Broadcast remote identification adaptive transmission method and system for unmanned aerial vehicle battery status

Abstract

The application discloses a broadcasting remote identification adaptive transmission method and system for a battery state of a UAV, and relates to the technical field of UAV communication and battery management. The method collects and standardizes battery state information, combines a battery risk evaluation value, dynamically selects a data item set to be sent according to a preset risk mapping relationship, and adaptively determines a sending mode as an original mode or a compression mode according to a communication bandwidth and the battery state. In the compression mode, differential data is generated and compressed by using a pseudo-random projection rule, thereby effectively reducing the amount of transmission data. Delayed disclosure authentication information is generated to ensure the security and integrity of data transmission. An extension flag in a broadcasting remote identification enhanced frame is used to indicate the existence of the data item to be sent in a data content field, a receiving end analyzes the battery state information according to the extension flag, and differential data is recovered according to the compressed representation in the compression mode, thereby improving the transmission efficiency of the battery state information.

Description

Technical Field

[0001] This invention relates to the field of unmanned aerial vehicle (UAV) communication and battery management technology, and in particular to a broadcast-based remote identification and adaptive transmission method and system for UAV battery status. Background Technology

[0002] With the widespread application of drone technology, drones are playing an increasingly important role in various missions. Monitoring the battery status of drones during flight is crucial for extending their operating time, ensuring safety, and improving performance. Traditional drone battery management systems (BMS) assess battery health by periodically reading parameters such as battery voltage, current, and temperature. However, with the increasing complexity of drone applications, especially in long-distance flights and extended operations, real-time monitoring and transmission of battery status information are necessary. Frequent data transmission not only increases the communication burden but also leads to additional battery consumption, especially under unstable signal or bandwidth-limited conditions, resulting in higher transmission latency and packet loss rates, severely impacting drone operating efficiency. Drones typically need to monitor multiple battery-related indicators, such as voltage, current, temperature, and state of charge (SOC). However, frequent transmission of battery status information can lead to wasted transmission bandwidth, especially when battery status changes are not significant, as redundant information increases the system burden. Drone battery management systems often rely on static parameters, such as battery voltage and temperature, when monitoring and judging battery health, lacking dynamic change detection. This leads to a failure to take timely and effective measures when battery failures occur, further exacerbating the risks for drones in long-distance missions.

[0003] Currently, existing drone battery status monitoring typically relies on dedicated telemetry links or cellular links for transmission, increasing hardware and power consumption. Furthermore, link interruptions can lead to critical fault alarms becoming unreachable, failing to adequately address these issues, especially under conditions of limited wireless communication bandwidth and battery resources. Therefore, effectively compressing battery status information and optimizing data transmission through adaptive transmission strategies—which can not only improve communication efficiency but also reduce energy consumption—is a core challenge in enhancing drone battery management systems. Summary of the Invention

[0004] To address the above problems, this invention proposes a broadcast-based remote identification and adaptive transmission method and system for UAV battery status, aiming to improve communication efficiency, reduce energy consumption, and enhance fault detection capabilities.

[0005] The present invention achieves the above objectives through the following technical solutions:

[0006] A broadcast-based adaptive transmission method for remote identification of drone battery status, the method comprising:

[0007] Collect and standardize the battery status information output by the battery management system, and obtain the battery risk assessment value based on the fault status and time sequence changes in the battery status information according to the preset risk mapping relationship.

[0008] Under the single-frame payload budget constraint, a set of data items to be sent is selected from the battery status information based on the battery risk assessment value and encoded to generate a payload to be sent. The transmission mode is determined to be either the original mode or the compressed mode according to the preset switching conditions. In the compressed mode, a set of differential data items is determined from the set of data items to be sent. For each differential data item in the set of differential data items, differential data relative to the previous time is generated. A compressed representation of the differential data is generated according to the pseudo-random projection rule determined by the broadcast slot index and written into the payload to be sent.

[0009] Generate delayed disclosure authentication information for the payload to be sent, the delayed disclosure authentication information including a message authentication code and an authentication key disclosed in a preset delay time slot;

[0010] The payload to be transmitted is encapsulated into a broadcast remote identification enhancement frame and broadcast. The data flag field of the broadcast remote identification enhancement frame includes an extended flag bit, which is used to indicate the existence of the value of the corresponding data item in the set of data items to be transmitted in the data content field.

[0011] The receiving end receives the broadcast remote identification enhancement frame, verifies the message authentication code after obtaining the authentication key, and obtains the battery status information by parsing the extended flag bit when the verification is successful. When the broadcast remote identification enhancement frame contains the compressed representation, the differential data is restored and the corresponding data item is reconstructed according to the pseudo-random projection rule, and the battery alarm or status result is output.

[0012] A further improvement of the present invention is that the battery status information includes at least: battery pack voltage, individual cell voltage, battery pack current, state of charge (SOC), state of health (SOH), temperature, number of cycles, remaining capacity, and fault bitmap.

[0013] The preset risk mapping relationship is specifically a parameterized self-calibrated basic risk mapping, including:

[0014] Calculate the basic risk value based on standardized battery state information. : ;

[0015] In the formula, For the first Each standardized state data item takes a value. A standardized set of state data items for risk assessment. For risk mapping parameters, A function to normalize the result to [0,1];

[0016] A new fault occurrence event is used as a self-calibration trigger signal. The new fault occurrence event is obtained by comparing the fault bit map at the current time with that at the previous time. When there is a fault bit that changes from 0 to 1, it is determined that a new fault occurrence event has occurred.

[0017] When a new fault event is detected, the risk mapping parameters are updated online using the gradient descent algorithm based on the boundary constraint loss function to obtain the updated base risk value. ;

[0018] Where the boundary constraint loss function satisfy:

[0019] ;

[0020] In the formula, This is an indicator of a new fault occurrence; it is set to 1 when a new fault occurs, and 0 otherwise. This is the lower bound of the fault trigger risk. This represents the upper bound of the risk of no failure. The regularization coefficient is . For the reason The set of parameters that make up the composition.

