Bluetooth lock system based on transformer substation

Abstract

The invention discloses a Bluetooth lockset system based on a transformer substation, and belongs to the technical field of transformer substation safety, remote control of a lockset in the transformer substation becomes possible by arranging a Bluetooth lockset terminal, and a mobile key terminal and a key server are arranged, so that the Bluetooth lockset terminal can be unlocked only after authentication, and the safety of the transformer substation is improved. By using the lock, a person can be accurately positioned, event tracing is more convenient, a dynamic secret key of the Bluetooth lock terminal can be realized, and the safety of the lock is greatly improved.

Description

Technical Field

[0001] The invention belongs to the technical field of substation safety, and in particular relates to a Bluetooth lock system based on a substation. Background Art

[0002] Substation is an important power facility. Its main function is to receive high-voltage electricity, convert it into low-voltage electricity suitable for users through transformers, and distribute and control it. With the rapid development of my country's power industry, the number and scale of substations are constantly expanding. In order to ensure the safe operation of substations, it is necessary to regularly inspect and maintain the equipment in substations. However, the existing substation locks have the following problems: the locks are not safe and can be easily damaged by malicious means. Summary of the invention

[0003] The invention provides a Bluetooth lock system based on a substation, which is used to solve the technical problem of insufficient security in the existing substation.

[0004] A substation-based Bluetooth lock system, comprising: a Bluetooth lock terminal, a mobile key terminal, and a key server; The mobile key terminal is used to obtain a facial image input by a staff member and transmit the facial image to a key server; wherein each mobile key terminal corresponds to a staff member one by one; The key server is used to obtain the face image transmitted by the mobile key terminal, and recognize the face image through the pre-deployed identity authentication model to determine the face recognition result; wherein the face recognition result includes identity recognition passing or identity recognition failing; The key server is further configured to transmit unlocking information to the mobile key terminal when the face recognition result is that the identity recognition is passed; wherein the unlocking information includes an unlocking key and a randomly generated new unlocking key; The mobile key terminal is used to transmit the unlocking key and the new unlocking key to the Bluetooth lock terminal through a Bluetooth encrypted connection after receiving the unlocking information transmitted by the key server; The Bluetooth lock terminal is used to verify whether the unlocking key is correct after receiving the unlocking key and the new unlocking key transmitted by the mobile key terminal. If so, the unlocking operation is performed and the new unlocking key is used to replace the original unlocking key. Otherwise, the unlocking is refused.

[0005] Further, obtaining a face image input by a staff member and transmitting the face image to a key server includes: When the staff requests to unlock the door, the camera on the mobile key terminal is dispatched to capture the staff's face image; After the face image is encrypted using the public key of the key server, an encrypted face image is obtained, and the encrypted face image is transmitted to the key server.

[0006] Furthermore, the face image transmitted by the mobile key terminal is taken, and the face image is recognized by a pre-deployed identity authentication model to determine the face recognition result, including: The encrypted face image transmitted by the mobile key terminal is obtained, and the key server decrypts the encrypted face image using its own private key to obtain the decrypted face image; The decrypted face image is preprocessed to obtain a preprocessed face image, and the preprocessed face image is used as input data of a pre-deployed identity authentication model to obtain output data of the identity authentication model and obtain a face recognition result.

[0007] Furthermore, the pre-deployment method of the identity authentication model includes: The neural network model is used to build the identity authentication model, and the population is initialized based on the hyperparameters of the identity authentication model; Obtaining the error function value corresponding to each individual in the population, and dividing the population into a first sub-population, a second sub-population, and a third sub-population in order from small to large according to the error function values; For each first individual in the first subpopulation, a neighborhood search based on Gaussian distribution is performed on the first individual to obtain the first individual after the search; For each second individual in the second subpopulation, a social learning search is performed on the second individual guided by the first individual and in cooperation with the second individual to obtain the second individual after the search; For each third individual in the third subpopulation, a locally guided perturbation search is performed on the third individual to obtain the third individual after the search; The first individual after the search, the second individual after the search, and the third individual after the search are merged into a population again to obtain the individual crowding factor; When the individual crowding factor is less than a preset threshold, a global mutation search is performed on the individuals in the fused population to obtain individuals after the global mutation search; Determine whether the current number of training times has reached the maximum number of training times. If so, determine the optimal individual based on the first individual after the search, the second individual after the search, and the third individual after the search, apply the hyperparameters contained in the optimal individual to the identity authentication model, and deploy the identity authentication model on the key server. Otherwise, based on the individual after the global mutation search, return to the population division step.

