Unmanned aerial vehicle battery charging control method and related device

Abstract

The invention provides an unmanned aerial vehicle battery charging control method and a related device, and relates to the technical field of unmanned aerial vehicles. The method comprises the following steps: in response to a detected unmanned aerial vehicle battery charging instruction, obtaining charging and discharging state parameters of a to-be-charged unmanned aerial vehicle battery; the charge and discharge state parameters are input into a charge state determination model, the charge state of the unmanned aerial vehicle battery is determined according to the charge and discharge state parameters, the charge state estimation value of the unmanned aerial vehicle battery is obtained, and the charge state determination model is determined by a convolutional neural network and a recurrent neural network; determining the performance of the unmanned aerial vehicle battery according to the charge state estimation value; and when the performance of the unmanned aerial vehicle battery is determined to be qualified, controlling a charging device to charge the unmanned aerial vehicle battery. According to the invention, the accuracy of the unmanned aerial vehicle battery evaluation result is improved.

Description

Technical Field

[0001] This application relates to the technical field of unmanned aerial vehicles, and particularly to a method for controlling the charging of an unmanned aerial vehicle battery and related devices. Background Art

[0002] With the rapid development of unmanned aerial vehicle technology, the problem of performance loss of unmanned aerial vehicle batteries has gradually emerged, becoming a key factor affecting the operation stability and economy of unmanned aerial vehicles. Therefore, developing an efficient and accurate method for evaluating the performance of unmanned aerial vehicle batteries has become an urgent problem to be solved in the current technical field of unmanned aerial vehicles.

[0003] Currently, the performance evaluation of unmanned aerial vehicle batteries usually adopts a charge-discharge test method. This method first requires removing the battery from the unmanned aerial vehicle, then connecting it to a dedicated test device to measure the charge-discharge state parameters, and finally obtaining the battery performance evaluation result through manual calculation. However, since the battery evaluation process relies on manual calculation, errors are easily introduced, resulting in a low accuracy of the evaluation result. Summary of the Invention

[0004] This application provides a method for controlling the charging of an unmanned aerial vehicle battery and related devices to improve the accuracy of the evaluation result of the unmanned aerial vehicle battery.

[0005] In a first aspect, this application provides a method for controlling the charging of an unmanned aerial vehicle battery, including:

[0006] In response to detecting an unmanned aerial vehicle battery charging instruction, obtaining the charge-discharge state parameters of the unmanned aerial vehicle battery to be charged;

[0007] Inputting the charge-discharge state parameters into a state of charge determination model, determining the state of charge of the unmanned aerial vehicle battery according to the charge-discharge state parameters, and obtaining an estimated value of the state of charge of the unmanned aerial vehicle battery. The state of charge determination model is determined by a convolutional neural network and a recurrent neural network;

[0008] Determining the performance of the unmanned aerial vehicle battery according to the estimated value of the state of charge;

[0009] When the performance of the unmanned aerial vehicle battery is determined to be qualified, controlling a charging device to perform a charging operation on the unmanned aerial vehicle battery.

[0010] In a possible implementation manner, when the performance of the unmanned aerial vehicle battery is determined to be qualified, controlling a charging device to perform a charging operation on the unmanned aerial vehicle battery includes:

[0011] When the performance of the unmanned aerial vehicle battery is determined to be qualified, adjusting the charging parameters of the charging device to ideal charging parameters matching the unmanned aerial vehicle battery to complete the charging operation on the unmanned aerial vehicle battery, where the charging parameters include charging current, charging voltage, and charging rate;

[0012] Or,

[0013] When the performance of the UAV battery is determined to be qualified and the estimated state of charge of the UAV battery is less than or equal to the constant current charging threshold, control the charging device to charge the UAV battery in the constant current mode; when the performance of the UAV battery is determined to be qualified and the state of charge of the UAV battery is greater than the constant current threshold, control the charging device to charge the UAV battery in the constant voltage mode.

[0014] In a possible implementation manner, determining the state of charge of the UAV battery according to the charge and discharge state parameters to obtain an estimated value of the state of charge of the UAV battery includes:

[0015] Input the charge and discharge state parameters into a convolutional neural network to extract local features between the charge and discharge state parameters;

[0016] Input the local features into a recurrent neural network. In the recurrent neural network, capture the long-term dependencies of the local features in the time series, and determine the estimated value of the state of charge of the UAV battery according to the long-term dependencies.

[0017] In a possible implementation manner, the convolutional neural network includes an input layer, a one-dimensional convolutional layer, and a one-dimensional pooling layer. The output of the input layer serves as the input of the connected one-dimensional convolutional layer, and the output of the one-dimensional convolutional layer serves as the output of the connected one-dimensional pooling layer; the recurrent neural network includes a long short-term memory network, a Dropout layer, and a fully connected layer. The output of the one-dimensional pooling layer serves as the input of the long short-term memory network, the output of the long short-term memory network serves as the input of the Dropout layer, and the output of the Dropout layer serves as the input of the fully connected layer; where:

[0018] The input layer is used to receive the charge and discharge state parameters;

[0019] The one-dimensional convolutional layer is used to extract local features in the charge and discharge state parameters;

[0020] The one-dimensional pooling layer is used to reduce the dimension of the local features;

[0021] The long short-term memory network is used to capture the long-term dependencies of the local features in the time series;

[0022] The Dropout layer is used to randomly deactivate some neurons in the output of the long short-term memory network during the training stage of the state of charge determination model;

[0023] The fully connected layer is used to determine the estimated value of the state of charge of the UAV battery through linear transformation according to the output in the Dropout layer.

[0024] In a possible implementation manner, the hyperparameters of the state of charge determination model can be determined by the following method:

[0025] Obtain the velocities and positions of a particle swarm composed of multiple particles, where the position of each particle represents a set of hyperparameter combinations in the charge state determination model, and the hyperparameters at least include the learning rate, the number of hidden layer units, and the maximum number of training iterations, and the velocity of each particle represents the direction and step size of the particle moving in the search space;

[0026] Based on the positions of the particle swarm, train the charge state determination model, and calculate the fitness value corresponding to each set of hyperparameters to determine the individual optimal position and the global optimal position in the particle swarm;

[0027] According to the individual optimal position and the global optimal position, apply the velocity update equation in the particle swarm optimization algorithm to update the velocities of the particle swarm, and adjust the positions of the particle swarm based on the updated velocities;

[0028] Based on the positions of the updated particle swarm, retrain the charge state determination model, and update the individual optimal position and the global optimal position in the particle swarm;

[0029] Judge whether the current iteration number of the charge state determination model reaches the maximum number of training iterations;

[0030] If the current iteration number reaches the maximum number of training iterations, determine the hyperparameter combination corresponding to the global optimal position as the hyperparameters of the charge state determination model;

[0031] If the current iteration number does not reach the number of training iterations, return to train the charge state determination model, update the individual optimal position and the global optimal position in the particle swarm, and continue iterative optimization.