[0021] A further improvement of the present invention is that the temporal variation of the battery state information is used to calculate the variation intensity term, and the variation intensity term is obtained through a fault intensity-driven dual-time-scale adaptive statistic, including:

[0022] The adaptive update coefficient is determined based on the number of bits set in the fault bitmap, satisfying the following:

[0023] ;

[0024] In the formula, For short-term update coefficients, For long-term update coefficients, preset ; This is the sensitivity coefficient; This represents the number of fault bits that are set in the fault bitmap at the current moment. This represents the total number of bits in the fault bitmap. As a preset boundary, and satisfying ;

[0025] Calculate the short-time mean for each standardized state data item. Compared with long-term average During the first sampling, the short-term and long-term means are initialized to the corresponding standardized state data item values; in subsequent sampling, they are recursively updated using the following formula:

[0026] ;

[0027] Calculate the intensity of change term : ,in For safety weights, satisfy , The normalized scale threshold;

[0028] Determine the validation lag factor : ,in The preset number of delay slots; The broadcast period for enhanced frames in broadcast-based remote identification; This indicates the verification waiting delay. Preset reference lag time;

[0029] Based on fault strength term , Intensity of change Validate the lag factor and the updated base risk value Calculate the battery risk assessment value: ;

[0030] In the formula, This is the battery risk assessment value, and ; These are preset non-negative coefficients used to adjust the contributions of the fault intensity term, the change intensity term, and the verification lag factor to the battery risk assessment value, respectively.

[0031] A further improvement of the present invention is that the preset switching condition is specifically: the battery risk assessment value reaches a preset risk threshold or a fault bit is set in the fault bit map;

[0032] When the preset switching conditions are met, the transmission mode is determined to be the original mode; otherwise, the transmission mode is determined to be the compressed mode. The payload to be transmitted in the compressed mode includes the original value of the fault bitmap and the compressed representation.

[0033] The preset risk thresholds include a first risk threshold for entering the original mode and a second risk threshold for returning to the compressed mode, and the first risk threshold is greater than the second risk threshold. When the battery risk assessment value is greater than or equal to the first risk threshold, the original mode is entered and a preset holding window duration is maintained. After the holding window duration ends, the compressed mode is switched only when the battery risk assessment value is less than or equal to the second risk threshold and there is no set fault bit in the fault bitmap.

[0034] A further improvement of the present invention is that, in compression mode, determining the differential data item set from the set of data items to be sent includes:

[0035] Remove the data items corresponding to the fault bitmap from the set of data items to be sent to obtain a candidate set;

[0036] Calculate the normalized difference score for each candidate data item in the candidate set. ,satisfy: ;

[0037] in, For the current moment The standardized values ​​of each candidate data item. For the previous moment The standardized values ​​of each candidate data item. For the first The preset normalization scale threshold corresponding to each candidate data item;

[0038] The candidate set is sorted from largest to smallest according to the normalized difference score, and candidate data items are selected and added to the difference data item set in the sorted order until further additions would cause the byte length of the compressed representation corresponding to the difference data item set to exceed the byte budget reserved for the compressed representation in the compression mode. The selection stops when further additions are made. The byte budget reserved for the compressed representation is equal to the single frame payload budget minus the fixed field byte length and the remaining byte length after deducting the data item byte length that needs to be written to the payload to be sent with the original value in the compression mode.

[0039] A further improvement of the present invention is that the pseudo-random projection rule is generated by a pseudo-random number generator, the seed of the pseudo-random number generator is obtained by hashing the broadcast time slot index, and the input of the hashing operation further includes the UAV identifier;

[0040] The identification information used to generate the seed is written into the payload to be sent and overwritten by the message authentication code, so that after the receiver passes the verification, it generates a consistent pseudo-random projection rule based on the same broadcast slot index and UAV identifier to recover the differential data.

[0041] A further improvement of the present invention is that the set of differential data items is divided into a critical differential data subset and a non-critical differential data subset, and different measurement resources are allocated to the two subsets to generate the compressed representation, wherein:

[0042] The critical differential data subset includes at least one or more of the battery pack voltage differential, battery pack current differential, and temperature differential, and the non-critical differential data subset consists of data items from the differential data item set other than the critical differential data subset.

[0043] Under the constraint of the byte budget reserved in the compressed representation, the total number of available measurements for generating the measurement value sequence in this frame is determined based on the number of encoded bytes of a single measurement value and the number of bytes of the reconstruction parameter field; wherein the total number of available measurements is the maximum number of measurements that ensures the byte length of the compressed representation does not exceed the byte budget reserved in the compressed representation, and the total number of available measurements is obtained by subtracting the number of bytes of the reconstruction parameter field from the byte budget reserved in the compressed representation, dividing by the number of encoded bytes of a single measurement value, and taking the integer part;

[0044] Based on the battery risk assessment value, the proportion of key differential data subsets in the differential data item set, and the preset stratification coefficient, the number of key measurements is determined and the remaining number of measurements is determined as non-key measurements. The number of key measurements does not decrease as the battery risk assessment value increases, does not decrease as the proportion of key differential data subsets increases, and the number of key measurements is not greater than the total number of available measurements.

[0045] Key measurement matrices corresponding to key measurement numbers and non-key measurement matrices corresponding to non-key measurement numbers are generated according to the pseudo-random projection rules. A linear projection is performed on the difference data vectors corresponding to the key difference data subsets to obtain a key measurement value sequence, and a linear projection is performed on the difference data vectors corresponding to the non-key difference data subsets to obtain a non-key measurement value sequence. The reconstruction parameter field, the key measurement value sequence, and the non-key measurement value sequence together constitute the compressed representation. The reconstruction parameter field includes at least the key and non-key measurement numbers and is used to instruct the receiving end to generate a consistent measurement matrix according to the corresponding measurement numbers to recover the difference data.

