A UAV cluster collaborative processing method

Abstract

The present invention relates to the technical field of unmanned aerial vehicles, and in particular to a method for collaborative processing of unmanned aerial vehicle clusters, comprising: generating a private key for each unmanned aerial vehicle in the unmanned aerial vehicle cluster to establish communication with a central host; obtaining historical driving data of each unmanned aerial vehicle; inputting the historical driving data into a prediction model for training to obtain predicted driving data; after the central host obtains task information, generating an execution ciphertext for each unmanned aerial vehicle based on the predicted driving data, and sending each execution ciphertext to a corresponding unmanned aerial vehicle; each unmanned aerial vehicle decrypts the corresponding execution ciphertext through a private key to obtain an execution plaintext, and executes the task based on the execution plaintext, and when each unmanned aerial vehicle establishes communication with the central host and the tasks sent by the central host are transmitted after being encrypted, thereby avoiding leakage of task information and solving the problem that the task information released by the central host to the node is in plaintext and is easy to be stolen and leaked.

Description

Technical Field

[0001] The present invention relates to the technical field of unmanned aerial vehicles, and in particular to a collaborative processing method for a cluster of unmanned aerial vehicles. Background Art

[0002] The coordinated control of multi-UAV clusters is divided into centralized control, distributed control and distributed control. The centralized control method has a unique central host that controls the global nodes and has absolute decision-making power. This control structure has low system complexity and low data communication pressure.

[0003] In the above method, the task information released by the central host to the node is in plain text, which is easy to be stolen and leaked, thereby causing the task execution to fail. Summary of the invention

[0004] The purpose of the present invention is to provide a drone cluster collaborative processing method, aiming to solve the problem that the task information released by the central host to the nodes is in plain text and can be easily stolen and leaked.

[0005] To achieve the above object, the present invention provides a UAV cluster collaborative processing method, comprising the following steps:

[0006] Each drone in the drone cluster generates its own private key to establish communication with the central host;

[0007] Obtaining historical driving data of each of the drones;

[0008] Inputting the historical driving data into a prediction model for training to obtain predicted driving data;

[0009] After acquiring the task information, the central host generates an execution ciphertext for each of the drones based on the predicted driving data, and sends each of the execution ciphertexts to the corresponding drone;

[0010] Each of the drones decrypts the corresponding execution ciphertext using a private key to obtain execution plaintext, and executes the task based on the execution plaintext.

[0011] The specific method for each drone to generate its own private key and establish communication with the central host is:

[0012] The drone obtains the identity information of the central host;

[0013] The drone uses a symmetric key obtained by a random number generator;

[0014] The drone sets the length of time for the symmetric key to communicate with the central host;

[0015] The drone generates its own private key to encrypt the symmetric key, the time length and the timestamp to obtain a first ciphertext;

[0016] The drone encrypts the first ciphertext and the identity information of the drone through the identity information of the central host to obtain a second ciphertext, and sends the second ciphertext to the central host;

[0017] The central host generates its own private key to decrypt the second ciphertext to obtain a decrypted plaintext;

[0018] The central host verifies the identity information of the drone in the decrypted plaintext, and upon successful verification, obtains the symmetric key and the time length;

[0019] The central host notifies the drone of successful decryption through the symmetric key, and communication establishment is complete.

[0020] The specific method of inputting the historical driving data into the prediction model for training to obtain the predicted driving data is:

[0021] Obtain public driving datasets;

[0022] Build a neural network model;

[0023] Using the public driving data set to train the neural network model to obtain a prediction model;

[0024] The historical driving data is input into a prediction model for training to obtain predicted driving data.

[0025] The specific method of using the public driving data set to train the neural network model to obtain the prediction model is:

[0026] Preprocessing the public driving data set to obtain a preprocessed data set;

[0027] Dividing the preprocessed data set into a training set and a validation set;

[0028] Using the training set to train the neural network model to obtain a training model;

[0029] The training model is verified using the verification set, and if the verification passes, a prediction model is obtained.

[0030] The ratio of the training set to the validation set is 8:2.

[0031] The specific method of preprocessing the public driving data set to obtain the preprocessed data set is as follows:

[0032] Filtering the public driving data set to obtain a filtered data set;

[0033] The format of the filtered data set is adjusted based on the input format of the neural network model to obtain a preprocessed data set.