[0032] In a possible implementation manner, determine the performance of the UAV battery according to the charge state estimation value, including:

[0033] Judge whether the charge state estimation value is greater than or equal to the estimation threshold;

[0034] When the charge state estimation value is greater than or equal to the estimation threshold, determine that the performance of the UAV battery is qualified;

[0035] When the charge state estimation value is less than the estimation threshold, determine that the performance of the UAV battery is unqualified.

[0036] In a possible implementation manner, it further includes:

[0037] When it is determined that the performance of the UAV battery is unqualified, display the charge state estimation value and a prompt message indicating that the battery performance is unqualified on the visualization big screen;

[0038] And / or,

[0039] When it is determined that the performance of the UAV battery is unqualified, control the buzzer to emit a sound warning message;

[0040] and / or,

[0041] When it is determined that the performance of the UAV battery is unqualified, control the running lights to emit visual warning information.

[0042] In a possible implementation manner, it further includes:

[0043] Perform data preprocessing operations on the charge and discharge state parameters, and the data preprocessing operations include at least one of the following: Kalman filtering processing, moving average processing, and data cleaning.

[0044] In a possible implementation manner, it further includes:

[0045] Obtain the model of the UAV battery;

[0046] Based on the performance of the UAV battery at different times, determine the change law of the performance state of the UAV battery of the model;

[0047] According to the change law of the performance state, predict the moment when the performance of the UAV battery of the model is unqualified;

[0048] Based on the moment when the performance is unqualified, determine the management measures for the UAV battery.

[0049] In a second aspect, the present application provides a UAV battery charging control device, including:

[0050] An acquisition module, configured to respond to detecting a UAV battery charging instruction and acquire the charge and discharge state parameters of the UAV battery to be charged;

[0051] A first determination module, configured to input the charge and discharge state parameters into a state of charge determination model, determine the state of charge of the UAV battery according to the charge and discharge state parameters, and obtain an estimated value of the state of charge of the UAV battery. The state of charge determination model is determined by a convolutional neural network and a recurrent neural network;

[0052] A second determination module, configured to determine the performance of the UAV battery according to the estimated value of the state of charge;

[0053] A charging module, configured to control a charging device to perform a charging operation on the UAV battery when the performance of the UAV battery is determined to be qualified.

[0054] In a possible implementation manner, the charging module is specifically configured to:

[0055] When the performance of the UAV battery is determined to be qualified, adjust the charging parameters of the charging device to ideal charging parameters matching the UAV battery, and complete the charging operation on the UAV battery, where the charging parameters include charging current, charging voltage, and charging rate;

[0056] Or,

[0057] When the performance of the UAV battery is determined to be qualified and the estimated value of the state of charge of the UAV battery is less than or equal to the constant current charging threshold, control the charging device to charge the UAV battery in the constant current mode; when the performance of the UAV battery is determined to be qualified and the state of charge of the UAV battery is greater than the constant current threshold, control the charging device to charge the UAV battery in the constant voltage mode.

[0058] In a possible implementation manner, the first determination module is specifically configured to:

[0059] Input the charge and discharge state parameters into the convolutional neural network, and extract the local features between the charge and discharge state parameters;

[0060] Input the local features into the recurrent neural network. In the recurrent neural network, capture the long-term dependence relationship of the local features in the time series, and determine the estimated value of the state of charge of the UAV battery according to the long-term dependence relationship.

[0061] In a possible implementation manner, the convolutional neural network includes an input layer, a one-dimensional convolutional layer, and a one-dimensional pooling layer. The output of the input layer serves as the input of the connected one-dimensional convolutional layer, and the output of the one-dimensional convolutional layer serves as the output of the connected one-dimensional pooling layer; the recurrent neural network includes a long short-term memory network, a Dropout layer, and a fully connected layer. The output of the one-dimensional pooling layer serves as the input of the long short-term memory network, the output of the long short-term memory network serves as the input of the Dropout layer, and the output of the Dropout layer serves as the input of the fully connected layer; wherein:

[0062] The input layer is used to receive the charge and discharge state parameters;

[0063] The one-dimensional convolutional layer is used to extract the local features in the charge and discharge state parameters;

[0064] The one-dimensional pooling layer is used to reduce the dimension of the local features;

[0065] The long short-term memory network is used to capture the long-term dependence relationship of the local features in the time series;

[0066] The Dropout layer is used to randomly deactivate some neurons output by the long short-term memory network during the training stage of the state of charge determination model;

[0067] The fully connected layer is used to determine the estimated value of the state of charge of the UAV battery through linear transformation according to the output in the Dropout layer.

[0068] In a possible implementation manner, the hyperparameters of the state of charge determination model can be determined by the following method:

[0069] Obtain the velocities and positions of a swarm of particles composed of multiple particles, where the position of each particle represents a set of hyperparameter combinations in the charge state determination model, and the hyperparameters at least include the learning rate, the number of hidden layer units, and the maximum number of training iterations, and the velocity of each particle represents the direction and step size of the particle moving in the search space;

[0070] Based on the positions of the particle swarm, train the charge state determination model, and calculate the fitness value corresponding to each set of hyperparameters to determine the individual optimal position and the global optimal position in the particle swarm;

[0071] According to the individual optimal position and the global optimal position, apply the velocity update equation in the particle swarm optimization algorithm to update the velocities of the particle swarm, and adjust the positions of the particle swarm based on the updated velocities;

[0072] Based on the positions of the updated particle swarm, retrain the charge state determination model, and update the individual optimal position and the global optimal position in the particle swarm;

[0073] Judge whether the current iteration number of the charge state determination model reaches the maximum number of training iterations;

[0074] If the current iteration number reaches the maximum number of training iterations, determine the hyperparameter combination corresponding to the global optimal position as the hyperparameters of the charge state determination model;

[0075] If the current iteration number does not reach the number of training iterations, return to train the charge state determination model, update the individual optimal position and the global optimal position in the particle swarm, and continue iterative optimization.

[0076] In a possible implementation manner, the second determination module is specifically configured to:

[0077] Judge whether the charge state estimation value is greater than or equal to the estimation threshold;

[0078] When the charge state estimation value is greater than or equal to the estimation threshold, determine that the performance of the UAV battery is qualified;

[0079] When the charge state estimation value is less than the estimation threshold, determine that the performance of the UAV battery is unqualified.

[0080] In a possible implementation manner, the UAV battery charging control further includes a processing module, and the processing module is specifically configured to:

[0081] When it is determined that the performance of the UAV battery is unqualified, display the charge state estimation value and a prompt message indicating that the battery performance is unqualified on the visualization large screen;

[0082] And / or,

[0083] When it is determined that the performance of the UAV battery is unqualified, control the buzzer to emit a sound warning message;

[0084] and / or

[0085] When it is determined that the performance of the UAV battery is unqualified, control the running water lamp to emit a visual warning message.

[0086] In a possible implementation, the processing module is further configured to:

[0087] Perform data preprocessing operations on the charge and discharge state parameters, and the data preprocessing operations at least include one of the following: Kalman filtering processing, moving average processing, and data cleaning.

