Remote control and authority management system for intelligent lock of electric power facility

Abstract

The invention relates to the technical field of intelligent locks, in particular to an electric power facility intelligent lock remote control and authority management system which comprises an intelligent lock module, an identity verification module, an authority management module, a multi-device collaborative optimization module and a data recording and analyzing module. The intelligent lock module is physically connected with a lock of an electric power facility through a prefabricated mounting interface; the identity verification module is used for carrying out multi-mode identity verification on an operator; the authority management module dynamically allocates authority levels of operators; the multi-device collaborative optimization module optimizes operation and resource allocation among devices; and the data recording and analyzing module records an operation log of the intelligent lock in real time, analyzes the operation log, detects an abnormal behavior and automatically performs permission adjustment. According to the invention, high-efficiency cooperation of equipment and safety, flexibility and operation efficiency of the system are ensured, and complex electric power facility management requirements are effectively met.

Description

Technical Field

[0001] The present invention relates to the technical field of smart locks, and in particular to a remote control and authority management system for smart locks of electric power facilities. Background Art

[0002] As the scale and complexity of power facilities continue to increase, the security, stability and efficiency management of power systems has become crucial. Traditional power facilities usually rely on physical keys for lock management and maintenance operations, but this management method has many inconveniences, especially in remote monitoring, authority management and multi-device collaboration. In order to improve the management efficiency of power facilities, more and more smart lock technologies are being used in the management of power facilities. Through the combination of smart locks and remote control systems, the safe management of key equipment can be achieved. At the same time, the development of identity authentication technology also provides a guarantee for system security.

[0003] At present, although some power facilities have begun to apply smart locks and identity authentication technologies, the existing systems have many shortcomings. First of all, the existing smart locks mostly use a single identity authentication method, which is easily affected by misjudgment and external attacks and has low security. In addition, the existing smart lock system lacks dynamic authority allocation and multi-device collaborative optimization mechanism. The operator's authority is often statically allocated and cannot be dynamically adjusted according to actual operation needs, resulting in low management efficiency. For resource and task allocation in a multi-device environment, the existing system also lacks effective collaborative optimization and cannot be intelligently adjusted according to the correlation between devices and operation requirements. In addition, the anomaly detection function of the existing system is imperfect and cannot analyze the operation logs and environmental data of the equipment in real time, resulting in safety hazards not being discovered and handled in time.

[0004] In order to address the deficiencies in the prior art, the present invention provides a remote control and authority management system for smart locks of power facilities, aiming to improve the security, flexibility and management efficiency of power facilities and ensure efficient and stable operation of equipment in complex environments. Summary of the invention

[0005] The present invention provides a remote control and authority management system for intelligent locks of electric power facilities.

[0006] The remote control and authority management system for smart locks of electric power facilities includes a smart lock module, an identity authentication module, an authority management module, a multi-device collaborative optimization module, and a data recording and analysis module, among which;

[0007] The smart lock module is physically connected to the lock of the power facility through a prefabricated installation interface, performs locking, unlocking and status feedback operations, and communicates with the remote server through the network;

[0008] The identity verification module performs multi-modal identity verification on the operator, combining multiple biometric features for comprehensive verification;

[0009] The authority management module dynamically allocates the authority level of operators according to the results of the comprehensive verification, and through multi-level authority management, enables personnel with different authority levels to operate power facilities within their authorized scope;

[0010] The multi-device collaborative optimization module collaboratively optimizes the operation and authority allocation of multiple smart lock devices in the power facility, and optimizes the operation and resource allocation between devices by analyzing the association relationship between smart lock devices, specifically including:

[0011] Smart lock device feature extraction: Extract key features from each smart lock device, including the physical location of the device, usage frequency, permission allocation, and operation history, analyze the correlation between smart lock devices, and build a correlation graph between smart lock devices to define the relationship between smart lock devices;

[0012] Collaborative optimization calculation: Based on the association graph between smart lock devices, the graph neural network (GNN) model is used to optimize the operation and permission allocation of multiple smart lock devices;

[0013] Strategy adjustment: Based on the optimization results of the operation and authority allocation of multiple smart lock devices, dynamically adjust the coordination strategy between devices, including dynamic authority allocation adjustment, operation priority adjustment, resource allocation adjustment, load balancing adjustment and emergency response adjustment;

[0014] The data recording and analysis module records the operation log of the smart lock in real time, including identity authentication information, operation time, authority allocation, lock status and environmental data, analyzes the operation log and detects abnormal behavior, and automatically adjusts the authority when an abnormal situation is detected.

[0015] Optionally, the smart lock module includes:

[0016] Fixed installation of smart lock module: physical connection with the physical lock of power facilities through prefabricated installation interface;

[0017] Execute locking and unlocking operations: The smart lock module is integrated with a motor drive device to perform locking and unlocking operations, and control the opening and closing of the lock according to the instructions of the authority management module;

[0018] State feedback detection: The smart lock module is equipped with a state sensor to detect the current state of the lock (locked or unlocked) in real time and generate state feedback information;

[0019] Network communication and remote control: Establish a communication connection with the remote server through a wireless communication unit (Wi-Fi, cellular network or dedicated communication protocol), upload the lock status feedback information to the server in real time, and receive operation instructions from the remote server to realize remote control and status monitoring functions.

[0020] Optionally, the identity authentication module includes:

[0021] Biometrics collection and preprocessing: responsible for collecting various biometrics of operators, including fingerprint images, facial and iris images, and voice data, and preprocessing the collected biometrics, including noise filtering and formatting;

[0022] Multimodal feature fusion and verification: A deep fusion algorithm is used to fuse the pre-processed biometric features, and the fused biometric features are authenticated through pre-stored template data to generate a comprehensive verification result.

[0023] Optionally, the biometric feature collection and preprocessing includes:

[0024] Biometric collection: fingerprint images are acquired through fingerprint sensors, facial and iris images are acquired through cameras, and voice data is acquired through microphones;

[0025] Noise filtering: Use Gaussian filters to filter noise from the collected fingerprint images, facial and iris images, and use bandpass filters to reduce noise from voice data;

[0026] Data formatting: The fingerprint image, face and iris image after noise filtering are scaled by bilinear interpolation, and the signal amplitude of the voice data after noise reduction is adjusted to a fixed range by normalization.