[0046] A further improvement of the present invention is that the broadcast-type remote identification enhancement frame includes a data type field, a version number field, a data length field, a data flag field, and a data content field; wherein the data flag field is extended from a basic length to an extended length, the basic length being 3 bytes and the extended length being 5 bytes, and the newly added 4th and 5th bytes constitute the extended flag bit.

[0047] A further improvement of the present invention is that, when broadcast transmission is interrupted, the payload to be transmitted and the authentication context corresponding to the delayed disclosure authentication information are cached; after broadcast is resumed, the broadcast-type remote identification enhancement frame is retransmitted, and the correspondence between the authentication key and the message authentication code is maintained, so that the receiving end can parse and output the battery status information after obtaining the authentication key and completing the verification; and when the interruption duration exceeds a preset duration threshold, only the payload to be transmitted containing a fault bitmap or a battery risk assessment value not less than a preset cache threshold is cached.

[0048] A broadcast-based remote identification adaptive transmission system for drone battery status, the system comprising:

[0049] The battery status acquisition and standardization module is used to acquire battery status information output by the battery management system and perform standardization processing.

[0050] The risk assessment module is used to obtain a battery risk assessment value based on the fault status and time sequence changes in the battery status information, according to a preset risk mapping relationship.

[0051] The payload budget and data item selection module is used to select a set of data items to be sent from the battery status information based on the battery risk assessment value under the single-frame payload budget constraint.

[0052] The payload encoding module is used to encode the set of data items to be sent and generate the payload to be sent.

[0053] The mode switching control module is used to determine the transmission mode as either raw mode or compressed mode according to preset switching conditions.

[0054] The differential generation module is used to determine a differential data item set from the set of data items to be sent when the transmission mode is compression mode, and to generate differential data relative to the previous time for each differential data item in the differential data item set.

[0055] The pseudo-random projection compression module is used to generate a compressed representation of the differential data according to the pseudo-random projection rules determined by the broadcast time slot index, and write the compressed representation into the payload to be transmitted;

[0056] The delayed disclosure authentication module is used to generate delayed disclosure authentication information for the payload to be sent. The delayed disclosure authentication information includes a message authentication code and an authentication key disclosed in a preset delay time slot.

[0057] The enhanced frame encapsulation and broadcast transmission module is used to encapsulate the payload to be transmitted into a broadcast remote identification enhanced frame and broadcast it, wherein the data flag field of the broadcast remote identification enhanced frame includes an extended flag bit, which is used to indicate the existence of the value of the corresponding data item in the set of data items to be transmitted in the data content field.

[0058] The receiving verification and parsing module is used to receive the broadcast remote identification enhancement frame, verify the message authentication code after obtaining the authentication key, and parse the battery status information according to the extended flag bit when the verification is successful.

[0059] The differential recovery and reconstruction module is used to recover differential data and reconstruct corresponding data items according to the pseudo-random projection rule when the broadcast remote identification enhancement frame contains the compressed representation;

[0060] The alarm and status output module is used to output battery alarms or status results based on the battery status information obtained from parsing or reconstruction.

[0061] The beneficial effects of this invention are as follows: By assessing the risk of fault states and time-series changes in battery status information, the set of data items to be sent can be dynamically selected. This adaptive selection based on risk assessment values ​​avoids transmitting redundant information when battery status changes are small or stable, thereby reducing communication bandwidth usage and improving transmission efficiency. In compression mode, the sending end first generates differential data by comparing the battery status information at the current moment with that at the previous moment. For the differential data, a pseudo-random projection rule is used for compression. This compression method can significantly reduce data redundancy while maintaining information integrity, especially under conditions of limited wireless bandwidth, effectively reducing the amount of data and thus improving data transmission efficiency. By generating delayed disclosure authentication information and disclosing the authentication key in a preset time slot, data security and reliability during transmission are ensured. After obtaining the authentication key, the receiving end can verify the message authentication code to ensure that the received data has not been tampered with, enhancing system security. By setting an extended flag bit in the data flag field, the receiving end can quickly parse and identify the existence of each data item in the data content field. This mechanism not only simplifies the parsing process but also reduces the computational burden on the receiving end, further improving the system's response speed. Attached Figure Description

[0062] To more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort. Wherein:

[0063] Figure 1 This is a flowchart of the method of the present invention;

[0064] Figure 2 This is a schematic diagram of the modular structure of the system in an embodiment of the present invention. Detailed Implementation

[0065] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. All other embodiments obtained by those skilled in the art based on the described embodiments of the present invention are within the scope of protection of the present invention.

[0066] like Figure 1 As shown, this is an embodiment of the present invention, which provides a broadcast-based remote identification adaptive transmission method for UAV battery status, including the following steps:

[0067] S1: Collect and standardize the battery status information output by the battery management system. Based on the fault status and time sequence changes in the battery status information, obtain the battery risk assessment value according to the preset risk mapping relationship.

[0068] In one specific embodiment, step S1 includes:

[0069] S11: The drone battery management system (BMS) periodically outputs battery status information, including at least: battery pack voltage, individual cell voltage, battery pack current, state of charge (SOC), state of health (SOH), temperature, cycle count, remaining capacity, and fault bitmap.

[0070] The fault bitmap is used to represent the fault state. Each bit in the fault bitmap corresponds to a preset fault item (such as overcharge, over-discharge, overcurrent, high temperature, low temperature, sensor failure, cell failure, etc.). When a bit is 1, it means that the fault item has occurred.