[0034] A drone cluster collaborative processing method of the present invention establishes communication with a central host by having each drone in the drone cluster generate its own private key; obtains historical driving data of each drone; inputs the historical driving data into a prediction model for training to obtain predicted driving data; after the central host obtains task information, it generates an execution ciphertext for each drone based on the predicted driving data, and sends each execution ciphertext to the corresponding drone; each drone decrypts the corresponding execution ciphertext through a private key to obtain an execution plaintext, and executes the task based on the execution plaintext, each drone uses a ciphertext to establish communication with the central host, and at the same time, the task sent by the central host is also an execution ciphertext generated after encryption, and the drone corresponding to the execution ciphertext needs to use a private key to decrypt and generate plaintext, thereby avoiding the leakage of the task information and solving the problem that the task information released by the central host to the node is in plaintext and is easy to be stolen and leaked. BRIEF DESCRIPTION OF THE DRAWINGS

[0035] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in 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 described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0036] Figure 1 It is a flow chart of a UAV cluster collaborative processing method provided by the present invention.

[0037] Figure 2 It is a flowchart for each drone in the drone cluster to generate its own private key to establish communication with the central host.

[0038] Figure 3 The historical driving data is input into the prediction model for training to obtain a flow chart of predicted driving data.

[0039] Figure 4 The present invention is a flowchart of using the public driving data set to train the neural network model to obtain a prediction model. DETAILED DESCRIPTION

[0040] Embodiments of the present invention are described in detail below, examples of which are shown in the accompanying drawings, wherein the same or similar reference numerals throughout represent the same or similar elements or elements having the same or similar functions. The embodiments described below with reference to the accompanying drawings are exemplary and are intended to be used to explain the present invention, and should not be construed as limiting the present invention.

[0041] See also Figures 1 to 4 The present invention provides a UAV cluster collaborative processing method, comprising the following steps:

[0042] S1 generates its own private key for each drone in the drone cluster to establish communication with the central host;

[0043] The specific method is:

[0044] S11 The drone obtains the identity information of the central host;

[0045] S12: The drone uses a symmetric key obtained by a random number generator;

[0046] S13, the drone sets the time length for the symmetric key to communicate with the central host;

[0047] S14, the drone generates its own private key to encrypt the symmetric key, the time length and the timestamp to obtain a first ciphertext;

[0048] S15, the drone encrypts the first ciphertext and the identity information of the drone through the identity information of the central host to obtain a second ciphertext, and sends the second ciphertext to the central host;

[0049] S16 The central host generates its own private key to decrypt the second ciphertext to obtain a decrypted plaintext;

[0050] S17 The central host verifies the identity information of the drone in the decrypted plaintext, and obtains the symmetric key and the time length after successful verification;

[0051] Specifically, the drone and the central host can communicate only within the time length. If the time length is exceeded, steps S11 to S17 need to be executed again to establish a communication connection.

[0052] S18 The central host notifies the drone of successful decryption through the symmetric key, and the communication is established.

[0053] Specifically, the drone may use the symmetric key, the new key, the allowed time of the new key, and the new key timestamp to perform key replacement with the central host.

[0054] Specifically, key replacement: Since using the same symmetric key for communication for a long time may give attackers an advantage in decryption, it is necessary to replace the key for a long time to ensure the security of the system. Therefore, the key replacement process can occur at any stage when the original key has not expired or has expired. In the case that the original key expires, the user and the central host need to re-execute the above key negotiation process. If the original key has not expired, if the user wants to replace the key, the original key, the new key, the allowed time of the new key, and the timestamp must be used as information.

[0055] S2 obtains historical driving data of each of the drones;

[0056] Specifically, a MAP map is searched according to historical working conditions, and information such as the speed, position, current driving path, etc. of the UAV is obtained based on the MAP map to obtain historical driving data.

[0057] S3 inputs the historical driving data into a prediction model for training to obtain predicted driving data;

[0058] The specific method is:

[0059] S31 obtains a public driving data set;

[0060] Specifically, the public driving data is crawled on the website through crawler technology.

[0061] S32 builds a neural network model;

[0062] S33 uses the public driving data set to train the neural network model to obtain a prediction model;

[0063] The specific method is:

[0064] S331 preprocesses the public driving data set to obtain a preprocessed data set;

[0065] Specifically, the public driving data set is filtered so that redundant data in the public driving data set is filtered out to obtain a filtered data set; the format of the filtered data set is adjusted based on the input format of the neural network model to obtain a preprocessed data set, thereby increasing the accuracy of subsequent training of the neural network model.

[0066] S332 divides the preprocessed data set into a training set and a validation set;

[0067] Specifically, the ratio of the training set to the validation set is 8:2.