[0027] Optionally, the multimodal feature fusion and verification includes:

[0028] Feature extraction: For each biometric feature (fingerprint, face, iris, voice), a convolutional neural network (CNN) is used to extract features and generate feature vectors;

[0029] Feature fusion: After feature extraction is completed, the extracted feature vectors are fused using the fully connected layer to generate a comprehensive feature vector F fused ;

[0030] Identity verification: The fused comprehensive feature vector F fused Matching with pre-stored template data, and determining identity matching by calculating cosine similarity;

[0031] Generation of comprehensive verification results: Based on the cosine similarity results, determine whether the matching degree reaches the set verification threshold. If the similarity exceeds the set verification threshold, the verification is passed, a comprehensive verification result is generated, and a verification pass signal is sent to the authority management module.

[0032] Optionally, the rights management module includes:

[0033] Permission level setting: pre-define multiple permission levels, each of which corresponds to different operation scopes and access rights, including basic operation permissions, intermediate maintenance permissions, and advanced management permissions;

[0034] Dynamic authority allocation: Compare the comprehensive verification results output by the identity authentication module with the preset upper and lower limits of the credibility threshold, and divide the identity credibility of the operator into basic credibility (the comprehensive verification result is lower than the lower limit of the credibility threshold), intermediate credibility (the comprehensive verification result is between the lower and upper limits of the credibility threshold), and advanced credibility (the comprehensive verification result is higher than the upper limit of the credibility threshold). According to the credibility level, the corresponding authority level is automatically allocated, including basic operation authority for basic credibility, intermediate maintenance authority for intermediate credibility, and advanced management authority for advanced credibility;

[0035] Multi-level authority management: Through the multi-level authority management mechanism, based on the results of dynamic authority allocation, operators with different authority levels can operate and control power facilities within their authorized scope.

[0036] Optionally, the smart lock device feature extraction includes:

[0037] Feature data collection: Collect key feature data from each smart lock device, including the physical location P i 、Usage frequency U i 、Permission allocation A i and operation history H i ,in:

[0038] Physical location P i Determined by GPS or network topology information;

[0039] Frequency of use i Calculated by the number of times or time intervals of equipment operation;

[0040] Permission AssignmentA i Set up collection based on current device permissions;

[0041] Operation History i Includes past operation records, error logs, and maintenance records;

[0042] Correlation analysis: Analyze the characteristic data of each smart lock device, calculate the correlation between smart lock devices, and calculate the correlation R between smart lock device i and device j through multi-factor comprehensive analysis. ij ;

[0043] Association graph construction: based on the calculated association R ij , construct the association graph G = (V, E) between smart lock devices, where V represents the smart lock device node, E represents the association edge between nodes, and the weight of the association edge is the association degree R ij , define the relationships between smart lock devices.

[0044] Optionally, the graph neural network (GNN) model includes:

[0045] Feature initialization of smart lock device node: Each smart lock device v i As a node, initialize the node's feature vector The feature vector is determined by the physical position P of the device i 、Usage frequency U i 、Permission allocation A i and operation history H i composition;

[0046] Adjacency matrix and graph structure construction: Based on the constructed association graph G = (V, E) between smart lock devices, the adjacency matrix A represents the relationship between nodes in the association graph, and the matrix elements are the association weights between devices;

[0047] Message passing and feature updating: In the process of message passing, a weight adjustment mechanism related to device status is introduced. For each node v i , receive the information of its neighbor nodes at each layer l and update the feature vector of the node;

[0048] Node feature aggregation and update: Each node updates its own feature vector at each layer by interacting with neighboring nodes and self-loop information;

[0049] Output and optimization calculation of the final feature vector: After L-layer message transmission and aggregation, the final feature vector of each device node is output The ultimate optimization goal is to globally optimize the operation and authority allocation of the entire device network. i , output its operation strategy S through the fully connected layer i and authority assignment A i .

[0050] Optionally, the policy adjustment includes:

[0051] Dynamic permission allocation adjustment: When the usage frequency of the smart lock device is Ui The frequency is higher than the set threshold, and the operation authority of the smart lock device is increased. ij When the correlation degree is higher than the set threshold, the operator will be automatically promoted to the equipment v j authority;

[0052] Operation priority adjustment: if the urgency of the task of the smart lock device is higher than the preset emergency threshold, the operation priority of the smart lock device is adjusted in real time;

[0053] Resource allocation adjustment: current device load factor L i When the load factor L of the device is higher than the preset load threshold, the network bandwidth and power resource allocation will be automatically increased. i If the load falls below a preset threshold, the allocated resource allocation is reduced;

[0054] Load balancing adjustment: through the load status L of the device i and equipment capacity C i If the load ratio of the equipment is L i / C i ≥1, some tasks will be transferred to the adjacent low-load device v j ;

[0055] Emergency response adjustment: When equipment failure or emergency tasks are detected, the operation priority of emergency smart lock devices is increased.

[0056] Optionally, the data recording and analysis module includes:

[0057] Real-time operation log: Real-time log information of each smart lock operation, including authentication information, operation time, permission allocation, lock status and environmental data, including:

[0058] The identity verification information is the identity verification data of the operator, including the verification results of fingerprint, face, iris or voice recognition;

[0059] The operation time is the timestamp of the operation execution, including the time of unlocking and locking operations;

[0060] The authority allocation refers to the authority level of the current operator when performing an operation, including basic authority, intermediate authority or advanced authority;

[0061] The lock status is the current status of the smart lock device, whether the lock is unlocked or locked;

[0062] The environmental data is environmental data related to the operation, including temperature and humidity;

[0063] Log analysis and anomaly detection: Real-time analysis of operation logs, using behavioral pattern detection algorithms to analyze the matching of operation time, identity authentication, and permission allocation, and detect abnormal behaviors, including abnormal operation time (such as operations during non-working hours), abnormal permission allocation (such as mismatch between operation permissions and current task permissions), and abnormal environmental data (such as abnormal temperature, which may indicate abnormal equipment or environment);

[0064] Permission adjustment: When abnormal operation behavior is detected, the operator's permissions are automatically adjusted, including reducing operation permissions or locking sensitive operation permissions, and triggering security responses based on the degree of abnormality (such as alarms or notifying administrators to intervene).