[0008] Further, for each first individual in the first subpopulation, a neighborhood search based on Gaussian distribution is performed on the first individual to obtain the first individual after the search, including: For each first individual in the first subpopulation, the Gaussian distribution parameter is obtained according to the error function value corresponding to the first individual:

[0009]

[0010] in, represents the Gaussian distribution parameter, represents an exponential function with the natural constant e as the base, Indicates i The fitness corresponding to the first individual is represents the fitness corresponding to the first randomly selected individual, Indicates i The error function value corresponding to the first individual, represents the preset noise term, and =0.00001; Based on the Gaussian distribution parameters, a neighborhood search is performed on the first body, and the first body after the search is obtained as:

[0011] in, Indicates t During the training i The first individual, Indicates the first body after the search, i =1,2,…,I, I represents the total number of the first individual, The mean is 0 and the standard deviation is A random number generated from a Gaussian distribution.

[0012] Further, for each second individual in the second subpopulation, a social learning search is performed on the second individual guided by the first individual and in cooperation with the second individual to obtain the second individual after the search, including: For each second individual in the second subpopulation, randomly match the second individual with a first individual and a different second individual to obtain a first target individual and a second target individual; According to the first target individual and the second target individual, a first social learning factor and a second social learning factor are generated as follows:

[0013]

[0014] in, represents the first social learning factor, represents the second social learning factor, represents the fitness corresponding to the first target individual, represents the fitness corresponding to the second target individual, Indicates j The fitness corresponding to the second individual; According to the first target individual, the second target individual, the first social learning factor and the second social learning factor, a social learning search is performed on the second individual, and the second individual after the search is obtained as:

[0015] in, Indicates t During the training j A second individual, represents the second individual after the search, j =1,2,…,J, J represents the total number of the second individuals, represents the first random number between (0,1), represents the second random number between (0,1), represents the first target individual, Represents the second target individual.

[0016] Furthermore, for each third individual in the third subpopulation, a locally guided perturbation search is performed on the third individual to obtain the third individual after the search, including: Based on the second individual in the second subpopulation, the average position corresponding to all the second individuals is obtained to obtain the third target individual; Based on the current number of training times, the adaptive inertia weight coefficient is obtained as:

[0017] in, represents the adaptive inertia weight coefficient, Indicates the preset maximum number of training times. Indicates the preset maximum value of the adaptive inertia weight coefficient, Indicates the preset minimum value of the adaptive inertia weight coefficient; According to the third target individual and the adaptive inertia weight coefficient, a disturbance search is performed on the third individual, and the third individual after the search is obtained as follows:

[0018] in, Indicates t During the training m A third individual, Indicates the third individual after the search, m =1,2,…,M, M represents the total number of the third individuals, represents the third target individual, represents a perturbation factor randomly generated from (0,2).

[0019] Furthermore, individual crowding factors are obtained, including: The first individual after the search, the second individual after the search, and the third individual after the search are merged into the population again, and the Euclidean distance between each individual and all other individuals is obtained; For any individual, the Euclidean distance between the individual and all other individuals is averaged to obtain the individual crowding factor.