[0071] The preset sampling period is (e.g., 100ms), with the sampling sequence number representing the time. To ensure that state data items of different dimensions can be compared and used for the same risk mapping relationship, this embodiment standardizes each state data item participating in the risk assessment. Let the set of state data items participating in the risk assessment be... , of which The original value of each state data item is Its standardized scale threshold is Its reference benchmark value is (For example, the rated value or normal operating median value of this item). Then the standardized state data item takes the value... Calculate using the following formula: ,in Indicates will The data is truncated to the [-1, 1] interval. Through the above standardization, different data items such as battery voltage, current, and temperature are mapped to a unified numerical range, facilitating subsequent calculations. Standardized As input for subsequent risk calculations.

[0072] In one feasible implementation, the set of state data items involved in risk assessment and change intensity calculation At least the following should be included: battery pack voltage, battery pack current, battery pack temperature, state of charge (SOC), and state of health (SOH); when the BMS provides individual cell voltages, the maximum or minimum deviation of the individual cell voltage can be added to the aggregate. For sets For each data item, the baseline value Take the nominal value or the median value of normal operation statistics for this data item; scale threshold. Take the upper limit of the allowable deviation for this data item under normal operating conditions. For example: battery pack voltage. A preset ratio of the rated voltage can be used; the battery pack temperature... A preset temperature rise threshold can be selected; SOC A preset percentage change threshold can be selected; the threshold is determined by the battery specification sheet or BMS calibration parameters and stored in non-volatile memory.

[0073] S12: Assume the total number of bits in the fault bitmap is... (e.g., 16-bit or 32-bit), the number of fault bits set to 1 in the fault bitmap at the current moment is: The fault intensity term is then defined as follows: The larger this ratio, the more types of failures occur simultaneously, and the higher the risk.

[0074] Let the fault bitmap at the current moment be... The fault bitmap at the previous moment was A new fault occurrence event is determined when a fault bit satisfies the condition that it was 0 in the previous time step and 1 in the current time step. Formalized as: if Make and If the condition is met, the new fault occurrence event is true; otherwise, it is false. This can be implemented using bitwise logical operations.

[0075] S13: Construct the linear input of the basic risk mapping : ,in For bias terms, For the first Risk mapping parameters for each standardized state data item For parameter set;

[0076] Using normalization function Will Mapped to the [0,1] interval as the base risk value : In this embodiment, Select logical functions: ;

[0077] An online self-calibration is performed only when a "new fault occurrence event" occurs, to correct the issue. This makes the basic risk values ​​more consistent with the safety requirements of fault-triggered scenarios;

[0078] Define the new fault occurrence event indicator. When a new fault occurs ,otherwise Let the lower bound of the fault triggering risk be... (e.g., 0.7), the upper bound of the no-fault risk is (e.g., 0.3), and satisfies Define the boundary constraint loss function. satisfy: ;

[0079] in, This is the regularization coefficient, used to suppress numerical instability caused by excessively large parameters; Let be the square of the L2 norm of the parameter set.

[0080] Update parameters using gradient descent: ,in Represents the boundary constraint loss function For parameter set The gradient (or subgradient) vector is composed of the partial derivatives of each parameter: Learning step size The range can be from 0.001 to 0.05; A value of 0.0001 to 0.01 can be used to stabilize parameters without affecting the response.

[0081] In this embodiment, because Piecewise functions have non-differentiable points, which can be addressed using subgradient calculations. An embedded implementation is equivalent to conditionally selecting whether to participate in gradient calculation.

[0082] like and ,but The derivative term takes effect;

[0083] like and ,but The derivative term takes effect;

[0084] Otherwise, the gradient of this term is 0, and only the gradient of the regularization term is retained.

[0085] For ease of implementation, the derivative of the basic risk value with respect to the linear input is given: ,and ;

[0086] Therefore, when the fault trigger item is in effect (i.e. and When ), we have: ;

[0087] When the fault-free upper bound is in effect (i.e.) and When ), we have: ;

[0088] Bias term The gradient can Replace with 1 to get the result. The above gradient calculation only involves addition, subtraction, multiplication, and exponentiation operations, which is suitable for real-time operation of embedded controllers.

[0089] The updated baseline risk value is obtained after self-calibration. This will be used in the next section to calculate the final battery risk assessment value. .

[0090] S14: The temporal changes in battery state information are used to calculate the change intensity term, which is obtained through a fault intensity-driven dual-time-scale adaptive statistic, including:

[0091] According to the fault intensity item Determine the adaptive update coefficients to satisfy:

[0092] ;

[0093] In the formula, For short-term update coefficients, For long-term update coefficients, preset ; This is a sensitivity coefficient used to adjust the degree of influence of fault intensity on the update coefficient; it is determined and solidified by experimental calibration. This represents the number of fault bits that are set in the fault bitmap at the current moment. This represents the total number of bits in the fault bitmap. As a preset boundary, and satisfying ;

[0094] For each standardized state data item Calculate the short-time averages separately Compared with long-term average :

[0095] Initialization during the first sampling Subsequent sampling is recursively calculated using the following formula:

[0096] ;

[0097] Calculate the intensity of change term : ,in The normalized scale threshold; For safety weights, satisfy In one feasible implementation, the weight of the temperature term is no less than the weight of the voltage term, and the weight of the voltage term is no less than the weight of the SOC term, so that factors more sensitive to safety, such as high temperature and overvoltage, have a greater impact on the intensity of change of the voltage term. The contribution is greater, and the weight can be determined by battery safety specifications or historical failure statistics, and can be updated through software configuration;

[0098] This formula quantifies both types of variables in a unified way:

[0099] Reflects trend drift (short-term mean deviates from long-term mean);

[0100] It reflects transient shocks (the current value deviates from the short-term mean).

