SVM) is defined as:

[0107] f(x i )=w i ·φ(x i )+b (2-1)

[0108] In the formula, f(x i ) is the regression function; w i is the definable weight vector; φ(x i ) is the vector mapping; b is the bias, also known as offset or threshold. At this time, the insensitive loss function L ε can be used to evaluate the error between the predicted variable y i and the regression function f(x i ).

[0109]

[0110] In the formula, ε intervenes in the fitting ability of the approximation function by affecting the number of support vectors. Combining with hyperplane analysis, w i and b can be further determined as:

[0111]

[0112] In the formula, ξi and are the introduced slack variables to measure the sample deviation outside the insensitive region. In the linearly separable case, find the optimal solutions of w i and b, at which time the objective function value is minimized:

[0113]

[0114] Introduce the dual form of the Lagrange multiplier. Based on the KKT (Karush-Kuhn-Tucker) theory, the objective and constraint conditions can be transformed into a quadratic equation in two variables and a linear function:

[0115]

[0116] where a i and are the values of the Lagrange multipliers. Furthermore, the regression function of the SVM model is obtained:

[0117]

[0118] where K(x, x i ) is the kernel function. The kernel function adopted in the SVM algorithm of the present invention is the radial basis function (RBF), which has the advantages of strong non-linear modeling ability, good global approximation ability, being unrestricted by the input dimension and having high-dimensional feature mapping, as well as interpretability. Its corresponding formula is as follows:

[0119] K(x, y) = exp(-||x - y|| 2 / 2σ 2 ) (2 - 7)

[0120] where σ is the width of the REF kernel.

[0121] Finally, combine the ant colony algorithm with the support vector machine algorithm to achieve global search optimization and improve efficiency.

[0122] In summary, through the above description, the collection of information security in the park can be made safer, and through digital means, great optimization of management, great reduction of costs, great improvement of decision-making quality, and ultimately the construction of core competitiveness through efficiency can be achieved; while improving the user experience and efficiency through digital means, it will bring changes to the organizational mode and governance method of the entire park, stimulate the vitality of the organization, and bring long-term changes.

[0123] In addition, to provide a concise description of the exemplary embodiments, all features of the actual embodiments may not be described (i.e., those features that are not relevant to the currently considered best mode of implementing the present invention, or those features that are not relevant to implementing the present invention).

[0124] It should be understood that, during the development of any actual implementation, such as in any engineering or design project, a large number of specific implementation decisions can be made. Such development efforts may be complex and time-consuming, but for those of ordinary skill in the art who benefit from this disclosure, without undue experimentation, such development efforts will be routine work in design, manufacturing, and production.

[0125] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit them. Although the present invention has been described in detail with reference to the preferred embodiments, those of ordinary skill in the art should understand that the technical solutions of the present invention can be modified or equivalently replaced without departing from the spirit and scope of the technical solutions of the present invention, and they should all be covered by the scope of the claims of the present invention.

Claims

1. An information security acquisition system for an intelligent park, characterized in that Including: Distributed sensor nodes for real-time monitoring and collection of physical environment data, device operation status information, and network traffic within a smart park; At least one gateway device communicatively connected to multiple sensor nodes for aggregating data from each node and performing preliminary processing; A central server connected to the gateway device via a network, responsible for receiving, storing, analyzing, and managing data from the gateway; A security protection module integrated in the central server or independently provided, offering intrusion detection, abnormal behavior recognition, and threat response mechanisms; A user terminal interface allowing authorized users to access the system and obtain reports or alerts; Joint training and learning of multimodal data based on a recurrent neural network to form a comprehensive information collection model; Combining a fuzzy extractor and blockchain technology to optimize the training process of the information collection model and improve the generalization ability and stability of the information collection model; Wherein the sensor node further has a self-diagnosis function capable of automatically detecting its own faults and sending a maintenance request to the gateway.

2. The information security acquisition system for an intelligent park according to claim 1, wherein The gateway device further includes an encryption unit for encrypting data during transmission to ensure data security.

3. The information security acquisition system for an intelligent park according to claim 1, characterized in that: Wherein the central server further includes a machine learning algorithm library for continuously optimizing the data analysis model and improving the accuracy and efficiency of the system.

4. The information security acquisition system for an intelligent park according to claim 1, characterized in that, The security protection module adopts a multi-layer defense architecture, combining a firewall, intrusion prevention system, and anti-virus software to form a security barrier.

5. The information security acquisition system for an intelligent park according to claim 1, characterized in that: Wherein the user terminal interface supports multiple access methods, including but not limited to a web browser, mobile application, and desktop client.

6. The information security acquisition system for an intelligent park according to claim 1, wherein It further includes a cloud service platform for expanding storage capacity, enhancing computing power, and enabling remote monitoring.

7. An information security acquisition method for a smart park as described in any one of claims 1-6, characterized in that, Including: S1: Real-time collection of various types of information within the park through distributed sensor nodes; S2: Uploading the collected data to the central server via the gateway device; S3: Analyzing the data on the central server using preset rules or machine learning models; S4: Triggering corresponding security measures or issuing warnings to users based on the analysis results.

8. The information security acquisition method for an intelligent park according to claim 7, characterized in that: The S1 includes visual sensors, sound sensors, and dynamic sensors.

9. The information security collection method for an intelligent park according to claim 7, characterized in that: The S2 includes neural network algorithms, similarity extraction algorithms, and analytic hierarchy process to improve information collection efficiency. The neural network algorithms include the ant colony algorithm, which improves the algorithm performance by dynamically setting the evaporation factor, extracts data with high similarity by improving the calculation metrics of the similarity extraction algorithm and the support vector machine algorithm, and then determines whether the data is extracted based on the similarity and transmits it to the central server.

Images