[0020] Furthermore, a global mutation search is performed on the individuals in the fused population to obtain individuals after the global mutation search, including: For the fused population, randomly select N fourth target individuals and find the historical optimal positions corresponding to the N fourth target individuals; The historical optimal positions corresponding to the N fourth target individuals are averaged to obtain the fifth target individual, and the fifth target individual is mutated through the chaotic sequence as follows:

[0021] in, Indicates the fifth target individual d Dimension parameters, d =1,2,…,D, where D represents the total number of parameters in an individual. Indicates the fifth target individual after mutation d Dimension parameters, represents the mutation search radius of the d-th dimension parameter, represents the chaotic sequence generated by chaotic mapping; According to the fifth target individual after mutation, a global mutation search is performed on the individual, and the individual after the global mutation search is obtained as follows:

[0022] in, represents the individual to be mutated, Represents the individual after global mutation search , represents the historical optimal value corresponding to the individual to be mutated, represents the fifth target individual after mutation, represents the third social learning factor preset as a constant term, represents the fourth social learning factor preset as a constant term, represents the third random number between (0,1), Represents the fourth random number between (0,1).

[0023] Furthermore, when the face recognition result is that the identity recognition is passed, the unlocking information is transmitted to the mobile key terminal, including: When the face recognition result is that the identity recognition is passed, a new unlocking key is randomly generated and the unlocking key is found; After encrypting the new unlocking key and the unlocking key by the public key of the Bluetooth lock terminal, the encrypted new unlocking key and the unlocking key are obtained; Generate an unlocking instruction, and generate unlocking information according to the unlocking instruction, the encrypted new unlocking key and the unlocking key.

[0024] The present invention provides a substation-based Bluetooth lock system, which makes it possible to remotely control the locks in the substation by setting up a Bluetooth lock terminal, and is provided with a mobile key terminal and a key server. Not only does the Bluetooth lock terminal need to be authenticated before it can be unlocked, but the use of the lock can be accurately located to an individual, making it more convenient to trace events, and the dynamic key of the Bluetooth lock terminal can be realized, greatly improving the security of the lock. BRIEF DESCRIPTION OF THE DRAWINGS

[0025] The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the invention and, together with the description, serve to explain the principles of the invention.

[0026] Figure 1 A schematic structural diagram of a substation-based Bluetooth lock system provided by an embodiment of the present invention.

[0027] The above drawings have shown clear embodiments of the present invention, which will be described in more detail below. These drawings and text descriptions are not intended to limit the scope of the present invention in any way, but to illustrate the concept of the present invention for those skilled in the art by referring to specific embodiments. DETAILED DESCRIPTION

[0028] Exemplary embodiments will be described in detail herein, examples of which are shown in the accompanying drawings. When the following description refers to the drawings, the same numbers in different drawings represent the same or similar elements unless otherwise indicated. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with the present invention. Instead, they are merely examples of devices and methods consistent with some aspects of the present invention as detailed in the appended claims.

[0029] The embodiments of the present invention are described in detail below with reference to the accompanying drawings.

[0030] like Figure 1 As shown, an embodiment of the present invention provides a substation-based Bluetooth lock system, including: a Bluetooth lock terminal 101, a mobile key terminal 102, and a key server 103; The mobile key terminal 102 is used to obtain a facial image input by a staff member and transmit the facial image to the key server 103; wherein each mobile key terminal 102 corresponds to a staff member one by one; The high-definition camera provided on the mobile key terminal 102 can be used to capture the facial image of the staff member, which is then encrypted using the public key disclosed by the key server 103 before being transmitted, thereby ensuring data security.

[0031] The key server 103 is used to obtain the face image transmitted by the mobile key terminal 102, and recognize the face image through the pre-deployed identity authentication model to determine the face recognition result; wherein the face recognition result includes identity recognition passing or identity recognition failing; The pre-deployed identity authentication model may be a face recognition neural network. After the face recognition neural network is trained, it is possible to identify whether the current face image to be identified is an authorized person.

[0032] Optionally, a record may be made after each facial image recognition, and unlocking information may be recorded subsequently, thereby facilitating traceability and effectively improving operational management efficiency.

[0033] The key server 103 is further configured to transmit unlocking information to the mobile key terminal 102 when the face recognition result is that the identity recognition is passed; wherein the unlocking information includes an unlocking key and a randomly generated new unlocking key; The mobile key terminal 102 is used to transmit the unlocking key and the new unlocking key to the Bluetooth lock terminal 101 through a Bluetooth encrypted connection after receiving the unlocking information transmitted by the key server 103; In the embodiment of the present invention, various data are running in the background, and the data are transmitted and identified through encryption technology, which can effectively ensure data security.