[0101] S15: Determine the validation lag factor , used to characterize the additional amplification effect of the waiting delay at the receiving end from receiving the broadcast remote identification enhancement frame to obtaining the authentication key on the risk, and satisfying: ,in The preset number of delay time slots (the number of time slots for the authentication key disclosure relative to the authentication code transmission delay, which is a system configuration constant); The broadcast period for enhanced frames in broadcast-based remote identification; This represents the verification waiting delay (also known as the verification lag time), which is the time interval from the sending of the enhanced frame containing the message authentication code to the receiving end being able to obtain the authentication key corresponding to that authentication code. The preset reference lag time is used for normalization;

[0102] Based on fault strength term , Intensity of change Validate the lag factor and the updated base risk value Calculate the battery risk assessment value: ;

[0103] In the formula, This is the battery risk assessment value, and ; To pre-determine non-negative coefficients, used to adjust the contributions of the fault intensity term, the change intensity term, and the verification lag factor to the battery risk assessment value, one feasible implementation plan is as follows: Take not less than The value is set to ensure that the impact of fault triggering on risk assessment is no less than that of state change; A positive value is used to reflect the amplifying effect of the verification lag window on risk.

[0104] S2: Under the single-frame payload budget constraint, select the set of data items to be sent from the battery status information based on the battery risk assessment value and encode them to generate the payload to be sent. Determine the transmission mode as the original mode or the compressed mode according to the preset switching conditions. In the compressed mode, determine the differential data item set from the set of data items to be sent, generate differential data relative to the previous time for each differential data item in the differential data item set, and generate a compressed representation of the differential data according to the pseudo-random projection rule determined by the broadcast time slot index, and write it into the payload to be sent.

[0105] In this embodiment, the transmitting end performs the following processing in each broadcast cycle: First, it determines whether the transmission mode is the original mode or the compressed mode based on the battery risk assessment value and the fault bitmap; then, under the determined transmission mode, it selects the set of data items to be transmitted according to the single frame payload budget constraint and encodes and generates the payload to be transmitted, ensuring that the broadcast remote identification enhancement frame will not overflow, while ensuring that the receiving end can parse the actual transmitted data item value based on the extended flag bit.

[0106] The inputs to the transmitter include: the battery risk assessment value calculated for the current broadcast cycle, the fault bitmap at the current moment, and the battery status information output by the battery management system. The transmitter is pre-configured with the following parameters:

[0107] (1) The first risk threshold for entering the original mode The second risk threshold with return compression mode ,satisfy ;

[0108] (2) Window duration With enhanced frame broadcast period ;

[0109] (3) Single-frame payload budget (bytes); fixed field overhead (bytes); Length of authentication key disclosure field (byte);

[0110] (4) Risk grading thresholds used for field selection grading and To ensure consistency in caliber, let ,and .

[0111] For example: , ; , Broadcast cycle Seconds; window duration =10 seconds; Single-frame payload budget Bytes; fixed field overhead Bytes; Length of authentication key disclosure field Bytes; the key is disclosed once every 5 time slots.

[0112] in, , Used only for mode switching hysteresis; , It is only used to determine the number and range of fields in a given pattern. Although the two have different uses, the parameters satisfy the above relationship to avoid ambiguity in interpretation.

[0113] The sending end maintains the current mode state (raw mode or compressed mode) and discretizes the hold window duration into the number of hold window frames: ;

[0114] When switching from compressed mode to raw mode, the hold window counter is activated. , set as And if each frame is sent Then decrease by 1 until it reaches 0.

[0115] Within each broadcast cycle, mode switching is performed according to the following deterministic rules:

[0116] (1) If the risk assessment value is not less than If a fault bit is set in the fault bitmap, then enter the original mode; if switching from compression mode to original mode, then... .

[0117] (2) When in the original mode and Keep the original mode unchanged.

[0118] (3) When in the original mode and At that time, only when the risk assessment value is not greater than If there is no fault bit set in the fault bitmap, switch to compression mode; otherwise, continue to maintain the original mode.

[0119] (4) When in compression mode and the conditions for entering the original mode are not met, the compression mode remains unchanged.

[0120] pass The hysteresis interval avoids mode jitter, and the window is maintained to avoid frequent switching during high-risk phases, which can cause load organization instability.

[0121] After determining the above mode, the transmitter performs payload organization within that mode. The transmitter budgets the payload for each frame. It is broken down into fixed field costs and variable field budgets. It includes at least the extended flag field, the broadcast slot index field, and the message authentication code field; the authentication key disclosure field is carried only in the disclosure frame. Let the disclosure frame indicator be... If the variable field value is 1 in the disclosure frame and 0 otherwise, the variable field budget for this frame is: ;

[0122] Example calculation: If it is a non-disclosure frame, then Bytes; if it is a disclosure frame, then byte.

[0123] To enable the selection process, the sending end pre-establishes a table of data item byte lengths and uses little-endian encoding with preset precision. For example: battery pack voltage 2 bytes (0.01V), battery pack current 2 bytes (0.01A), temperature 2 bytes (0.01℃), SOC 2 bytes (0.01%), remaining capacity 4 bytes (mAh), fault bits... Figure 2 Bytes (16 bits), etc.

[0124] The sending end determines the field selection level for this frame based on the risk assessment value and determines the set of data items to be sent under budget constraints. The risk classification rules are as follows:

[0125] If the risk assessment value is less than It has been determined to be at a low risk level.

[0126] If the risk assessment value is not less than and less than It has been determined to be at a medium risk level.

[0127] If the risk assessment value is not less than It is identified as a high-risk level (and at this point it has usually entered the original mode).