[0065] Beneficial effects of the present invention:

[0066] The present invention, by combining a smart lock module, an identity authentication module, a permission management module, a multi-device collaborative optimization module and a data recording and analysis module, can realize remote control, real-time permission management and dynamic collaborative optimization of the smart locks of power facilities. The smart lock module is securely physically connected to the locks of the power facilities through a prefabricated installation interface, and is combined with a motor drive device and a status sensor to provide real-time feedback on the status information of the locks, and realize remote control through wireless communication, which greatly improves the degree of automation of equipment management and the convenience of operation.

[0067] The present invention, through multimodal identity authentication, combines the fusion verification of multiple biometric features such as fingerprint, face, iris and voice recognition, effectively enhances the security and accuracy of identity authentication, extracts and verifies comprehensive features through deep fusion algorithm, avoids the limitations of a single biometric recognition method, ensures the efficient operation and accurate verification of the system in complex environments, reduces the risk of misjudgment and security vulnerabilities, and provides safety protection for the operation of the entire power facility.

[0068] The present invention globally optimizes the operation and permission allocation of smart lock devices through a graph neural network model. The system can dynamically adjust the collaboration strategy and permission allocation between devices based on multi-dimensional information such as device usage frequency, physical location, and operation history. Coupled with a real-time data recording and analysis module, the system can detect abnormal behavior in a timely manner and automatically adjust operating permissions to ensure efficient collaboration of devices and the security, flexibility, and operational efficiency of the system, effectively responding to complex power facility management needs. BRIEF DESCRIPTION OF THE DRAWINGS

[0069] In order to more clearly illustrate the technical solutions in the present invention or the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings in the following description are only for the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0070] Figure 1 A schematic diagram of system function modules according to an embodiment of the present invention;

[0071] Figure 2 Schematic diagram of a multi-device collaborative optimization module according to an embodiment of the present invention. DETAILED DESCRIPTION

[0072] The present invention is described in detail below in conjunction with the accompanying drawings and specific embodiments. At the same time, it is explained here that in order to make the embodiments more detailed, the following embodiments are the best and preferred embodiments, and those skilled in the art may also adopt other alternatives to implement some known technologies; and the accompanying drawings are only for more specific description of the embodiments, and are not intended to specifically limit the present invention.

[0073] It should be noted that the references to "one embodiment", "an embodiment", "an exemplary embodiment", "some embodiments" and the like in the specification indicate that the embodiments described may include specific features, structures or characteristics, but not every embodiment may include the specific features, structures or characteristics. In addition, when a specific feature, structure or characteristic is described in conjunction with an embodiment, it should be within the knowledge of a person skilled in the art to implement such feature, structure or characteristic in conjunction with other embodiments (whether or not explicitly described).

[0074] In general, a term can be understood, at least in part, from its use in context. For example, depending, at least in part, on the context, the term "one or more" as used herein can be used to describe any feature, structure, or characteristic in the singular sense, or can be used to describe a combination of features, structures, or characteristics in the plural sense. Additionally, the term "based on" can be understood as not necessarily intended to convey an exclusive set of factors, but can instead, depending, at least in part, on the context, allow for the presence of other factors that are not necessarily explicitly described.

[0075] like Figure 1-Figure 2 As shown, the remote control and authority management system for smart locks of electric power facilities includes a smart lock module, an identity authentication module, an authority management module, a multi-device collaborative optimization module, and a data recording and analysis module, wherein;

[0076] The smart lock module is physically connected to the lock of the power facility through a prefabricated installation interface, performs locking, unlocking and status feedback operations, and communicates with the remote server through the network;

[0077] The identity verification module performs multi-modal identity verification on operators, combining multiple biometric features for comprehensive verification to ensure the security and accuracy of identity verification;

[0078] The authority management module dynamically allocates the authority level of operators based on the results of comprehensive verification. Through multi-level authority management, personnel with different authority levels can operate power facilities within their authorized scope.

[0079] The multi-device collaborative optimization module collaboratively optimizes the operation and authority allocation of multiple smart lock devices in power facilities. By analyzing the association between smart lock devices, the operation and resource allocation between devices are optimized, including:

[0080] Smart lock device feature extraction: Extract key features from each smart lock device, including the physical location of the device, usage frequency, permission allocation, and operation history, analyze the correlation between smart lock devices, and build a correlation graph between smart lock devices to define the relationship between smart lock devices;

[0081] Collaborative optimization calculation: Based on the association graph between smart lock devices, the graph neural network (GNN) model is used to optimize the operation and permission allocation of multiple smart lock devices;

[0082] Strategy adjustment: Based on the optimization results of the operation and authority allocation of multiple smart lock devices, dynamically adjust the coordination strategy between devices, including dynamic authority allocation adjustment, operation priority adjustment, resource allocation adjustment, load balancing adjustment and emergency response adjustment;

[0083] The data recording and analysis module records the operation log of the smart lock in real time, including authentication information, operation time, authority allocation, lock status and environmental data, analyzes the operation log and detects abnormal behavior. When an abnormal situation is detected, the authority is automatically adjusted;

[0084] Through the above content, efficient collaboration and resource allocation between devices are ensured. At the same time, combined with real-time strategy adjustment and data recording and analysis, the security, flexibility and operational efficiency of the system are improved. It can not only effectively respond to the complex management needs of power facilities, but also ensure the reliability and adaptability of equipment in different operating environments through real-time monitoring and dynamic optimization.

[0085] The smart lock module includes:

[0086] Fixed installation of the smart lock module: Physically connect the smart lock to the physical lock of the power facility through a prefabricated installation interface to ensure the secure connection between the smart lock and the power facility;

[0087] Execute locking and unlocking operations: The smart lock module is integrated with a motor drive device to perform locking and unlocking operations, and control the opening and closing of the lock according to the instructions of the authority management module;

[0088] State feedback detection: The smart lock module is equipped with a state sensor to detect the current state of the lock (locked or unlocked) in real time and generate state feedback information;

[0089] Network communication and remote control: Establish a communication connection with the remote server through a wireless communication unit (Wi-Fi, cellular network or dedicated communication protocol), upload the lock status feedback information to the server in real time, and receive operation instructions from the remote server to realize remote control and status monitoring functions;

[0090] Through the above contents, the safe integration with the power facilities is ensured, and a built-in motor drive device is used to automatically perform locking and unlocking operations, which improves the degree of automation of the operation. The module is equipped with a status sensor, which can detect the lock status in real time and provide accurate status feedback information, ensuring the real-time monitoring capability of the system. In addition, through the wireless communication unit, the smart lock module can interact with the remote server for data to achieve remote control and monitoring, thereby improving the system's management efficiency and ability to respond to emergencies, and greatly improving the safety of power facilities, ease of operation and intelligent management level.