[0034] The Bluetooth lock terminal 101 is used to verify whether the unlocking key is correct after receiving the unlocking key and the new unlocking key transmitted by the mobile key terminal 102. If so, it performs the unlocking operation and replaces the original unlocking key with the new unlocking key, otherwise it refuses to unlock.

[0035] It is worth noting that each Bluetooth lock terminal 101 should be initialized with a public key and a corresponding private key when leaving the factory. The public key can be transmitted to other devices via Bluetooth, such as to the mobile key terminal 102, and then to the key server 103 via the mobile key terminal 102, while the private key is stored locally in the Bluetooth lock terminal 101 and is prohibited from being obtained in any way; then, when transmitting the unlocking key and the randomly generated new unlocking key, the public key of the Bluetooth lock terminal 101 can be used for encryption, so that intermediate devices cannot obtain the unlocking key and the new unlocking key, thereby achieving high data security.

[0036] At the same time, only after verifying that the unlocking key is correct can the new unlocking key be used to replace the original unlocking key, thereby ensuring that each unlocking key is safe and effective. Furthermore, key replacement instruction information can be preset in the Bluetooth lock terminal 101, and the unlocking key can be replaced only after the key replacement instruction information is unlocked, further ensuring the security of the Bluetooth lock terminal 101.

[0037] The present invention provides a substation-based Bluetooth lock system, which makes it possible to remotely control the locks in the substation by setting up a Bluetooth lock terminal 101, and is provided with a mobile key terminal 102 and a key server 103. Not only does it require authentication to unlock the Bluetooth lock terminal 101, but the use of the lock can be accurately located to an individual, making it more convenient to trace events, and the dynamic key of the Bluetooth lock terminal 101 can also be realized, greatly improving the security of the lock.

[0038] Optionally, a tester can also be set up. When the Bluetooth lock terminal 101 receives the unlocking key, the tester is dispatched to detect whether the device to be operated is powered off. If so, it is unlocked. Otherwise, a prompt message indicating that power needs to be turned off is returned to the mobile key terminal 102. The device is unlocked after the power is turned off, thereby effectively ensuring the safety of the staff.

[0039] In the embodiment of the present invention, obtaining a facial image input by a staff member and transmitting the facial image to the key server 103 includes: When the staff requests to unlock the door, the camera on the mobile key terminal 102 is dispatched to collect the staff's face image; After the face image is encrypted using the public key of the key server 103 , an encrypted face image is obtained, and the encrypted face image is transmitted to the key server 103 .

[0040] In the embodiment of the present invention, taking the face image transmitted by the mobile key terminal 102 and identifying the face image through the pre-deployed identity authentication model to determine the face recognition result includes: Taking the encrypted face image transmitted by the mobile key terminal 102, the key server 103 decrypts the encrypted face image using its own private key to obtain the decrypted face image; The decrypted face image is preprocessed to obtain a preprocessed face image, and the preprocessed face image is used as input data of a pre-deployed identity authentication model to obtain output data of the identity authentication model and obtain a face recognition result.

[0041] In an embodiment of the present invention, a method for pre-deploying an identity authentication model includes: S201, constructing an identity authentication model using a neural network model, and initializing a population based on hyperparameters of the identity authentication model; For example, a convolutional neural network can be used to build an identity authentication model. The convolutional neural network is first trained using the face of an authorized user. The training labels are authorized user labels and unauthorized user labels, thereby achieving identity authentication.

[0042] S202, obtaining an error function value corresponding to each individual in the population, and dividing the population into a first sub-population, a second sub-population, and a third sub-population in order from small to large according to the error function values; For example, facial images and authorized user labels of authorized users, as well as facial images and unauthorized user labels of unauthorized users, can be obtained. The convolutional neural network can be trained with these data, and the error function value corresponding to each individual can be obtained through the cross entropy loss function or the root mean square error function.