[0128] For different risk levels, a pre-defined sequence of candidate data items (arranged from high to low in terms of safety relevance) is provided, for example:

[0129] Low-risk candidate sequences: fault bitmap, battery pack voltage, battery pack current, temperature, SOC;

[0130] Medium-risk candidate sequences: Based on low-risk sequences, increase SOH, number of cycles, and remaining capacity;

[0131] High-risk candidate sequences: Add extensions such as individual unit voltage to the medium-risk sequence (if budget allows).

[0132] The field addition rule is: Initialize the number of bytes used. Try adding items one by one according to the candidate sequence. If the data item added is less than the desired result... No more than Then add the data item to the set of data items to be sent and update it. Otherwise, stop adding data. This yields a unique set of data items to be sent.

[0133] The sending end sets an extended flag bit based on the set of data items to be sent: if a data item is selected, its corresponding flag bit is set to 1; otherwise, it is set to 0. Then, the values ​​of each data item are encoded and written into the payload to be sent in a preset order, along with the broadcast slot index, message authentication code, and the authentication key disclosure field in the disclosure frame. After encoding, the total byte length is verified to not exceed [a certain value]. If the limit is exceeded, backtrack by removing the last data item of the candidate sequence and re-encoding until the budget is met.

[0134] When the transmitter is in compression mode, the payload to be transmitted contains at least the original values ​​and compressed representation of the fault bitmap; the original values ​​of the fault bitmap are used by the receiver to directly parse the fault state, and the compressed representation is used to carry the compression result of the differential data. In the original mode, the payload is organized according to the original mode encoding rules.

[0135] When the sending end is in compression mode, it needs to determine the differential data item set from the set of data items to be sent, and ensure that the byte length of the compressed representation corresponding to the differential data item set does not exceed the byte budget reserved for the compressed representation in compression mode. This ensures that the compressed representation can be written into the payload to be sent without exceeding the single-frame payload budget. The determination method of the single-frame payload budget, the fixed field byte length, and the data item byte length written into the payload to be sent with the original value in compression mode can be found in the budget splitting and overhead deduction rules in the previous embodiment. This embodiment further explains the rules for determining the differential set and the determination criteria for the byte length of the compressed representation.

[0136] The sending end removes the data items corresponding to the fault bitmap from the set of data items to be sent, obtaining a candidate set. The reason for this is that the fault bitmap is used for alarm triggering, and in compressed mode, it needs to be written to the payload to be sent with its original value to ensure that the receiving end can directly parse it. Therefore, the fault bitmap does not participate in the candidate sorting of the differential set.

[0137] Calculate the normalized difference score for each candidate data item in the candidate set. ,satisfy: ;

[0138] in, For the current moment The standardized values ​​of each candidate data item. For the previous moment The standardized values ​​of each candidate data item (the values ​​of the previous time step are cached by the sender at the end of the previous broadcast period). For the first The preset normalization scale threshold corresponding to each candidate data item is used to normalize the difference amplitude of data items with different dimensions to a comparable range; for example, for the four types of data items, namely battery pack voltage, battery pack current, temperature, and SOC, separate settings can be configured. The values ​​are 0.10, 0.20, 0.15, and 0.05 under their standardized scale (example values; actual values ​​are determined by calibration or specification allowable deviations).

[0139] The sending end obtains the budget reserved bytes for compressed representation according to the aforementioned budget splitting rules, denoted as . This budget equals the single-frame payload budget minus the fixed field byte length and the remaining byte length after deducting the data item byte length that needs to be written to the payload in its original value under compressed mode. For example: single-frame payload budget 200 bytes; fixed field 12 bytes; at least original write fault bits under compressed mode. Figure 2 byte, then Bytes. If the current frame is a key disclosure frame and the disclosure field is 16 bytes, the fixed field becomes 28 bytes. byte.

[0140] To ensure that the compressed representation's byte length does not exceed the budget, this embodiment specifies that the compressed representation consists of a sequence of measured values ​​and a reconstruction parameter field:

[0141] The measurement sequence contains M measurements; each measurement is encoded using a fixed number of bytes, denoted as . The reconstruction parameter field is the minimum set of parameters required by the receiving end to recover the differential data; its length in bytes is denoted as [missing information]. This includes at least: measurement number identifier, measurement value encoding precision identifier, and identifier information used to generate pseudo-random projection rules (such as seed check summary or equivalent identifier).

[0142] Therefore, compression indicates byte length. Calculate using the following formula: To make A definite relationship is established with the size of the differential data item set, which can then be used for stopping the decision. In this embodiment, the measurement number M is configured as the number of elements in the differential data item set. Linear functions: ,in and These are preset parameters.

[0143] Sender scores based on normalized differential scores Sort the candidate set from largest to smallest, and add the candidate data items to the difference data item set in the sorted order. Each time a candidate data item is added, the sender updates... Based on this, update the measurement number M and calculate the compressed representation byte length. .

[0144] When attempting to add the next candidate data item, it will result in the updated... Greater than If this happens, immediately stop adding new members to the current set. This refers to the set of difference data items. This rule ensures that the boundary of the difference data item set is uniquely determined by the budget stopping condition, thus avoiding uncertainty.

[0145] After the set of differential data items is determined, the transmitter generates differential data relative to the previous time for each differential data item in the set, and generates a compressed representation according to the pseudo-random projection rule determined by the broadcast time slot index. The transmitter writes the original value of the fault bitmap and the compressed representation into the payload to be transmitted, and sets an extended flag bit to indicate the existence of the actual data item value written in this frame.

[0146] The following embodiments illustrate how the transmitter generates a seed for a pseudo-random projection rule based on the broadcast slot index and the UAV identifier, and writes the identification information used to generate the seed into the payload to be sent and overwritten by the message authentication code, so that the receiver generates a consistent pseudo-random projection rule after successful verification to recover the differential data.