[0091] The authentication modules include:

[0092] Biometrics collection and preprocessing: responsible for collecting various biometrics of operators, including fingerprint images, facial and iris images, and voice data, and preprocessing the collected biometrics, including noise filtering and formatting;

[0093] Multimodal feature fusion and verification: Use deep fusion algorithm to fuse pre-processed biometric features, and authenticate the fused biometric features through pre-stored template data to generate comprehensive verification results;

[0094] Through the above content, the security and accuracy of identity authentication are improved. Comprehensive verification is carried out by combining multiple biometric features such as fingerprint, face, iris and voice, which effectively avoids misjudgment or security loopholes caused by a single biometric recognition method. The preprocessing process ensures data quality, and the deep fusion algorithm improves the robustness and anti-interference of identity authentication, so that the system can still accurately verify the identity of the operator in a complex environment, thereby improving the security and reliability of the overall system.

[0095] Biometric collection and preprocessing include:

[0096] Biometric collection: fingerprint images are acquired through fingerprint sensors, facial and iris images are acquired through cameras, and voice data is acquired through microphones;

[0097] Noise filtering: Gaussian filters are used to filter noise from the collected fingerprint images, facial and iris images, and bandpass filters are used to reduce noise from the voice data, which can be expressed as:

[0098]

[0099] Where I(x,y) is the original image pixel value, I'(x,y) is the pixel value after filtering, and σ is the standard deviation of the filter kernel;

[0100]

[0101] Where x(t) is the original speech signal, h(t) is the impulse response of the bandpass filter, and y(t) is the filtered speech signal;

[0102] Data formatting: The fingerprint image, face and iris image after noise filtering are scaled by bilinear interpolation, and the signal amplitude of the voice data after noise reduction is adjusted to a fixed range by normalization, which is expressed as:

[0103] I"(x,y)=(1-a)(1-b)I'(x,y)+a(1-b)I'(x+1,y)+(1-a)bI'(x,y+1)+abI'(x+1,y+1);

[0104] Where I'(x,y) is the image pixel after noise filtering, I"(x,y) is the pixel of the image after scaling, and a and b are interpolation factors;

[0105]

[0106] Among them, x(t) is the original speech signal, and x'(t) is the normalized signal;

[0107] Through the above content, the accuracy of the data is ensured. The pre-processed data goes through a formatting step to ensure the standardization of the biometric data for subsequent multimodal feature fusion and verification. The application of these processing algorithms improves data quality, reduces interference from external noise, and ensures the efficiency and accuracy of the identity authentication process.

[0108] Multimodal feature fusion and verification include:

[0109] Feature extraction: For each biometric feature (fingerprint, face, iris, voice), a convolutional neural network (CNN) is used to extract features and generate a feature vector, which is expressed as:

[0110] (1) Convolution operation: Each input biometric feature X i (Fingerprint image, face and iris image, voice data) pass through multiple convolutional layers to obtain feature maps, expressed as:

[0111] F i =ReLU(W i *X i +b i );

[0112] Among them, X i is the input biometric image or data, W i is the convolution kernel (weight matrix), which represents the parameters of the convolution layer, b i is the bias term, ReLU(x)=max(0,x) is the activation function, F i It is the feature map after the convolution operation, which contains the local pattern of the biological feature;

[0113] (2) Feature dimensionality reduction (Pooling operation): Reduce the dimensionality of the feature map and use maximum pooling to retain important features and reduce the amount of data, which can be expressed as:

[0114]

[0115] in, is the feature vector after pooling;

[0116] (3) Feature vector generation: After multiple convolutional layers and pooling layers, the feature map is flattened into a feature vector, expressed as:

[0117]

[0118] Among them, F fingerprint , F face , F iris , F voice They are fingerprint feature vector, facial feature vector, iris feature vector, and voice feature vector respectively;

[0119] Feature fusion: After feature extraction is completed, the extracted feature vectors are fused using the fully connected layer to generate a comprehensive feature vector F fused , expressed as:

[0120] (1) Feature concatenation: The feature vectors of multiple biometric features are concatenated and expressed as:

[0121] F concat =[F fingerprint , F face , F iris , F voice ];

[0122] Among them, F concat It is the concatenated feature vector, which combines the feature vectors of fingerprint, face, iris and voice;

[0123] (2) Fully connected layer operation: Through the fully connected layer, the concatenated feature vectors are mapped to a low-dimensional space to generate a comprehensive feature vector F fused , expressed as:

[0124] Ffused =W FC ·F concat +b FC ;

[0125] Among them, W FC is the weight matrix of the fully connected layer, b FC is the bias term of the fully connected layer, F fused It is the integrated feature vector after fusion, which contains the comprehensive information of multiple biometric features;

[0126] Identity verification: The fused comprehensive feature vector F fused Match with the pre-stored template data and determine the identity match by calculating the cosine similarity, expressed as:

[0127]

[0128] Among them, F template is the template feature vector stored in the database, Sim(F fused ,F template ) is the similarity between the fusion feature and the template feature;

[0129] Comprehensive verification result generation: Based on the cosine similarity result, determine whether the matching degree reaches the set verification threshold. If the similarity exceeds the set verification threshold, the verification is passed, a comprehensive verification result is generated, and a verification pass signal is sent to the authority management module;

[0130] The verification threshold is determined based on the similarity distribution of the operator's verification results, including:

[0131] Data collection and similarity calculation: collect verification data of a group of legitimate users and data of a group of illegitimate users, and calculate their respective similarity distributions;

[0132] Define FAR and FRR: Calculate the false acceptance rate FAR and the false rejection rate FRR;

[0133] Threshold setting: By analyzing the similarity distribution of legitimate users and illegitimate users, the optimal threshold that balances FAR and FRR is found, which is expressed as:

[0134]

[0135] Where T is the verification threshold, FAR(T) and FRR(T) are the false acceptance rate and false rejection rate respectively as the threshold changes;

[0136] Through the above content, the security and accuracy of identity authentication can be improved. The comprehensive vector of fusion of various features can effectively make up for the limitations of single biometric features and reduce the occurrence of misjudgment. The convolutional neural network is used to extract high-dimensional features and generate a unified feature vector through a deep fusion algorithm. The system has stronger anti-interference ability and can perform accurate identity recognition in complex environments, thereby improving the robustness and reliability of the system.