[0043] S203, for each first individual in the first subpopulation, performing a neighborhood search based on Gaussian distribution on the first individual to obtain the first individual after the search; S204, for each second individual in the second subpopulation, performing a social learning search on the second individual guided by the first individual and in cooperation with the second individual, to obtain a second individual after the search; S205, for each third individual in the third subpopulation, performing a locally guided perturbation search on the third individual to obtain the third individual after the search; S206, re-merging the first individual after the search, the second individual after the search, and the third individual after the search into a population, and obtaining an individual crowding factor; S207, when the individual crowding factor is less than a preset threshold, a global variation search is performed on the individuals in the fused population to obtain individuals after the global variation search; S208. Determine whether the current number of training times has reached the maximum number of training times. If so, determine the optimal individual based on the first individual after the search, the second individual after the search, and the third individual after the search, apply the hyperparameters contained in the optimal individual to the identity authentication model, and deploy the identity authentication model on the key server 103. Otherwise, based on the individual after the global variation search, return to the population division step.

[0044] Since convolutional neural networks need to be trained before they can achieve predetermined tasks, in the prior art, intelligent optimization algorithms such as particle swarm algorithms are often used for training, which have problems such as being prone to falling into local optimality and slow convergence accuracy. Therefore, an embodiment of the present invention provides a new deployment method to improve the training effect of the algorithm and improve the accuracy of face recognition.

[0045] In the embodiment of the present invention, for each first individual in the first subpopulation, a neighborhood search based on Gaussian distribution is performed on the first individual to obtain the first individual after the search, including: For each first individual in the first subpopulation, the Gaussian distribution parameter is obtained according to the error function value corresponding to the first individual:

[0046]

[0047] in, represents the Gaussian distribution parameter, represents an exponential function with the natural constant e as the base, Indicates i The fitness corresponding to the first individual is represents the fitness corresponding to the first randomly selected individual, Indicates i The error function value corresponding to the first individual, represents the preset noise term, and =0.00001; Based on the Gaussian distribution parameters, a neighborhood search is performed on the first body, and the first body after the search is obtained as:

[0048] in, Indicates t During the training i The first individual, Indicates the first body after the search, i =1,2,…,I, I represents the total number of the first individual, The mean is 0 and the standard deviation is A random number generated from a Gaussian distribution.

[0049] The Gaussian distribution-based neighborhood search provided in the embodiment of the present invention can enable individuals with better positions to implement neighborhood search, thereby realizing the development of better areas, and can effectively improve the ability of the algorithm to find the local optimal position.

[0050] In the embodiment of the present invention, for each second individual in the second subpopulation, a social learning search is performed on the second individual guided by the first individual and coordinated by the second individual to obtain the second individual after the search, including: For each second individual in the second subpopulation, randomly match the second individual with a first individual and a different second individual to obtain a first target individual and a second target individual; According to the first target individual and the second target individual, a first social learning factor and a second social learning factor are generated as follows:

[0051]

[0052] in, represents the first social learning factor, represents the second social learning factor, represents the fitness corresponding to the first target individual, represents the fitness corresponding to the second target individual, Indicates j The fitness corresponding to the second individual; According to the first target individual, the second target individual, the first social learning factor and the second social learning factor, a social learning search is performed on the second individual, and the second individual after the search is obtained as:

[0053] in, Indicates t During the training j A second individual, represents the second individual after the search, j =1,2,…,J, J represents the total number of the second individuals, represents the first random number between (0,1), represents the second random number between (0,1), represents the first target individual, Represents the second target individual.

[0054] The social learning search provided by the embodiment of the present invention can enable individuals to learn information about other locations while learning a better location, thereby achieving exploration of more local areas. The introduction of the first social learning factor and the second social learning factor can achieve adaptive search and improve search accuracy in the later stage of the algorithm.