[0147] Input includes:

[0148] Broadcast slot index: denoted as , is an unsigned integer, generated by the sender in increments according to the broadcast period (e.g., incremented by 1 for each frame sent, modulo 1). (Circling)

[0149] Drone identifier: Described as ID, it is a fixed identifier and can be the Basic ID for broadcast remote identification or its hash digest; the sending and receiving ends have the same rules for the value of ID;

[0150] Preset parameters: Hash algorithm (SHA-256), seed length (32 bits), pseudo-random number generator (XorShift128+ or equivalent reproducible PRNG), domain separation constant (used to prevent different matrices in the same frame from using the same random stream; in this embodiment, two constant strings "C" and "N" are used), seed identifier information length (4 bytes, 32-bit checksum).

[0151] Field definitions (written in the payload to be sent and overridden by the message authentication code): Time slot index field (2 bytes), UAV identifier field (4 bytes; if the actual ID is longer, its 32-bit truncated digest can be written, but it is sufficient for sending / receiving), seed checksum field (4 bytes, used for consistency verification at the receiving end).

[0152] Before generating the compressed representation of this frame, the sending end performs the following deterministic process:

[0153] Constructing hash input: Concatenate the byte order of the byte ID with the ID in a fixed byte order (either little-endian or big-endian, but it must be fixed in the system) to obtain a byte string. ;

[0154] Generate seeds:

[0155] Calculate hash value Take the first 4 bytes of H as a 32-bit seed; to ensure consistency verification at the receiving end, take the last 4 bytes of H as a seed check digest tag (or take the first 4 bytes XORed with a fixed constant, but a fixed rule is required).

[0156] Write payload and overwrite authentication: The sender writes the slot index field, drone identifier field and seed verification digest field to a fixed position in the payload to be sent (e.g., immediately after the extension flag or immediately after the compression representation parameter field), and ensures that the message authentication code generated subsequently overwrites the payload to be sent, which includes the above fields.

[0157] In this way, if an attacker tampers with the i / ID / tag, the message authentication code verification will fail, and the receiving end will not enter the recovery process.

[0158] To avoid the correlation caused by using the same random flow for both key and non-key matrices, this embodiment employs domain separation:

[0159] Key matrix random stream seed: hash the seed again with the constant "C" to get the seed. C ;

[0160] Non-critical matrix random stream seed: hash the seed again with the constant "N" to obtain the seed. N .

[0161] In terms of implementation, it can be directly defined as follows:

[0162] ;

[0163] ;

[0164] Subsequently, the sending end used seed respectively C With seed N Initialize the PRNG, generating measurement matrix elements row-wise and column-wise. Matrix elements can take the following values: The random symbol or fixed-point random value (such as 16-bit fixed-point) is used, and each row is normalized to control the range of values ​​(the normalization rule can be fixed).

[0165] After obtaining the disclosure key and verifying the message authentication code, the receiving end performs the following steps:

[0166] Parse the time slot index field, the drone identifier field, and the seed verification digest field;

[0167] Constructed using the same byte order and concatenation rules as the sending end. ,calculate Get the local seed and local tag;

[0168] Verify that the local tag matches the seed verification digest field in the payload; if they do not match, determine that the data has been tampered with or parsed incorrectly, and terminate the recovery process.

[0169] Generate seeds according to the domain separation rules. C seed N It also initializes the PRNG and generates a measurement matrix consistent with that of the transmitter, which is then used for subsequent differential data recovery and reconstruction.

[0170] Through the above mechanism, the pseudo-random projection rules and measurement matrix can be deterministically reproduced at the receiving end. At the same time, since the seed-related fields are protected by the message authentication code, it can effectively prevent the recovery of the wrong matrix caused by tampering.

[0171] The following embodiment illustrates that in compression mode, the sending end divides the differential data item set into a critical differential data subset and a non-critical differential data subset, and determines the total number of available measurements under the constraint of reserved bytes budget for compressed representation. Then, based on the battery risk assessment value, critical proportion and preset hierarchical coefficient, the number of critical / non-critical measurements is allocated, two measurement value sequences are generated and a compressed representation is constructed.

[0172] Input includes:

[0173] Differential data item set Battery risk assessment value Compression indicates that reserved bytes are allocated. Number of bytes encoded for a single measurement value (Fixed configuration), number of bytes in refactored parameter fields (Fixed configuration, including at least critical measurement data, non-critical measurement data, accuracy identifier, seed summary, etc.)

[0174] Preset parameters include:

[0175] Stratification coefficient Non-negative, balancing the impact of risk assessment value and key proportion on allocation;

[0176] Key data item list: Includes at least one or more of the following: battery pack voltage difference, battery pack current difference, and temperature difference (system configuration is fixed, and sending / receiving are consistent).

[0177] The sending end, based on the list of key data items configured in the system, from Key difference data items were selected to form a key difference data subset. The remainder constitutes a non-critical subset of the difference data. Then The differential data within each segment are concatenated in a fixed order to form the key difference vector. ,Will The individual difference data are concatenated into a non-critical difference vector. The splicing order is fixed in the protocol (e.g., incremented by data item number) to ensure that the receiving end can reconstruct the data consistently.

[0178] The sending end determines the total number of available measurements M (number of measurement values) according to budget constraints, using deterministic calculation:

[0179] ;

[0180] like Then set M=0 and enter the exception handling path that only sends the original value of the fault bitmap and not the compressed representation (this exception handling is a system configuration option; under normal configuration, by selecting sufficient...). Avoid M=0 with field overhead.

[0181] The sending end determines the key measurement numbers according to the following deterministic rules. Non-critical measurement numbers :

[0182] Calculate key percentages :make ;like Then the compressed representation is skipped and generated directly;

[0183] Calculate the proportion of key measurements :Will and By stratification coefficient Weighted and will Truncate to [0,1]; Requirement Follow Increase without decreasing, follow Increase without decreasing. An enforceable deterministic rule is:

[0184] when At that time, take ;

[0185] when At that time, take ;

[0186] Let key measurement numbers for Rounded up and not exceeding M, non-critical measurement number .