[0088] In a possible implementation, the processing module is further configured to:

[0089] Obtain the model of the UAV battery;

[0090] Based on the performance of the UAV battery at different times, determine the change law of the performance state of the UAV battery of the model;

[0091] According to the change law of the performance state, predict the moment when the performance of the UAV battery of the model is unqualified;

[0092] Based on the moment when the performance is unqualified, determine the management measures for the UAV battery.

[0093] In a third aspect, the present application provides an electronic device, including: a memory, a processor;

[0094] The memory stores computer execution instructions;

[0095] The processor executes the computer execution instructions stored in the memory, so that the processor executes the above first aspect and / or various possible implementation manners of the first aspect.

[0096] In a fourth aspect, the present application provides a computer-readable storage medium, in which computer execution instructions are stored, and when the computer execution instructions are executed by a processor, they are used to implement the above first aspect and / or various possible implementation manners of the first aspect.

[0097] In a fifth aspect, the present application provides a computer program product, including a computer program, and when the computer program is executed by a processor, it implements the above first aspect and / or various possible implementation manners of the first aspect.

[0098] The drone battery charging control method and related device provided by this application relate to the field of drone technologies. The method includes: in response to detecting a drone battery charging instruction, obtaining the charge and discharge state parameters of the drone battery to be charged; inputting the charge and discharge state parameters into a state of charge determination model, and determining the state of charge of the drone battery according to the charge and discharge state parameters to obtain an estimated value of the state of charge of the drone battery, where the state of charge determination model is determined by a convolutional neural network and a recurrent neural network; determining the performance of the drone battery according to the estimated value of the state of charge; when the performance of the drone battery is determined to be qualified, controlling a charging device to perform a charging operation on the drone battery. By obtaining the charge and discharge state parameters of the drone battery to be charged in real time after detecting the drone charging instruction, this application does not need to wait for a complete charge and discharge time, which is beneficial to reducing the time cost; adopting a state of charge determination model combining a convolutional neural network and a recurrent neural network to learn and analyze the charge and discharge state parameters of the drone battery, realizing a high-precision estimation of the state of charge of the drone battery; evaluating the performance of the drone battery based on the estimated value of the state of charge can improve the accuracy of the evaluation result of the drone battery; when the performance of the drone battery is determined to be qualified, controlling the charging device to perform a charging operation on the drone battery can comprehensively evaluate the battery state before the charging operation, ensuring that only batteries with qualified performance are charged. This design effectively avoids charging batteries with poor performance or potential faults, thereby reducing safety hazards such as battery overheating and fire, and avoiding waste of charging resources at the same time. Description of the Drawings

[0099] The drawings herein are incorporated into the specification and constitute a part of this specification, showing embodiments consistent with this application, and are used together with the specification to explain the principles of this application.

[0100] Figure 1 Flow diagram of the drone battery charging control method provided by this application Figure 1 ;

[0101] Figure 2 Flow diagram of the drone battery charging control method provided by an embodiment of this application Figure 2 ;

[0102] Figure 3 Schematic diagram of the principle of the SOC prediction model based on PSO-CNN-LSTM provided by an embodiment of this application;

[0103] Figure 4 Schematic diagram of the structure of the drone battery charging control device provided by this application;

[0104] Figure 5 Schematic diagram of the structure of an electronic device provided by an embodiment of this application.

[0105] Through the above-mentioned accompanying drawings, specific embodiments of the present application have been shown, and will be described in more detail hereinafter. These drawings and the textual description are not intended to limit the scope of the concept of the present application in any way, but to illustrate the concept of the present application to those skilled in the art by referring to specific embodiments. Detailed Description of the Embodiments

[0106] Exemplary embodiments will be described in detail herein, and examples thereof are shown in the accompanying drawings. When the following description refers to the accompanying drawings, unless otherwise indicated, the same numbers in different drawings represent the same or similar elements. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with the present application. On the contrary, they are merely examples of devices and methods consistent with some aspects of the present application as detailed in the appended claims.

[0107] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data for analysis, stored data, displayed data, etc.) involved in the present application are all information and data authorized by the user or fully authorized by all parties. Moreover, the processing of relevant data, such as collection, use, processing, transmission, provision, disclosure, and application, complies with relevant laws, regulations, and standards, takes necessary confidentiality measures, does not violate public order and good customs, and provides corresponding operation entrances for users to choose to authorize or refuse.

[0108] To address the above problems, the present application provides a method for controlling the charging of an unmanned aerial vehicle (UAV) battery. After detecting a UAV charging instruction, the charging and discharging state parameters of the UAV battery to be charged are obtained in real time. A state-of-charge determination model combining a convolutional neural network and a recurrent neural network is used to learn and analyze the charging and discharging state parameters of the UAV battery, so as to achieve a high-precision estimation of the state of charge of the UAV battery. Based on the estimated state-of-charge value, the performance of the UAV battery is evaluated to improve the accuracy of the UAV battery evaluation result.

[0109] The technical solution of the present application and how the technical solution of the present application solves the above technical problems will be described in detail below with specific embodiments. These several specific embodiments below can be combined with each other, and the same or similar concepts or processes may not be repeated in some embodiments. The embodiments of the present application will be described below with reference to the accompanying drawings.

[0110] Figure 1 Flow diagram of the method for controlling the charging of the UAV battery provided by the present application Figure 1 , as Figure 1 shown, the method includes:

[0111] S101. In response to detecting a UAV battery charging instruction, obtain the charging and discharging state parameters of the UAV battery to be charged.

[0112] In this step, it can be understood that when a drone battery charging instruction is detected, the initial stage of the charging process will be entered, that is, the charge and discharge state parameters of the drone battery to be charged are obtained. The purpose of this step is to evaluate the current operating state of the drone battery and provide data support for subsequent charging operations.

[0113] In one implementation, to collect the charge and discharge state parameters of the drone battery, the drone battery needs to be removed from the drone and connected to a measuring device. The battery model identification module built into the measuring device can detect the drone battery model and the number of battery cycles, and can achieve plug-and-measure and plug-and-charge of the battery. The sensors built into the measuring device will collect the charge and discharge state parameters of the drone battery in real time, including voltage, internal resistance, current, temperature, etc. Among them, the voltage is collected by the direct voltage measurement method, the internal resistance is collected by the Kelvin measurement method, and the current is collected by the direct current measurement method. The charge and discharge state parameters collected by this invasive method are more in line with the actual situation, that is, high accuracy, but this method requires physical disassembly of the battery, which may damage the battery.

[0114] In another implementation, the charge and discharge state parameters of the drone battery can be directly read through the communication interface of the battery management system built into the drone battery. This method does not require physical disassembly of the battery and can avoid damage to the battery caused by physical disassembly of the battery.

[0115] S102. Input the charge and discharge state parameters into the state of charge determination model, determine the state of charge of the drone battery according to the charge and discharge state parameters, and obtain the estimated value of the state of charge of the drone battery. The state of charge determination model is determined by a convolutional neural network and a recurrent neural network.