[0055] In the embodiment of the present invention, for each third individual in the third subpopulation, a locally guided perturbation search is performed on the third individual to obtain the third individual after the search, including: Based on the second individual in the second subpopulation, the average position corresponding to all the second individuals is obtained to obtain the third target individual; Based on the current number of training times, the adaptive inertia weight coefficient is obtained as:

[0056] in, represents the adaptive inertia weight coefficient, Indicates the preset maximum number of training times. Indicates the preset maximum value of the adaptive inertia weight coefficient, Indicates the preset minimum value of the adaptive inertia weight coefficient; According to the third target individual and the adaptive inertia weight coefficient, a disturbance search is performed on the third individual, and the third individual after the search is obtained as follows:

[0057] in, Indicates t During the training m A third individual, Indicates the third individual after the search, m =1,2,…,M, M represents the total number of the third individuals, represents the third target individual, represents a perturbation factor randomly generated from (0,2).

[0058] The locally guided perturbation search provided by the embodiment of the present invention can enable the third individual to learn information of a better area, and introduce a perturbation factor so that the third individual has a certain ability to escape the local optimum while ensuring the training speed.

[0059] In the embodiment of the present invention, obtaining the individual crowding factor includes: The first individual after the search, the second individual after the search, and the third individual after the search are merged into the population again, and the Euclidean distance between each individual and all other individuals is obtained; For any individual, the Euclidean distance between the individual and all other individuals is averaged to obtain the individual crowding factor.

[0060] In the embodiment of the present invention, performing a global variation search on individuals in the fused population to obtain individuals after the global variation search includes: For the fused population, randomly select N fourth target individuals and find the historical optimal positions corresponding to the N fourth target individuals; The historical optimal positions corresponding to the N fourth target individuals are averaged to obtain the fifth target individual, and the fifth target individual is mutated through the chaotic sequence as follows:

[0061] in, The fifth target individual d Dimension parameters, d =1,2,…,D, where D represents the total number of parameters in an individual. Indicates the fifth target individual after mutation d Dimension parameters, represents the mutation search radius of the d-th dimension parameter, Represents a chaotic sequence generated by chaotic mapping; for example, a parameter can be randomly generated first, and then multiple parameters can be generated based on this parameter through chaotic mapping to form a chaotic sequence, and the parameter dimension of the chaotic sequence is the same as the parameter dimension of the individual, thereby realizing mutation search.

[0062] According to the fifth target individual after mutation, a global mutation search is performed on the individual, and the individual after the global mutation search is obtained as follows:

[0063] in, represents the individual to be mutated, Represents the individual after global mutation search , represents the historical optimal value corresponding to the individual to be mutated, represents the fifth target individual after mutation, represents the third social learning factor preset as a constant term, represents the fourth social learning factor preset as a constant term, represents the third random number between (0,1), Represents the fourth random number between (0,1).

[0064] The smaller the individual crowding factor is, the more other individuals gather around the individual. Therefore, a global mutation search can be performed on it to help the algorithm escape from the local optimum.

[0065] Optionally, in the process of global mutation search, a greedy algorithm or annealing simulation algorithm can be introduced for process control to ensure the convergence speed of the algorithm. At the same time, after each search, individuals can be processed for out-of-bounds to ensure the validity of the parameters.

[0066] In the embodiment of the present invention, when the face recognition result is that the identity recognition is passed, the unlocking information is transmitted to the mobile key terminal 102, including: When the face recognition result is that the identity recognition is passed, a new unlocking key is randomly generated and the unlocking key is found; After encrypting the new unlocking key and the unlocking key by the public key of the Bluetooth lock terminal 101, the encrypted new unlocking key and the unlocking key are obtained; Generate an unlocking instruction, and generate unlocking information according to the unlocking instruction, the encrypted new unlocking key and the unlocking key.

[0067] Those skilled in the art will readily appreciate other embodiments of the present invention after considering the specification and practicing the invention disclosed herein. The present invention is intended to cover any variations, uses or adaptations of the present invention that follow the general principles of the present invention and include common knowledge or customary techniques in the art that are not disclosed by the present invention. It should be understood that the present invention is not limited to the precise structure described above and shown in the accompanying drawings, and various modifications and changes can be made without departing from the scope thereof. The scope of the present invention is limited only by the appended claims.