[0137] The rights management module includes:

[0138] Permission level setting: pre-define multiple permission levels, each of which corresponds to different operation scopes and access rights, including basic operation permissions, intermediate maintenance permissions, and advanced management permissions, to ensure that operators can operate within their authorized scope;

[0139] Dynamic authority allocation: Compare the comprehensive verification results output by the identity authentication module with the preset upper and lower limits of the credibility threshold, and divide the identity credibility of the operator into basic credibility (the comprehensive verification result is lower than the lower limit of the credibility threshold), intermediate credibility (the comprehensive verification result is between the lower and upper limits of the credibility threshold), and advanced credibility (the comprehensive verification result is higher than the upper limit of the credibility threshold). According to the credibility level, the corresponding authority level is automatically allocated, including basic operation authority for basic credibility, intermediate maintenance authority for intermediate credibility, and advanced management authority for advanced credibility;

[0140] Multi-level authority management: Through the multi-level authority management mechanism, based on the results of dynamic authority allocation, operators with different authority levels can operate and control power facilities within their authorized scope to prevent authority abuse or over-authority operation. The system will match the corresponding authority level according to the sensitivity and complexity of each operation task;

[0141] The upper and lower limits of the credibility threshold are set based on the statistical distribution of the identity authentication results, including:

[0142] Data collection and analysis: Collect multimodal identity authentication results of a group of operators, including comprehensive verification scores (cosine similarity scores) of fingerprint, face, iris and voice features, and perform statistical analysis on the verification results of legitimate users to obtain the distribution of verification scores P(x);

[0143] Set the lower limit of the credibility threshold: According to the distribution of legitimate user verification scores, select the 5% quantile as the lower limit of the credibility threshold T low , expressed as:

[0144] T low =Percentile(P(x),5%);

[0145] Set the upper limit of the credibility threshold: According to the distribution of legitimate user verification scores, select the 95th percentile as the upper limit of the credibility threshold T high , expressed as:

[0146] T high =Percentile(P(x),95%);

[0147] Through the above content, the system ensures that a balance is achieved between security and flexibility. By setting the credibility threshold, the credibility of operators can be accurately divided to ensure that highly credible personnel can perform advanced operations while preventing low-credible personnel from performing operations beyond their authority. In addition, the system can flexibly adjust permissions based on the complexity and sensitivity of the task, so that operators at different levels can only perform tasks within the scope of their authorization, thereby preventing abuse of authority and improving the security and operational efficiency of the system.

[0148] Smart lock device feature extraction includes:

[0149] Feature data collection: Collect key feature data from each smart lock device, including the physical location P i 、Usage frequency U i 、Permission allocation A i and operation history H i ,in:

[0150] Physical location P i Determined by GPS or network topology information;

[0151] Frequency of use i Calculated by the number of times or time intervals of equipment operation;

[0152] Permission AssignmentA i Set up collection based on current device permissions;

[0153] Operation History i Includes past operation records, error logs, and maintenance records;

[0154] Correlation analysis: Analyze the characteristic data of each smart lock device, calculate the correlation between smart lock devices, and calculate the correlation R between smart lock device i and device j through multi-factor comprehensive analysis. ij , expressed as:

[0155] R ij =α1·sim(P i , P j )+α2·sim(U i , U j )+·α3·sim(A i ,A j)+α4·sim(H i , H j );

[0156] Among them, sim(P i , P j ) represents the physical location similarity between devices i and j, which is calculated by Euclidean distance, i.e. sim(U i , U j ) represents the similarity of usage frequency, which is calculated using the normalized frequency value difference, i.e. sim(A i , A j ) and sim(H i , H j ) respectively represent the similarity of permission allocation and operation history, which are calculated by the matching degree of classification data;

[0157] Association graph construction: based on the calculated association R ij , construct the association graph G = (V, E) between smart lock devices, where V represents the smart lock device node, E represents the association edge between nodes, and the weight of the association edge is the association degree R ij ,define the relationship between smart lock devices;

[0158] Through the above content, we can accurately analyze the correlation between smart lock devices and then build a correlation graph between devices. This multi-factor analysis and graph model construction not only improves the efficiency of collaborative work between devices, but also provides data support for authority management and resource allocation, ensuring that the system can intelligently and dynamically optimize the collaboration and operation of multiple devices, significantly improving the intelligence level of power facility management and overall operating efficiency.

[0159] The graph neural network (GNN) model includes:

[0160] Feature initialization of smart lock device node: Each smart lock device v i As a node, initialize the node's feature vector The feature vector is determined by the physical position P of the device i 、Usage frequency U i 、Permission allocation A i and operation history H i Composition, expressed as:

[0161]

[0162] Adjacency matrix and graph structure construction: Based on the constructed association graph G = (V, E) between smart lock devices, the adjacency matrix A represents the relationship between nodes in the association graph, and the matrix elements are the association weights between devices, which are expressed as:

[0163]

[0164] Among them, R ij For smart lock devices i and v j The association weight between them;

[0165] Message transmission and feature update: In order to adapt to the optimization of collaborative operation between smart lock devices, a weight adjustment mechanism related to device status is introduced in the message transmission process to ensure that devices with frequent operations have greater influence. i , receiving the information of its neighbor nodes at each layer l, updating the feature vector of the node, and the message passing is expressed as:

[0166]

[0167] in, is the node v in the l+1th layer i The message vector, is node v i The neighbor set of A ij is the weight in the adjacency matrix, d i and d j They are node v i and neighbor node v j Degree, W (l) is the trainable weight matrix of the lth layer, β is the adjustment coefficient related to the frequency of device use, so that devices with higher frequency of use have greater influence in the optimization, is the weight matrix of the self-loop edge (used to retain the device’s own information), Represents node v j The feature vector at layer l, Represents node v i Feature vector at layer l;

[0168] Node feature aggregation and update: Each node updates its own feature vector at each layer by interacting with neighboring nodes and self-loop information. In each layer l+1, the update of node features is expressed as:

[0169]

[0170] Where σ is the ReLU activation function, is the weight matrix for node self-loop information;

[0171] Output and optimization calculation of the final feature vector: After L-layer message transmission and aggregation, the final feature vector of each device node is output The ultimate optimization goal is to globally optimize the operation and authority allocation of the entire device network. i , output its operation strategy S through the fully connected layer i and authority assignment A i , expressed as:

[0172]

[0173]

[0174] Among them, S i Indicates device v i The operation strategy of A i Indicates device v i The permission allocation strategy, and are the weight matrices corresponding to the output, b (S) and b (A) They are the bias terms of the corresponding outputs;

[0175] Through the above content, the overall efficiency and accuracy of the system can be improved. By initializing multi-dimensional features such as device usage frequency, physical location, permission allocation and operation history, the correlation between devices can be more comprehensively understood. At the same time, the message passing mechanism introduces the weight of device usage frequency to ensure that frequently used devices have greater influence in the optimization process, thereby improving the rationality of the system's operation scheduling and resource allocation. After multi-layer feature updates and aggregation, it can intelligently generate optimized operation strategies and permission allocations for each device, ensuring efficient equipment collaboration and dynamic permission management, thereby improving the security and flexibility of the entire power facility management system.

[0176] Strategy adjustments include:

[0177] Dynamic permission allocation adjustment: When the usage frequency of the smart lock device is U i The frequency is higher than the set threshold, and the operation authority of the smart lock device is increased. ij When the correlation degree is higher than the set threshold, the operator will be automatically promoted to the equipment v j authority;

[0178] Operation priority adjustment: if the urgency of the task of the smart lock device is higher than the preset emergency threshold, the operation priority of the smart lock device is adjusted in real time;

[0179] Resource allocation adjustment: current device load factor L i When the load factor L of the device is higher than the preset load threshold, the network bandwidth and power resource allocation will be automatically increased. i If the load falls below a preset threshold, the allocated resource allocation is reduced;

[0180] Load balancing adjustment: through the load status L of the device i and equipment capacity C i If the load ratio of the equipment is L i / C i ≥1, some tasks will be transferred to the adjacent low-load device v j , ensuring full utilization of system resources;

[0181] Emergency response adjustment: when equipment failure or emergency tasks are detected, the operation priority of emergency smart lock devices is increased;

[0182] The method for setting the frequency threshold specifically includes:

[0183] Data collection and analysis: By analyzing the usage frequency of each smart lock device i Collect data and collect statistics on the frequency of use of equipment;

[0184] Dynamic setting: The frequency threshold is set according to the quantile in the statistical results. The 90% quantile of the frequency distribution is used as the frequency threshold. For devices with higher usage frequency, the system will increase the operator's authority, which is expressed as:

[0185] U threshold =Percentile(U i ,90%);

[0186] Among them, U threshold is the frequency threshold, U i The usage frequency of smart lock device i, which indicates the number of times or duration of use of the device within a certain period of time. By setting it to the 90th percentile, it ensures that frequently used devices are given higher authority management;

[0187] The method for setting the correlation threshold specifically includes:

[0188] Device association calculation: Calculate the device association R ij ;

[0189] Threshold setting: By analyzing the correlation R between all devices ij , the 75% quantile of the association distribution is selected as the threshold to ensure that devices with close physical locations and close collaboration can get privilege escalation, which is expressed as:

[0190] R threshold =Percentile(R ij , 75%);

[0191] Among them, R threshold is the correlation threshold, R ij For device vi and device v j The degree of correlation between

[0192] The method for setting the emergency threshold specifically includes:

[0193] Urgent task definition: The urgency of the task priority According to the importance of the task, the execution time limit or the scope of the failure impact, the urgent task needs to be completed in a short time or has systemic risks;

[0194] Threshold setting: By analyzing the historical task data, we set the quantiles of task priority and select the 85% quantile as the threshold for urgent tasks, which means that only tasks with a priority higher than this value will be considered urgent tasks and will be executed by the system first, as shown in the following figure:

[0195] T threshold =Percentile(T priority , 85%);

[0196] Among them, T threshold is the emergency threshold, T priority Indicates the priority of the task;

[0197] The method for setting the load threshold specifically includes:

[0198] Equipment load factor calculation: Calculate the load factor L of the equipment i ;

[0199] Dynamic load threshold: The load threshold is set based on the distribution of the system's historical load data. The 80% percentile of the device load factor is selected as the load threshold. When the load exceeds this value, the system will start load balancing or resource reallocation, which is expressed as:

[0200] L threshold =Percentile(L i , 80%);

[0201] Among them, L threshold is the load threshold, L i is the equipment load factor;

[0202] Through the above content, key strategies such as authority allocation, resource allocation, and operation priority can be intelligently adjusted in different situations, ensuring that equipment with high frequency of use or emergency tasks can receive priority processing and resource support, while effectively preventing system overload and ensuring load balancing. By setting reasonable thresholds based on historical data and actual operating conditions, the system not only improves the overall operating efficiency, but also enhances emergency response capabilities and safety, ensuring efficient, stable and reliable operation of power facilities.

[0203] The data recording and analysis module includes:

[0204] Real-time operation log: Real-time log information of each smart lock operation, including authentication information, operation time, permission allocation, lock status and environmental data, including:

[0205] The identity verification information is the operator’s identity verification data, including the verification results of fingerprint, face, iris or voice recognition;

[0206] The operation time is the timestamp of the operation execution, including the time of unlocking and locking operations;

[0207] The permission allocation refers to the permission level of the current operator when performing the operation, including basic permission, intermediate permission or advanced permission;

[0208] The lock status is the current status of the smart lock device, whether the lock is unlocked or locked;

[0209] Environmental data refers to environmental data related to operations, including temperature and humidity, to ensure detection of environmental changes;

[0210] Log analysis and anomaly detection: Real-time analysis of operation logs, using behavioral pattern detection algorithms to analyze the matching of operation time, identity authentication, and permission allocation, and detect abnormal behaviors, including abnormal operation time (such as operations during non-working hours), abnormal permission allocation (such as mismatch between operation permissions and current task permissions), and abnormal environmental data (such as abnormal temperature, which may indicate abnormal equipment or environment);