[0116] Input the charge and discharge state parameters obtained in S101 into the state of charge determination model. The state of charge determination model is a model that combines a convolutional neural network and a recurrent neural network. The state of charge determination model will determine the state of charge of the drone battery according to the charge and discharge state parameters, so as to obtain the estimated value of the state of charge of the drone battery.

[0117] S103. Determine the performance of the drone battery according to the estimated value of the state of charge.

[0118] In this step, it can be understood that the performance of the drone battery is determined according to the estimated value of the state of charge obtained in step S102.

[0119] Exemplarily, according to the estimated state of charge, the performance of the UAV battery is determined, including: determining whether the estimated state of charge is greater than or equal to the estimated threshold; when the estimated state of charge is greater than or equal to the estimated threshold, determining that the performance of the UAV battery is qualified; when the estimated state of charge is less than the estimated threshold, determining that the performance of the UAV battery is unqualified. Among them, the estimated threshold can be set according to actual needs. For example, the estimated threshold is set to 30%.

[0120] When the performance of the UAV battery is determined to be qualified, step S104 is executed.

[0121] S104: When the performance of the UAV battery is determined to be qualified, control the charging device to perform a charging operation on the UAV battery.

[0122] In this step, it can be understood that a charging operation is performed on the UAV battery determined to have qualified performance. Specifically, control the charging device to perform a charging operation on the UAV battery. Further, the specific implementation manner of controlling the charging device to perform a charging operation on the UAV battery can be selected according to the implementation requirements, and the embodiments of the present application do not limit this.

[0123] Exemplarily, when the performance of the UAV battery is determined to be qualified, controlling the charging device to perform a charging operation on the UAV battery includes: when the performance of the UAV battery is determined to be qualified, adjusting the charging parameters of the charging device to ideal charging parameters matching the UAV battery to complete the charging operation on the UAV battery, where the charging parameters include charging current, charging voltage, and charging rate; or, when the performance of the UAV battery is determined to be qualified and the estimated state of charge of the UAV battery is less than or equal to the constant current charging threshold, controlling the charging device to charge the UAV battery in a constant current mode; when the performance of the UAV battery is determined to be qualified and the state of charge of the UAV battery is greater than the constant current threshold, controlling the charging device to charge the UAV battery in a constant voltage mode. It should be noted that the constant current charging threshold can be set according to the actual situation.

[0124] This example provides different specific implementation manners of controlling the charging device to perform a charging operation on the UAV battery, which can improve the flexibility of the UAV battery charging control method and meet the needs of different users. Further, this example improves the charging efficiency of the charging device for the UAV battery by adjusting the charging parameters of the charging device or switching the charging device mode.

[0125] In the embodiment of the present application, after detecting a drone charging instruction, the charge and discharge state parameters of the drone battery to be charged are obtained in real time, without waiting for a complete charge and discharge time, which is beneficial to reducing the time cost; a state of charge determination model combining a convolutional neural network and a recurrent neural network is used to learn and analyze the charge and discharge state parameters of the drone battery, so as to achieve a high-precision estimation of the state of charge of the drone battery; the performance of the drone battery is evaluated based on the estimated value of the state of charge, which can improve the accuracy of the evaluation result of the drone battery; when the performance of the drone battery is determined to be qualified, the charging device is controlled to perform a charging operation on the drone battery, which can comprehensively evaluate the battery state before the charging operation, ensuring that the charging operation is only performed on the battery with qualified performance. This design effectively avoids charging the battery with poor performance or potential faults, thereby reducing safety hazards such as battery overheating and fire, and at the same time avoiding waste of charging resources.

[0126] On the basis of the above embodiment, the state of charge of the drone battery is determined according to the charge and discharge state parameters, and an estimated value of the state of charge of the drone battery is obtained, including: inputting the charge and discharge state parameters into a convolutional neural network to extract local features between the charge and discharge state parameters; inputting the local features into a recurrent neural network, and in the recurrent neural network, capturing the long-term dependence relationship of the local features in the time series, and determining the estimated value of the state of charge of the drone battery according to the long-term dependence relationship.

[0127] In this embodiment, it can be understood that, first, the charge and discharge state parameters of the drone battery are input into a convolutional neural network, and the convolutional neural network extracts the local correlation and pattern between these parameters. Next, the local features extracted by the convolutional neural network are input into a recurrent neural network, and the recurrent neural network captures the long-term dependence relationship of these local features in the time dimension. By combining the local features and the long-term dependence relationship of the time series, the recurrent neural network can finally accurately estimate the state of charge of the battery.

[0128] Furthermore, the convolutional neural network includes an input layer, a one-dimensional convolutional layer, and a one-dimensional pooling layer. The output of the input layer serves as the input of the connected one-dimensional convolutional layer, and the output of the one-dimensional convolutional layer serves as the output of the connected one-dimensional pooling layer. The recurrent neural network includes a long short-term memory network, a Dropout layer, and a fully connected layer. The output of the one-dimensional pooling layer serves as the input of the long short-term memory network, the output of the long short-term memory network serves as the input of the Dropout layer, and the output of the Dropout layer serves as the input of the fully connected layer. Wherein: the input layer is used to receive charge and discharge state parameters; the one-dimensional convolutional layer is used to extract local features in the charge and discharge state parameters; the one-dimensional pooling layer is used to reduce the dimension of the local features; the long short-term memory network is used to capture the long-term dependencies of the local features in the time series; the Dropout layer is used to randomly deactivate some neurons in the output of the long short-term memory network during the training stage of the state of charge determination model; the fully connected layer is used to determine the estimated value of the state of charge of the UAV battery through linear transformation according to the output in the Dropout layer.

[0129] In the embodiment of the present application, through the convolutional neural network and the recurrent neural network, not only can the static associations of battery parameters be understood, but also the changing rules of the battery during the charge and discharge process can be dynamically analyzed, so as to greatly improve the accuracy of state of charge estimation and provide technical support for the efficient management of UAV batteries.

[0130] Based on the above embodiments, the hyperparameters of the state of charge determination model can be determined by the following method: obtaining the velocities and positions of a particle swarm composed of multiple particles, where the position of each particle represents a set of hyperparameter combinations in the state of charge determination model, and the hyperparameters at least include the learning rate, the number of hidden layer units, and the maximum number of training iterations, and the velocity of each particle represents the direction and step size of the particle moving in the search space; based on the positions of the particle swarm, training the state of charge determination model and calculating the fitness value corresponding to each set of hyperparameters to determine the individual optimal position and the global optimal position in the particle swarm; according to the individual optimal position and the global optimal position, applying the velocity update equation in the particle swarm optimization algorithm to update the velocities of the particle swarm, and adjusting the positions of the particle swarm based on the updated velocities; based on the positions of the updated particle swarm, retraining the state of charge determination model and updating the individual optimal position and the global optimal position in the particle swarm; determining whether the current iteration number of the state of charge determination model reaches the maximum number of training iterations; if the current iteration number reaches the maximum number of training iterations, determining the hyperparameter combination corresponding to the global optimal position as the hyperparameters of the state of charge determination model; if the current iteration number does not reach the training iteration number, returning to train the state of charge determination model, updating the individual optimal position and the global optimal position in the particle swarm, and continuing iterative optimization.