Claims

1. A Bluetooth lock system based on a substation, characterized in that: include: Bluetooth lock terminal, mobile key terminal, key server; The mobile key terminal is used to obtain a facial image input by a staff member and transmit the facial image to a key server; wherein each mobile key terminal corresponds to a staff member one by one; The key server is used to obtain the face image transmitted by the mobile key terminal, and recognize the face image through the pre-deployed identity authentication model to determine the face recognition result; wherein the face recognition result includes identity recognition passing or identity recognition failing; The key server is further configured to transmit unlocking information to the mobile key terminal when the face recognition result is that the identity recognition is passed; wherein the unlocking information includes an unlocking key and a randomly generated new unlocking key; The mobile key terminal is used to transmit the unlocking key and the new unlocking key to the Bluetooth lock terminal through a Bluetooth encrypted connection after receiving the unlocking information transmitted by the key server; The Bluetooth lock terminal is used to verify whether the unlocking key is correct after receiving the unlocking key and the new unlocking key transmitted by the mobile key terminal. If so, the unlocking operation is performed and the new unlocking key is used to replace the original unlocking key. Otherwise, the unlocking is refused.

2. The substation-based Bluetooth lock system according to claim 1, characterized in that: Acquiring a facial image input by a staff member and transmitting the facial image to a key server includes: When the staff requests to unlock the door, the camera on the mobile key terminal is dispatched to capture the staff's face image; After the face image is encrypted using the public key of the key server, an encrypted face image is obtained, and the encrypted face image is transmitted to the key server.

3. The substation-based Bluetooth lock system according to claim 2, characterized in that: The facial image transmitted by the mobile key terminal is obtained, and the facial image is recognized through the pre-deployed identity authentication model to determine the facial recognition result, including: The encrypted face image transmitted by the mobile key terminal is obtained, and the key server decrypts the encrypted face image using its own private key to obtain the decrypted face image; The decrypted face image is preprocessed to obtain a preprocessed face image, and the preprocessed face image is used as input data of a pre-deployed identity authentication model to obtain output data of the identity authentication model and obtain a face recognition result.

4. The substation-based Bluetooth lock system according to claim 3, characterized in that: Pre-deployment methods for authentication models, including: The neural network model is used to build the identity authentication model, and the population is initialized based on the hyperparameters of the identity authentication model; Obtaining the error function value corresponding to each individual in the population, and dividing the population into a first sub-population, a second sub-population, and a third sub-population in descending order of the error function values; For each first individual in the first subpopulation, a neighborhood search based on Gaussian distribution is performed on the first individual to obtain the first individual after the search; For each second individual in the second subpopulation, a social learning search is performed on the second individual guided by the first individual and in cooperation with the second individual to obtain the second individual after the search; For each third individual in the third subpopulation, a locally guided perturbation search is performed on the third individual to obtain the third individual after the search; The first individual after the search, the second individual after the search, and the third individual after the search are merged into a population again to obtain the individual crowding factor; When the individual crowding factor is less than a preset threshold, a global mutation search is performed on the individuals in the fused population to obtain individuals after the global mutation search; Determine whether the current number of training times has reached the maximum number of training times. If so, determine the optimal individual based on the first individual after the search, the second individual after the search, and the third individual after the search, apply the hyperparameters contained in the optimal individual to the identity authentication model, and deploy the identity authentication model on the key server. Otherwise, based on the individual after the global mutation search, return to the population division step.

5. The substation-based Bluetooth lock system according to claim 4, characterized in that: For each first individual in the first subpopulation, a neighborhood search based on Gaussian distribution is performed on the first individual to obtain the first individual after the search, including: For each first individual in the first subpopulation, the Gaussian distribution parameter is obtained according to the error function value corresponding to the first individual: in, represents the Gaussian distribution parameter, represents an exponential function with the natural constant e as the base, Indicates i The fitness corresponding to the first individual is represents the fitness corresponding to the first randomly selected individual, Indicates i The error function value corresponding to the first individual, represents the preset noise term, and =0.00001; Based on the Gaussian distribution parameters, a neighborhood search is performed on the first body, and the first body after the search is obtained as: in, Indicates t During the training process i The first individual, Indicates the first body after the search, i =1,2,…,I, I represents the total number of the first individual, The mean is 0 and the standard deviation is A random number generated from a Gaussian distribution.