[0068] S333 uses the training set to train the neural network model to obtain a training model;

[0069] Specifically, before training the neural network model, the model parameters, learning rate and number of iterations of the neural network model are set.

[0070] S334 uses the verification set to verify the training model, and if the verification passes, a prediction model is obtained.

[0071] Specifically, if the verification fails, the model parameters of the neural network model are adjusted and retrained.

[0072] S34 inputs the historical driving data into a prediction model for training to obtain predicted driving data.

[0073] S4 After the central host obtains the task information, it generates an execution ciphertext for each of the drones based on the predicted driving data, and sends each of the execution ciphertexts to the corresponding drone;

[0074] S5 Each of the drones decrypts the corresponding execution ciphertext using a private key to obtain execution plaintext, and executes the task based on the execution plaintext.

[0075] A drone cluster collaborative processing method of the present invention establishes communication with a central host by having each drone in the drone cluster generate its own private key; obtains historical driving data of each drone; inputs the historical driving data into a prediction model for training to obtain predicted driving data; after the central host obtains task information, it generates an execution ciphertext for each drone based on the predicted driving data, and sends each execution ciphertext to the corresponding drone; each drone decrypts the corresponding execution ciphertext through a private key to obtain an execution plaintext, and executes the task based on the execution plaintext, each drone uses a ciphertext to establish communication with the central host, and at the same time, the task sent by the central host is also an execution ciphertext generated after encryption, and the drone corresponding to the execution ciphertext needs to use a private key to decrypt and generate plaintext, thereby avoiding the leakage of the task information and solving the problem that the task information released by the central host to the node is in plaintext and is easy to be stolen and leaked.

[0076] What is disclosed above is only a preferred embodiment of a drone cluster collaborative processing method of the present invention. Of course, this cannot be used to limit the scope of rights of the present invention. Ordinary technicians in this field can understand that all or part of the processes of the above embodiments and equivalent changes made according to the claims of the present invention still fall within the scope of the invention.

Claims

1. A drone cluster collaborative processing method, characterized in that: The following steps are involved: Each drone in the drone cluster generates its own private key to establish communication with the central host; Obtaining historical driving data of each of the drones; Inputting the historical driving data into a prediction model for training to obtain predicted driving data; After acquiring the task information, the central host generates an execution ciphertext for each of the drones based on the predicted driving data, and sends each of the execution ciphertexts to the corresponding drone; Each of the drones decrypts the corresponding execution ciphertext using a private key to obtain an execution plaintext, and executes a task based on the execution plaintext; The specific method for each drone to generate its own private key and establish communication with the central host is: The drone obtains the identity information of the central host; The drone uses a symmetric key obtained by a random number generator; The drone sets the length of time for the symmetric key to communicate with the central host; The drone generates its own private key to encrypt the symmetric key, the time length and the timestamp to obtain a first ciphertext; The drone encrypts the first ciphertext and the identity information of the drone through the identity information of the central host to obtain a second ciphertext, and sends the second ciphertext to the central host; The central host generates its own private key to decrypt the second ciphertext to obtain a decrypted plaintext; The central host verifies the identity information of the drone in the decrypted plaintext, and upon successful verification, obtains the symmetric key and the time length; The central host notifies the drone of successful decryption through the symmetric key, and communication establishment is complete.

2. The UAV cluster collaborative processing method according to claim 1, characterized in that: The specific method of inputting the historical driving data into the prediction model for training to obtain the predicted driving data is: Obtain public driving datasets; Build a neural network model; Using the public driving data set to train the neural network model to obtain a prediction model; The historical driving data is input into a prediction model for training to obtain predicted driving data.

3. The UAV cluster collaborative processing method according to claim 2, characterized in that: The specific method of using the public driving data set to train the neural network model to obtain the prediction model is: Preprocessing the public driving data set to obtain a preprocessed data set; Dividing the preprocessed data set into a training set and a validation set; Using the training set to train the neural network model to obtain a training model; The training model is verified using the verification set, and if the verification passes, a prediction model is obtained.

4. The UAV cluster collaborative processing method according to claim 3, characterized in that: The ratio of the training set to the validation set is 8:

2.

5. The UAV cluster collaborative processing method according to claim 4, characterized in that: The specific method of preprocessing the public driving data set to obtain the preprocessed data set is: Filtering the public driving data set to obtain a filtered data set; The format of the filtered data set is adjusted based on the input format of the neural network model to obtain a preprocessed data set.

Images