[0211] Permission adjustment: When abnormal operation behavior is detected, the operator's permissions are automatically adjusted, including reducing operation permissions or locking sensitive operation permissions, and triggering security responses (such as alarms or notifying administrators to intervene) based on the degree of abnormality;

[0212] The behavioral pattern detection algorithm specifically includes:

[0213] Authentication and permission allocation matching test: record the authentication information of the operator (such as authentication success rate, identity type, etc.) and the current permission allocation information, and determine whether there is abnormal behavior by calculating the matching degree between the authentication result and the permission allocation, which is expressed as:

[0214]

[0215] Among them, I i Indicates the authentication information of operator i, A i Indicates the authority allocation information of operator i, I score (i) is the authentication score of operator i (such as the authentication success rate of fingerprint or iris recognition, ranging from 0 to 1), A level(i) is the permission level assigned to operator i (e.g., 1 for basic permission, 2 for intermediate permission, and 3 for advanced permission). i ,A i ) is lower than the preset threshold (0.5), indicating that the authentication result does not match the current permission allocation. This is considered an abnormal operation and measures are taken;

[0216] Operation time anomaly detection: Match the operation time with the preset time range (such as working time period) to detect whether there is any operation in an abnormal time period, expressed as:

[0217] T deviation =|T operation -T expected |;

[0218] Among them, T operation is the time when the operation actually occurs, T expected is the preset normal operation time (such as the median of the working time period). deviation If the deviation is greater than the preset threshold (1 hour), it is considered an abnormal operation;

[0219] Permission allocation anomaly detection: Check whether the currently assigned permissions match the task requirements. For example, some high-privilege tasks can only be performed by specific personnel. The system detects anomalies by the matching degree between permission levels and task requirements, expressed as:

[0220]

[0221] Among them, A level (i) is the operator’s authority level, T priority Is the priority of the task, representing the permission level required for the task. If the permission match degree Permission_Match(A i ,T requirement ) is less than 1, indicating that the operator's authority is lower than the task requirement, and the system will detect an abnormality;

[0222] Environmental data anomaly detection: Environmental data (temperature, humidity) exceeding the preset range indicates a physical environment anomaly or equipment failure, which is expressed as:

[0223] E deviation =|E current -E normal |;

[0224] E current is the current environmental parameters (temperature, humidity), E normal is the median value of the environmental parameter range under normal conditions. deviationIf the set threshold is exceeded (temperature changes by more than 10°C or humidity changes by more than 15%), it will be considered an environmental abnormality;

[0225] Through the above content, abnormal behavior can be detected quickly, such as operation time exceeding the normal range, authority allocation not matching operation or abnormal environmental conditions. When an abnormality occurs, the operation authority can be automatically adjusted to prevent potential security risks. This not only improves the security of the system, but also enhances the ability to respond to abnormal situations, ensuring that the smart lock system of power facilities operates stably and efficiently.

[0226] The present invention covers any substitution, modification, equivalent method and scheme made on the essence and scope of the present invention. In order to make the public have a thorough understanding of the present invention, specific details are described in detail in the following preferred embodiments of the present invention, but those skilled in the art can fully understand the present invention without the description of these details. In addition, in order to avoid unnecessary confusion about the essence of the present invention, well-known methods, processes, procedures, components and circuits are not described in detail.

[0227] The above is only a preferred embodiment of the present invention. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the principle of the present invention. These improvements and modifications should also be regarded as the scope of protection of the present invention.

Claims

1. Remote control and authority management system for smart locks of electric power facilities, characterized in that: It includes smart lock module, identity authentication module, permission management module, multi-device collaborative optimization module and data recording and analysis module, among which; The smart lock module is physically connected to the lock of the power facility through a prefabricated installation interface, performs locking, unlocking and status feedback operations, and communicates with the remote server through the network; The identity verification module performs multi-modal identity verification on the operator, combining multiple biometric features for comprehensive verification; The authority management module dynamically allocates the authority level of operators according to the results of the comprehensive verification, and through multi-level authority management, enables personnel with different authority levels to operate power facilities within their authorized scope; The multi-device collaborative optimization module collaboratively optimizes the operation and authority allocation of multiple smart lock devices in the power facility, and optimizes the operation and resource allocation between devices by analyzing the association relationship between smart lock devices, specifically including: Smart lock device feature extraction: Extract key features from each smart lock device, including the physical location of the device, usage frequency, permission allocation, and operation history, analyze the correlation between smart lock devices, and build a correlation graph between smart lock devices to define the relationship between smart lock devices; Collaborative optimization computing: Based on the association graph between smart lock devices, the graph neural network model is used to optimize the operation and permission allocation of multiple smart lock devices; Strategy adjustment: Based on the optimization results of the operation and authority allocation of multiple smart lock devices, dynamically adjust the coordination strategy between devices, including dynamic authority allocation adjustment, operation priority adjustment, resource allocation adjustment, load balancing adjustment and emergency response adjustment; The data recording and analysis module records the operation log of the smart lock in real time, including identity authentication information, operation time, authority allocation, lock status and environmental data, analyzes the operation log and detects abnormal behavior, and automatically adjusts the authority when an abnormal situation is detected.

2. The remote control and authority management system for smart locks of electric power facilities according to claim 1 is characterized in that: The smart lock module includes: Fixed installation of smart lock module: physical connection with the physical lock of power facilities through prefabricated installation interface; Execute locking and unlocking operations: The smart lock module is integrated with a motor drive device to perform locking and unlocking operations, and control the opening and closing of the lock according to the instructions of the authority management module; Status feedback detection: The smart lock module is equipped with a status sensor to detect the current status of the lock in real time and generate status feedback information; Network communication and remote control: Establish a communication connection with the remote server through the wireless communication unit, upload the lock status feedback information to the server in real time, and receive operation instructions from the remote server to realize remote control and status monitoring functions.

3. The remote control and authority management system for smart locks of electric power facilities according to claim 1 is characterized in that: The identity verification module comprises: Biometrics collection and preprocessing: responsible for collecting various biometrics of operators, including fingerprint images, facial and iris images, and voice data, and preprocessing the collected biometrics, including noise filtering and formatting; Multimodal feature fusion and verification: A deep fusion algorithm is used to fuse the pre-processed biometric features, and the fused biometric features are authenticated through pre-stored template data to generate a comprehensive verification result.