6. The substation-based Bluetooth lock system according to claim 5, characterized in that: For each second individual in the second subpopulation, a social learning search is performed on the second individual guided by the first individual and coordinated by the second individual to obtain the second individual after the search, including: For each second individual in the second subpopulation, randomly match the second individual with a first individual and a different second individual to obtain a first target individual and a second target individual; According to the first target individual and the second target individual, a first social learning factor and a second social learning factor are generated as follows: in, represents the first social learning factor, represents the second social learning factor, represents the fitness corresponding to the first target individual, represents the fitness corresponding to the second target individual, Indicates j The fitness corresponding to the second individual; According to the first target individual, the second target individual, the first social learning factor and the second social learning factor, a social learning search is performed on the second individual, and the second individual after the search is obtained as: in, Indicates t During the training j A second individual, represents the second individual after the search, j =1,2,…,J, J represents the total number of the second individuals, represents the first random number between (0,1), represents the second random number between (0,1), represents the first target individual, Represents the second target individual.

7. The substation-based Bluetooth lock system according to claim 6, characterized in that: For each third individual in the third subpopulation, a locally guided perturbation search is performed on the third individual to obtain the third individual after the search, including: Based on the second individual in the second subpopulation, the average position corresponding to all the second individuals is obtained to obtain the third target individual; Based on the current number of training times, the adaptive inertia weight coefficient is obtained as: in, represents the adaptive inertia weight coefficient, Indicates the preset maximum number of training times. Indicates the preset maximum value of the adaptive inertia weight coefficient, Indicates the preset minimum value of the adaptive inertia weight coefficient; According to the third target individual and the adaptive inertia weight coefficient, a disturbance search is performed on the third individual, and the third individual after the search is obtained as follows: in, Indicates t During the training process m A third individual, Indicates the third individual after the search, m =1,2,…,M, M represents the total number of the third individuals, represents the third target individual, represents a perturbation factor randomly generated from (0,2).

8. The substation-based Bluetooth lock system according to claim 7, characterized in that: Get individual crowding factors, including: The first individual after the search, the second individual after the search, and the third individual after the search are merged into the population again, and the Euclidean distance between each individual and all other individuals is obtained; For any individual, the Euclidean distance between the individual and all other individuals is averaged to obtain the individual crowding factor.

9. The substation-based Bluetooth lock system according to claim 3, characterized in that: Perform global mutation search on the individuals in the fused population to obtain individuals after global mutation search, including: For the fused population, randomly select N fourth target individuals and find the historical optimal positions corresponding to the N fourth target individuals; The historical optimal positions corresponding to the N fourth target individuals are averaged to obtain the fifth target individual, and the fifth target individual is mutated through the chaotic sequence as follows: in, Indicates the fifth target individual d Dimension parameters, d =1,2,…,D, where D represents the total number of parameters in an individual. Indicates the fifth target individual after mutation d Dimension parameters, represents the mutation search radius of the d-th dimension parameter, represents the chaotic sequence generated by chaotic mapping; According to the fifth target individual after mutation, a global mutation search is performed on the individual, and the individual after the global mutation search is obtained as follows: in, represents the individual to be mutated, Represents the individual after global mutation search , represents the historical optimal value corresponding to the individual to be mutated, represents the fifth target individual after mutation, represents the third social learning factor preset as a constant term, represents the fourth social learning factor preset as a constant term, represents the third random number between (0,1), Represents the fourth random number between (0,1).

10. The substation-based Bluetooth lock system according to claim 9, characterized in that: When the face recognition result is that the identity recognition is passed, the unlocking information is transmitted to the mobile key terminal, including: When the face recognition result is that the identity recognition is passed, a new unlocking key is randomly generated and the unlocking key is found; After encrypting the new unlocking key and the unlocking key by the public key of the Bluetooth lock terminal, the encrypted new unlocking key and the unlocking key are obtained; Generate an unlocking instruction, and generate unlocking information according to the unlocking instruction, the encrypted new unlocking key and the unlocking key.

Images