[0131] In this embodiment, it can be understood that the hyperparameters of the state of charge determination model are determined by the particle algorithm. By simulating swarm intelligence behavior, the particle algorithm can perform efficient global search in a complex parameter space, avoid falling into local optima, and thus find a better combination of hyperparameters to improve the overall performance of the state of charge determination model. Secondly, the particle algorithm has a fast convergence speed. Through information sharing and cooperation among particles, it can quickly approach the optimal solution, significantly reduce the time cost of hyperparameter tuning, and improve the development efficiency of the state of charge determination model.

[0132] Furthermore, the above-mentioned unmanned aerial vehicle battery charging control method further includes: when it is determined that the performance of the unmanned aerial vehicle battery is unqualified, displaying the estimated value of the state of charge and a prompt message indicating that the battery performance is unqualified on the visualization large screen; and / or, when it is determined that the performance of the unmanned aerial vehicle battery is unqualified, controlling the buzzer to emit a sound warning message; and / or, when it is determined that the performance of the unmanned aerial vehicle battery is unqualified, controlling the running lights to emit a visual warning message.

[0133] This example provides a multi-modal warning mechanism when the performance of the unmanned aerial vehicle battery is unqualified. First, by displaying the estimated value of the state of charge and a prompt message indicating that the battery performance is unqualified on the visualization large screen, it can intuitively transmit the battery state information to the operator, improving the operability and user experience. Secondly, by controlling the buzzer to emit a sound warning message, it can timely remind relevant personnel in a complex environment. For example, when the light is insufficient or the operator does not directly visualize the large screen, a sound warning is emitted in time to remind relevant personnel. In addition, a visual warning message can also be emitted by controlling the running lights, further enriching the warning method. This method of emitting a visual warning message by controlling the running lights is especially suitable for noisy environments or scenarios that require multi-sensory collaboration, improving the coverage and effectiveness of the warning.

[0134] The multi-modal warning mechanism provided by the embodiments of this application can not only meet the requirements in different scenarios, but also reflect the consideration of user experience.

[0135] Based on the above embodiments, the above drone battery charging control method further includes: performing data preprocessing operations on the charge and discharge state parameters, and the data preprocessing operations at least include one of the following: Kalman filtering processing, moving average processing, and data cleaning. This means that the collected charge and discharge state parameters are preprocessed to remove invalid data and improve the effectiveness of the charge and discharge state parameters. And the data preprocessing operations at least include one of the following: Kalman filtering processing, moving average processing, and data cleaning. Among them, Kalman filtering processing is a recursive optimal estimation algorithm, which is mainly used to extract the true state of the system from the observed data containing noise; moving average processing is a data smoothing method, also known as moving average processing, which mainly calculates the average value of several consecutive data points in the data sequence to eliminate random noise or short-term fluctuations in the data, so as to highlight the long-term trend or stable characteristics of the data; data cleaning refers to processing the original data to identify, correct, or delete incomplete, incorrect, inconsistent, or duplicate data to improve the quality and usability of the data.

[0136] Further, after determining the performance of the drone battery, it further includes: obtaining the model of the drone battery; determining the performance state change rule of the drone battery of the model based on the performance of the drone battery at different times; predicting the time when the performance of the drone battery of the model is unqualified according to the performance state change rule; determining the management measures for the drone battery based on the time of performance unqualified.

[0137] This embodiment provides a drone battery performance prediction measure, and its core process includes: first obtaining the model information of the drone battery as the basis for subsequent drone battery performance prediction analysis; then determining the performance state change rule of the battery of this model based on the performance data of the battery at different times; then using the prediction model to predict the battery performance state change rule according to the performance change rule, for example, using a long short-term memory network or a time series analysis network to predict the battery performance state change rule according to the performance change rule; finally, based on the prediction result, formulating and implementing corresponding management measures, such as replacing the drone battery in advance or sending an alarm message.

[0138] In summary, the embodiments of the present application realize the intelligent management of the drone battery by analyzing the performance data of the drone battery at different times, improve the safety and reliability of the drone operation, and at the same time reduce the operation cost and risk.

[0139] Next, an example will be given to illustrate how to use the drone battery charging control method provided by the embodiments of the present application. Figure 2 The flow chart of the drone battery charging control method provided by the embodiments of the present application Figure 2 . As Figure 2 shown, the method includes the following steps:

[0140] 1. Connect the battery under test to the measuring device through the input module. The battery model identification module built into the measuring device will determine the battery model. The input module is mainly a data transmission interface, which can be connected to various drone batteries on the market.

[0141] 2. Collect the state parameters of the battery through sensors, including voltage, internal resistance, current, temperature, etc.

[0142] 3. Preprocess the data according to the collected voltage, internal resistance, current, and temperature data. Specifically, remove the noise signal through the Kalman filter or the moving average algorithm.

[0143] 4. Input the preprocessed data and use the State of Charge (SOC) prediction model based on the Particle Swarm Optimization (PSO)-Convolutional Neural Network (CNN)-Long Short-Term Memory (LSTM) to achieve real-time estimation of the SOC of the drone battery.

[0144] 5. Determine the performance of the drone battery according to the SOC estimation value. Transmit the state parameters and the corresponding battery model of the qualified drone battery to the charging module. The charging module will adjust the charging parameters according to the received data, such as the current magnitude, voltage level, and charging rate. The charging module is mainly a charging controller.

[0145] 6. Display the SOC estimation value and the performance evaluation result of the unqualified drone battery on the user interface, and use a buzzer, a running water lamp, etc. to alarm the unqualified battery.

[0146] It should be noted that if the performance of the drone battery is qualified, the performance evaluation result of the drone battery will also be displayed on the user interface.

[0147] Furthermore, Figure 3 This is a schematic diagram of the principle of the SOC prediction model based on PSO-CNN-LSTM provided by the embodiment of the present application. As Figure 3 shown, the construction principle of this prediction model is as follows:

[0148] (1) Select the dataset: Select the batteries of the set model of the unmanned aerial vehicle as the experimental object. Take the charge and discharge data of different batteries of this model for 10 days as the sample set, with a sampling interval of 30 s, and a total of a set number of groups of data. Each group of data covers information such as internal resistance, voltage, current, and temperature during the charge and discharge process of the battery. Take the internal resistance, voltage, current, and temperature of the dataset as the input of the model. Randomly divide the dataset according to the ratio of 8:2, where 80% is used to train the network and the remaining 20% is used for testing;

[0149] (2) Preprocess the battery data to solve problems such as null values, missing values, and outliers in the dataset for use as the model input;

[0150] (3) Feed the preprocessed data into the PSO algorithm to optimize the parameters, mainly determining three key hyperparameters: the learning rate, the number of neurons in the hidden layer, and the number of training times of the LSTM model;

[0151] (4) After determining the optimal hyperparameters, feed the preprocessed battery data into the CNN model for convolution and pooling operations to extract key features;

[0152] (5) Feed the denoised feature data into the LSTM model;

[0153] (6) Introduce a Dropout layer to reduce the risk of model overfitting;

[0154] (7) Convert the multi-dimensional input data into a one-dimensional output through a fully connected layer to obtain the prediction result;

[0155] (8) The LSTM model is trained on the training dataset using the optimal hyperparameters, with the optimizer being Adam and the loss function being the mean squared error. Subsequently, it is tested on the test dataset, and its performance is comprehensively evaluated through key performance indicators such as EMAE, EMSE, and R2.