4. The remote control and authority management system for smart locks of electric power facilities according to claim 3 is characterized in that: The biometric feature collection and preprocessing includes: Biometric collection: fingerprint images are acquired through fingerprint sensors, facial and iris images are acquired through cameras, and voice data is acquired through microphones; Noise filtering: Use Gaussian filters to filter noise from the collected fingerprint images, facial and iris images, and use bandpass filters to reduce noise from voice data; Data formatting: The fingerprint image, face and iris image after noise filtering are scaled by bilinear interpolation, and the signal amplitude of the voice data after noise reduction is adjusted to a fixed range by normalization.

5. The remote control and authority management system for smart locks of electric power facilities according to claim 4 is characterized in that: The multimodal feature fusion and verification includes: Feature extraction: For each biological feature, a convolutional neural network is used to extract features and generate feature vectors; Feature fusion: After feature extraction is completed, the extracted feature vectors are fused using the fully connected layer to generate a comprehensive feature vector F fused ; Identity verification: The fused comprehensive feature vector F fused Matching with pre-stored template data, and determining identity matching by calculating cosine similarity; Generation of comprehensive verification results: Based on the cosine similarity results, determine whether the matching degree reaches the set verification threshold. If the similarity exceeds the set verification threshold, the verification is passed, a comprehensive verification result is generated, and a verification pass signal is sent to the authority management module.

6. The remote control and authority management system for smart locks of electric power facilities according to claim 1, characterized in that: The rights management module includes: Permission level setting: pre-define multiple permission levels, each of which corresponds to different operation scopes and access rights, including basic operation permissions, intermediate maintenance permissions, and advanced management permissions; Dynamic authority allocation: Compare the comprehensive verification results output by the identity authentication module with the preset upper and lower limits of the credibility threshold, divide the operator's identity credibility into basic credibility, intermediate credibility, and advanced credibility, and automatically allocate corresponding authority levels according to the credibility level, including basic operation authority for basic credibility, intermediate maintenance authority for intermediate credibility, and advanced management authority for advanced credibility; Multi-level authority management: Through the multi-level authority management mechanism, based on the results of dynamic authority allocation, operators with different authority levels can operate and control power facilities within their authorized scope.

7. The remote control and authority management system for smart locks of electric power facilities according to claim 1, characterized in that: The smart lock device feature extraction includes: Feature data collection: Collect key feature data from each smart lock device, including the physical location P i 、Usage frequency U i 、Permission allocation A i and operation history H i ,in: Physical location P i Determined by GPS or network topology information; Frequency of use i Calculated by the number of times or time intervals of equipment operation; Permission AssignmentA i Set up collection based on current device permissions; Operation History i Includes past operation records, error logs, and maintenance records; Correlation analysis: Analyze the characteristic data of each smart lock device, calculate the correlation between smart lock devices, and calculate the correlation R between smart lock device i and device j through multi-factor comprehensive analysis. ij ; Association graph construction: based on the calculated association R ij , construct the association graph G = (V, E) between smart lock devices, where V represents the smart lock device node, E represents the association edge between nodes, and the weight of the association edge is the association degree R ij , define the relationships between smart lock devices.

8. The remote control and authority management system for smart locks of electric power facilities according to claim 7, characterized in that: The graph neural network model includes: Feature initialization of smart lock device node: Each smart lock device v i As a node, initialize the node's feature vector The feature vector is determined by the physical position P of the device i 、Usage frequency U i 、Permission allocation A i and operation history H i composition; Adjacency matrix and graph structure construction: Based on the constructed association graph G = (V, E) between smart lock devices, the adjacency matrix A represents the relationship between nodes in the association graph, and the matrix elements are the association weights between devices; Message passing and feature updating: In the process of message passing, a weight adjustment mechanism related to device status is introduced. For each node v i , receive the information of its neighbor nodes at each layer l and update the feature vector of the node; Node feature aggregation and update: Each node updates its own feature vector at each layer by interacting with neighboring nodes and self-loop information; Output and optimization calculation of the final feature vector: After L-layer message transmission and aggregation, the final feature vector of each device node is output The ultimate optimization goal is to globally optimize the operation and authority allocation of the entire device network. i , output its operation strategy S through the fully connected layer i and authority assignment A i .

9. The remote control and authority management system for smart locks of electric power facilities according to claim 8, characterized in that: The strategy adjustment includes: Dynamic permission allocation adjustment: When the usage frequency of the smart lock device is U i The frequency is higher than the set threshold, and the operation authority of the smart lock device is increased. ij When the correlation degree is higher than the set threshold, the operator will be automatically promoted to the equipment v j authority; Operation priority adjustment: if the urgency of the task of the smart lock device is higher than the preset emergency threshold, the operation priority of the smart lock device is adjusted in real time; Resource allocation adjustment: current device load factor L i When the load factor L of the device is higher than the preset load threshold, the network bandwidth and power resource allocation will be automatically increased. i If the load falls below a preset threshold, the allocated resource allocation is reduced; Load balancing adjustment: through the load status L of the device i and equipment capacity C i If the load ratio of the equipment is L i / C i ≥1, some tasks will be transferred to the adjacent low-load device v j ; Emergency response adjustment: When equipment failure or emergency tasks are detected, the operation priority of emergency smart lock devices is increased.

10. The remote control and authority management system for smart locks of electric power facilities according to claim 1, characterized in that: The data recording and analysis module includes: Real-time operation log: Real-time log information of each smart lock operation, including authentication information, operation time, permission allocation, lock status and environmental data, including: The identity verification information is the identity verification data of the operator, including the verification results of fingerprint, face, iris or voice recognition; The operation time is the timestamp of the operation execution, including the time of unlocking and locking operations; The authority allocation refers to the authority level of the current operator when performing an operation, including basic authority, intermediate authority or advanced authority; The lock status is the current status of the smart lock device, whether the lock is unlocked or locked; The environmental data is environmental data related to the operation, including temperature and humidity; Log analysis and anomaly detection: Real-time analysis of operation logs, using behavioral pattern detection algorithms to analyze the matching of operation time, identity authentication, and authority allocation, and detect abnormal behaviors, including abnormal operation time, abnormal authority allocation, and abnormal environmental data; Permission adjustment: When abnormal operation behavior is detected, the operator's permissions are automatically adjusted, including reducing operation permissions or locking sensitive operation permissions, and triggering security responses based on the degree of abnormality.

Images