[0156] In summary, for the unmanned aerial vehicle battery charging control method provided in the embodiments of this application, by comparing the internal resistance and discharge performance of the battery and using multiple methods such as the measurement method and the voltage method, the measurement accuracy is significantly improved. This method features rapid detection, without consuming the complete charge and discharge cycle time, thus greatly enhancing the detection efficiency. In addition, this method has a low cost. Through low-cost integrated circuits and special combination designs, high-cost-performance battery performance testing is achieved, fully meeting the current requirements for battery testing. At the same time, this method is simple to operate and can be easily mastered in just three steps, greatly reducing the usage threshold and being applicable to a wide range of application scenarios.

[0157] Furthermore, the drone battery charging control method provided by the embodiments of the present application realizes rapid performance testing of drone batteries through the resistance, voltage characteristics, and design principles of drone batteries, and through special integration and wiring modes such as resistance and voltage. It is applicable to existing common drone models, enabling operators to conveniently, quickly, and effectively test the performance of drone batteries, avoiding drone crashes caused by problems such as battery loss or battery bulge, and reducing equipment damage and damage to the drones themselves due to the above problems.

[0158] The following is an embodiment of the device of the present application, which can be used to execute the method embodiment of the present application. For details not disclosed in the device embodiment of the present application, please refer to the method embodiment of the present application.

[0159] Figure 4 It is a schematic structural diagram of the drone battery charging control device provided by the present application, as Figure 4 shown. The drone battery charging control device 400 provided in this embodiment includes:

[0160] An acquisition module 401, configured to obtain charge and discharge state parameters of the drone battery to be charged in response to detecting a drone battery charging instruction;

[0161] A first determination module 402, configured to input the charge and discharge state parameters into a state of charge determination model, determine the state of charge of the drone battery according to the charge and discharge state parameters, and obtain an estimated value of the state of charge of the drone battery. The state of charge determination model is determined by a convolutional neural network and a recurrent neural network;

[0162] A second determination module 403, configured to determine the performance of the drone battery according to the estimated state of charge value;

[0163] A charging module 404, configured to control a charging device to perform a charging operation on the drone battery when the performance of the drone battery is determined to be qualified.

[0164] In a possible implementation manner, the charging module 404 is specifically configured to:

[0165] When the performance of the drone battery is determined to be qualified, adjust the charging parameters of the charging device to ideal charging parameters matching the drone battery, and complete the charging operation on the drone battery, where the charging parameters include charging current, charging voltage, and charging rate;

[0166] Or,

[0167] When the performance of the UAV battery is determined to be qualified and the estimated value of the state of charge of the UAV battery is less than or equal to the constant current charging threshold, control the charging device to charge the UAV battery in the constant current mode; when the performance of the UAV battery is determined to be qualified and the state of charge of the UAV battery is greater than the constant current threshold, control the charging device to charge the UAV battery in the constant voltage mode.

[0168] In a possible implementation manner, the first determination module 402 is specifically configured to:

[0169] Input the charge and discharge state parameters into the convolutional neural network to extract the local features between the charge and discharge state parameters;

[0170] Input the local features into the recurrent neural network. In the recurrent neural network, capture the long-term dependencies of the local features in the time series, and determine the estimated value of the state of charge of the UAV battery according to the long-term dependencies.

[0171] In a possible implementation manner, the convolutional neural network includes an input layer, a one-dimensional convolutional layer, and a one-dimensional pooling layer. The output of the input layer serves as the input of the connected one-dimensional convolutional layer, and the output of the one-dimensional convolutional layer serves as the output of the connected one-dimensional pooling layer; the recurrent neural network includes a long short-term memory network, a Dropout layer, and a fully connected layer. The output of the one-dimensional pooling layer serves as the input of the long short-term memory network, the output of the long short-term memory network serves as the input of the Dropout layer, and the output of the Dropout layer serves as the input of the fully connected layer; where:

[0172] The input layer is used to receive the charge and discharge state parameters;

[0173] The one-dimensional convolutional layer is used to extract the local features in the charge and discharge state parameters;

[0174] The one-dimensional pooling layer is used to reduce the dimension of the local features;

[0175] The long short-term memory network is used to capture the long-term dependencies of the local features in the time series;

[0176] The Dropout layer is used to randomly deactivate some neurons in the output of the long short-term memory network during the training stage of the state of charge determination model;

[0177] The fully connected layer is used to determine the estimated value of the state of charge of the UAV battery through linear transformation according to the output in the Dropout layer.

[0178] In a possible implementation manner, the hyperparameters of the state of charge determination model can be determined by the following method:

[0179] Obtain the velocities and positions of a particle swarm composed of multiple particles, where the position of each particle represents a set of hyperparameter combinations in the charge state determination model, and the hyperparameters at least include the learning rate, the number of hidden layer units, and the maximum number of training iterations, and the velocity of each particle represents the direction and step size of the particle moving in the search space;

[0180] Based on the positions of the particle swarm, train the charge state determination model, and calculate the fitness value corresponding to each set of hyperparameters to determine the individual optimal position and the global optimal position in the particle swarm;

[0181] According to the individual optimal position and the global optimal position, apply the velocity update equation in the particle swarm optimization algorithm to update the velocities of the particle swarm, and adjust the positions of the particle swarm based on the updated velocities;

[0182] Based on the positions of the updated particle swarm, retrain the charge state determination model, and update the individual optimal position and the global optimal position in the particle swarm;

[0183] Judge whether the current iteration number of the charge state determination model reaches the maximum number of training iterations;

[0184] If the current iteration number reaches the maximum number of training iterations, determine the hyperparameter combination corresponding to the global optimal position as the hyperparameters of the charge state determination model;

[0185] If the current iteration number does not reach the number of training iterations, return to train the charge state determination model, update the individual optimal position and the global optimal position in the particle swarm, and continue iterative optimization.

[0186] In a possible implementation manner, the second determination module 403 is specifically configured to:

[0187] Judge whether the charge state estimated value is greater than or equal to the estimation threshold;

[0188] When the charge state estimated value is greater than or equal to the estimation threshold, determine that the performance of the UAV battery is qualified;

[0189] When the charge state estimated value is less than the estimation threshold, determine that the performance of the UAV battery is unqualified.

[0190] In a possible implementation manner, the UAV battery charging control further includes a processing module (not shown), and the processing module is specifically configured to:

[0191] When it is determined that the performance of the UAV battery is unqualified, display the charge state estimated value and a prompt message indicating that the battery performance is unqualified on the visualization large screen;

[0192] And / or,

[0193] When it is determined that the performance of the UAV battery is unqualified, control the buzzer to emit a sound warning message;

[0194] and / or

[0195] When it is determined that the performance of the UAV battery is unqualified, control the running lights to emit visual warning information.

[0196] In a possible implementation manner, the processing module is further configured to:

[0197] Perform data preprocessing operations on the charge and discharge state parameters, and the data preprocessing operations include at least one of the following: Kalman filtering processing, moving average processing, and data cleaning.

[0198] In a possible implementation manner, the processing module is further configured to:

[0199] Obtain the model of the UAV battery;

[0200] Based on the performance of the UAV battery at different times, determine the performance state change law of the UAV battery of the model;

[0201] According to the performance state change law, predict the moment when the performance of the UAV battery of the model is unqualified;

[0202] Based on the moment when the performance is unqualified, determine the management measures for the UAV battery.

[0203] The UAV battery charging control device provided in this embodiment can execute the method provided in the above method embodiment, and its implementation principle and technical effect are similar, which will not be elaborated here in this embodiment.

[0204] It should be noted that it should be understood that the division of each module of the above device is only a logical function division. In actual implementation, it can be fully or partially integrated into a physical entity, or physically separated. And these modules can all be implemented in the form of software called by a processing element; they can also all be implemented in the form of hardware; or some modules can be implemented in the form of software called by a processing element, and some modules can be implemented in the form of hardware. For example, the processing module can be a separately established processing element, or can be integrated in a certain chip of the above device. In addition, it can also be stored in the memory of the above device in the form of program code, and called and executed by a certain processing element of the above device to perform the functions of the above processing module. The implementation of other modules is similar. In addition, all or part of these modules can be integrated together or can be independently implemented. Here, the processing element can be an integrated circuit with signal processing capabilities. In the implementation process, each step of the above method or each of the above modules can be completed by the integrated logic circuit in the processor element or the instruction in the form of software.

[0205] For example, the above-mentioned modules may be one or more integrated circuits configured to implement the above methods, such as: one or more Application Specific Integrated Circuits (ASICs), or, one or more Digital Signal Processors (DSPs), or, one or more Field Programmable Gate Arrays (FPGAs), etc. For another example, when a certain module above is implemented in the form of a processing element scheduling program code, the processing element may be a general-purpose processor, such as a Central Processing Unit (CPU) or other processors that can call program code. For another example, these modules may be integrated together and implemented in the form of a System-On-a-Chip (SOC).

[0206] Figure 5 Schematic diagram of the structure of an electronic device provided by an embodiment of the present application. As Figure 5 shown, the electronic device 500 provided by the embodiment of the present application may include: a processor 501, and a memory 502 communicatively connected to the processor, wherein:

[0207] The memory stores computer-executable instructions;

[0208] The processor executes the computer-executable instructions stored in the memory to implement the method described in the foregoing method embodiments.

[0209] It should be understood that the processor 501 may be a Central Processing Unit (CPU), or may also be other general-purpose processors, Digital Signal Processors (DSPs), Application Specific Integrated Circuits (ASICs), etc. The general-purpose processor may be a microprocessor or the processor may also be any conventional processor, etc. The steps of the method disclosed in combination with the application may be directly embodied as being executed and completed by a hardware processor, or may be executed and completed by a combination of hardware and software modules in the processor. The memory 502 may include high-speed random access memory (Random Access Memory, RAM), and may also include non-volatile storage NVM (non-volatile memory), such as at least one disk memory, and may also be a USB flash drive, a mobile hard disk, a read-only memory, a magnetic disk or an optical disc, etc.

[0210] Optionally, the electronic device 500 may further include a communication interface 503. In specific implementation, if the communication interface 503, the memory 502, and the processor 501 are implemented independently, the communication interface 503, the memory 502, and the processor 501 may be interconnected through a bus and communicate with each other. The bus may be an Industry Standard Architecture (ISA) bus, a Peripheral Component Interconnect (PCI) bus, an Extended Industry Standard Architecture (EISA) bus, or the like. The bus may be divided into an address bus, a data bus, a control bus, etc., but it does not mean that there is only one bus or one type of bus.

[0211] Optionally, in specific implementation, if the communication interface 503, the memory 502, and the processor 501 are integrated on a chip, the communication interface 503, the memory 502, and the processor 501 may communicate through an internal interface.

[0212] The embodiment of the present application further provides a computer-readable storage medium, in which computer-executable instructions are stored, and when the computer-executable instructions are executed, they are used to implement the method described in any of the foregoing embodiments.

[0213] It can be understood that the computer-readable storage medium may be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as a Static Random Access Memory (SRAM), an Electrically Erasable Programmable Read Only Memory (EEPROM), an Erasable Programmable Read Only Memory (EPROM), a Programmable Read Only Memory (PROM), a Read Only Memory (ROM), a magnetic memory, a flash memory, a magnetic disk, or an optical disc. The readable storage medium may be any available medium that can be accessed by a general-purpose or special-purpose computer.

[0214] An exemplary computer-readable storage medium is coupled to a processor, enabling the processor to read information from and write information to the computer-readable storage medium. Of course, the computer-readable storage medium can also be a component of the processor. The processor and the computer-readable storage medium can be located in an ASIC. Of course, the processor and the computer-readable storage medium can also exist as discrete components in an electronic device.

[0215] The integrated modules implemented in the form of software functional modules as described above can be stored in a computer-readable storage medium. The software functional modules stored in a computer-readable storage medium include several instructions for causing an electronic device (which can be a personal computer, a server, or a network device, etc.) or a processor to execute some steps of the methods described in various embodiments of the present application.

[0216] The embodiments of the present application also provide a computer program product, including a computer program, which when executed implements the methods described in any of the foregoing embodiments.

[0217] It should be noted that, for the foregoing method embodiments, for the sake of simple description, they are all expressed as a series of action combinations. However, those skilled in the art should know that the present application is not limited by the described action sequence, because according to the present application, certain steps can be performed in other sequences or simultaneously. Secondly, those skilled in the art should also know that the embodiments described in the specification are all optional embodiments, and the actions and modules involved are not necessarily essential to the present application.

[0218] Furthermore, it should be noted that although the steps in the flowcharts are displayed in sequence according to the arrows, these steps are not necessarily executed in the order indicated by the arrows. Unless there is a clear description in this article, the execution of these steps has no strict order limit, and these steps can be executed in other orders. Moreover, at least a part of the steps in the flowchart can include multiple sub-steps or multiple stages. These sub-steps or stages are not necessarily executed at the same time, but can be executed at different times. The execution order of these sub-steps or stages is not necessarily sequential, but can be executed alternately or in turn with at least a part of other steps or sub-steps or stages of other steps.

[0219] In the above embodiments, the descriptions of the various embodiments have their own focuses. For the parts not detailed in a certain embodiment, reference can be made to the relevant descriptions of other embodiments. The technical features of the above embodiments can be combined arbitrarily. For the sake of concise description, not all possible combinations of the technical features in the above embodiments are described. However, as long as the combinations of these technical features do not conflict, they should all be considered as the scope described in this specification.

[0220] Those skilled in the art will readily conceive of other embodiments of the present application after considering the specification and practicing the invention disclosed herein. The present application is intended to cover any variations, uses, or adaptations of the present application, which follow the general principles of the present application and include known common general knowledge or conventional technical means in the technical field not disclosed in the present application. The specification and examples are only regarded as exemplary, and the true scope and spirit of the present application are pointed out by the following claims.

[0221] It should be understood that the present application is not limited to the exact structures described above and shown in the drawings, and various modifications and changes can be made without departing from its scope. The scope of the present application is only limited by the appended claims.

Claims

1. A method for controlling battery charging of an unmanned aerial vehicle, characterized in that: include: In response to detecting a drone battery charging instruction, obtaining charging and discharging state parameters of the drone battery to be charged; Inputting the charge and discharge state parameters into a state of charge determination model, determining the state of charge of the drone battery according to the charge and discharge state parameters, and obtaining an estimated value of the state of charge of the drone battery, wherein the state of charge determination model is determined by a convolutional neural network and a recurrent neural network; Determining the performance of the drone battery based on the estimated state of charge; When the performance of the drone battery is determined to be qualified, the charging device is controlled to perform a charging operation on the drone battery.

2. The method according to claim 1, characterized in that: When the performance of the drone battery is determined to be qualified, controlling the charging device to charge the drone battery includes: When the performance of the drone battery is determined to be qualified, the charging parameters of the charging device are adjusted to ideal charging parameters matching the drone battery, and the charging operation of the drone battery is completed, wherein the charging parameters include charging current, charging voltage and charging rate; or, When the performance of the drone battery is determined to be qualified and the estimated state of charge of the drone battery is less than or equal to the constant current charging threshold, the charging device is controlled to charge the drone battery in a constant current mode; when the performance of the drone battery is determined to be qualified and the state of charge of the drone battery is greater than the constant current threshold, the charging device is controlled to charge the drone battery in a constant voltage mode.

3. The method according to claim 2, characterized in that Determining the state of charge of the drone battery according to the charge and discharge state parameter to obtain an estimated value of the state of charge of the drone battery includes: Inputting the charge and discharge state parameters into the convolutional neural network to extract local features between the charge and discharge state parameters; The local features are input into the recurrent neural network, in which the long-term dependencies of the local features in the time series are captured, and based on the long-term dependencies, an estimated value of the state of charge of the drone battery is determined.

4. The method according to claim 3, characterized in that The convolutional neural network comprises an input layer, a one-dimensional convolutional layer and a one-dimensional pooling layer, the output of the input layer is used as the input of the connected one-dimensional convolutional layer, and the output of the one-dimensional convolutional layer is used as the output of the connected one-dimensional pooling layer; the recurrent neural network comprises a long short-term memory network, a Dropout layer and a fully connected layer, the output of the one-dimensional pooling layer is used as the input of the long short-term memory network, the output of the long short-term memory network is used as the input of the Dropout layer, and the output of the Dropout layer is used as the input of the fully connected layer; wherein: The input layer is used to receive the charging and discharging state parameters; The one-dimensional convolution layer is used to extract local features in the charge and discharge state parameters; The one-dimensional pooling layer is used to reduce the dimension of the local features; The long short-term memory network is used to capture the long-term dependency of the local features in the time series; The Dropout layer is used to randomly inactivate some neurons output by the long short-term memory network during the state of charge determination model training phase; The fully connected layer is used to determine the estimated value of the state of charge of the drone battery through linear transformation according to the output in the Dropout layer.

5. The method according to any one of claims 1 to 4, characterized in that The hyperparameters of the state of charge determination model can be determined by the following method: Acquire the speed and position of a particle group consisting of a plurality of particles, wherein the position of each particle represents a set of hyperparameter combinations in the charge state determination model, the hyperparameters at least including a learning rate, a number of hidden layer units, and a maximum number of training iterations, and the speed of each particle represents a direction and a step size of movement of the particle in the search space; Based on the position of the particle swarm, the charge state determination model is trained, and the fitness value corresponding to each group of hyperparameters is calculated to determine the individual optimal position and the global optimal position in the particle swarm; According to the individual optimal position and the global optimal position, applying a speed update equation in a particle swarm optimization algorithm, updating the speed of the particle swarm, and adjusting the position of the particle swarm based on the updated speed; Based on the updated position of the particle swarm, retrain the charge state determination model, and update the individual optimal position and the global optimal position in the particle swarm; Determine whether the current iteration number of the state of charge determination model reaches the maximum training iteration number; If the current number of iterations reaches the maximum number of training iterations, determining the hyperparameter combination corresponding to the global optimal position as the hyperparameter of the state of charge determination model; If the current number of iterations does not reach the number of training iterations, return to training the state of charge determination model, update the individual optimal position and the global optimal position in the particle swarm, and continue iterative optimization.

6. The method according to any one of claims 1 to 4, characterized in that Determining the performance of the drone battery according to the estimated state of charge value includes: Determining whether the state of charge estimation value is greater than or equal to an estimation threshold; When the charge state estimation value is greater than or equal to the estimation threshold, determining that the drone battery performance is qualified; When the charge state estimation value is less than the estimation threshold, it is determined that the drone battery performance is unqualified.

7. The method according to claim 6, characterized in that Also includes: When it is determined that the battery performance of the drone is unqualified, the estimated value of the state of charge and a prompt message indicating that the battery performance is unqualified are displayed on a large visual screen; and / or, When it is determined that the battery performance of the drone is not up to standard, the buzzer is controlled to emit a sound warning message; and / or, When it is determined that the battery performance of the drone is not up to standard, the running light is controlled to issue a visual warning message.

8. The method according to any one of claims 1 to 4, characterized in that Also includes: A data preprocessing operation is performed on the charge and discharge state parameters, wherein the data preprocessing operation includes at least one of the following: Kalman filter processing, sliding mean processing, and data cleaning.

9. The method according to any one of claims 1 to 4, characterized in that Also includes: Obtain the model of the drone battery; Based on the performance of the drone battery at different times, determining the performance state change rule of the drone battery of the model; According to the performance status change law, predict the time when the performance of the battery of the drone of the model fails to meet the requirements; Based on the time when the performance fails, management measures for the drone battery are determined.

10. A drone battery charging control device, characterized in that: include: An acquisition module, used for acquiring charging and discharging state parameters of the battery of the drone to be charged in response to detecting a drone battery charging instruction; A first determination module is used to input the charge and discharge state parameters into a charge state determination model, determine the charge state of the drone battery according to the charge and discharge state parameters, and obtain an estimated value of the charge state of the drone battery, wherein the charge state determination model is determined by a convolutional neural network and a recurrent neural network; A second determination module is used to determine the performance of the drone battery according to the estimated state of charge value; The charging module is used to control the charging device to charge the battery of the drone when the performance of the drone battery is determined to be qualified.

Images