[0187] The allocation rule satisfies the following conditions: the higher the risk and the greater the proportion of critical differences, the more critical measurements there will be; at the same time, the total number of measurements is strictly equal to the total number of available measurements in the budget, and will not cause compressed representation overflow.

[0188] The sending end generates pseudo-random projection rules and measurement matrices:

[0189] Generation and key measurement data Matching key measurement matrix ;

[0190] Generation and non-critical measurement numbers Matching non-critical measurement matrix .

[0191] Then calculate separately:

[0192] Key measurement series ;

[0193] Non-critical measurement series ;

[0194] Finally, the compressed representation is constructed as follows:

[0195] Reconstruct parameter fields (at least containing) , (Measurement accuracy label, seed summary);

[0196] Key measurement series (common (one measurement value)

[0197] Non-critical measurement series (common (One measured value).

[0198] because And M is calculated from the budget, the compressed representation has a byte length equal to Therefore, not exceeding .

[0199] After successful authentication, the receiving end reads the parameter fields. , With seed summary, and reconstruct the same measurement matrix. , This allows for the recovery and reconstruction of corresponding data items for both critical and non-critical differences, using the same dimensions.

[0200] S3: Generate delayed disclosure authentication information for the payload to be transmitted. The delayed disclosure authentication information includes the message authentication code and the authentication key disclosed in the preset delay time slot.

[0201] S31: The transmitting end maintains the broadcast slot index. Its increment step size is 1, and it is related to the enhanced frame broadcast period. One-to-one correspondence: Each time an enhanced frame is sent, it will be... Increment by 1; wrap around when the preset maximum value (e.g., 16-bit upper limit) is reached. To ensure that the receiver can determine replay, the receiver records the set of most recently authenticated time slot indices after authentication (e.g., record the most recent W indices in a sliding window manner, where W can be 1000), and rejects the output of frames with duplicate indices.

[0202] S32: The sending end generates a one-way key chain. ,satisfy ,in Use SHA-256 or equivalent hashing. The sender will send the keychain commitment value (e.g., an initial commitment). Alternatively, the anchor value used to verify the disclosure key can be written into a fixed field of the enhancement frame and overwritten by the message authentication code, enabling the receiver to perform consistency verification after subsequent disclosure of the key.

[0203] The sending end is configured with a preset delay time slot number d: in the time slot Send a payload containing the message authentication code in the time slot. Disclose the authentication key used to verify time slot 𝑖.

[0204] S33: The sending end generates a message authentication code for the payload to be sent. and with time slot index Binding. To ensure deterministic authentication coverage, the sender uses the payload to be transmitted, the slot index field, and a digest of fixed header fields related to this frame as authentication input, and generates an authentication code using the HMAC symmetric authentication algorithm:

[0205]

[0206] Among them, Payload is the byte sequence of the payload to be sent (including battery data item values, compressed representation and necessary identification fields); 𝑖 is the time slot index of this frame; HeaderDigest is the digest after hashing the fixed fields (such as data type, version number, etc.) in the enhanced frame header (fixed fields can also be directly concatenated, and in engineering, the set of fixed fields can be kept constant).

[0207] In the time slot In the disclosure frame, the sender will send the authentication key. The authentication key disclosure field is written into the payload to be sent and is also overwritten by the message authentication code of the frame to prevent the disclosure key from being replaced.

[0208] S4: Encapsulate the payload to be sent into a broadcast remote identification enhancement frame and broadcast it. The data flag field of the broadcast remote identification enhancement frame includes an extended flag bit, which is used to indicate the existence of the value of the corresponding data item in the set of data items to be sent in the data content field.

[0209] S41: The sending end encapsulates the payload to be transmitted into a broadcast remote identification enhancement frame. The enhancement frame includes: data type field, version number field, data length field, data flag field, and data content field.

[0210] The data flag field has been expanded from a basic length to an extended length: the basic length is 3 bytes, and the extended length is 5 bytes. The newly added 4th and 5th bytes constitute the extended flag bits. The extended flag bits are used to indicate whether each data item to be sent exists in the data content field, so that the receiving end can parse the set of data items carried in this frame bit by bit.

[0211] S42: The sender allocates an extended flag bit for each type of sendable data item and forms a bit mapping table (stored in the protocol configuration of the sender and receiver). When a data item is written into the data content field, the corresponding bit is set to 1; otherwise, it is set to 0.

[0212] In one possible implementation, the extended flags cover at least: battery pack voltage, battery pack current, SOC, SOH, temperature, cycle count, remaining capacity, fault bitmap, compressed representation field, and authentication key disclosure field; wherein the authentication key disclosure field is set only in the disclosure frame.

[0213] For example: Byte 4, bit 0 indicates that the battery pack voltage exists; bit 1 indicates that the battery pack current exists; bit 2 indicates that the temperature exists; bit 3 indicates that the SOC exists; Byte 5, bit 0 indicates that the fault bitmap exists; bit 1 indicates that the compressed representation exists; bit 2 indicates that the authentication key disclosure field exists (disclosure frame is set to 1).

[0214] S43: The data length field is the total length of the data flag field (5 bytes) plus the actual length of the data content field. The data content field is written in a preset order.

[0215] 1) Time slot index field;

[0216] 2) Values ​​of each data item under the extended flag mapping (in little-endian encoding and fixed-precision quantization).

[0217] 3) Compressed representation (if it exists);

[0218] 4) Message authentication code;

[0219] 5) Authentication key disclosure field (disclosure frame exists).

[0220] Continuous quantities such as battery voltage, current, and temperature are quantized into integers with preset precision and written as such, for example, voltage 0.01V, current 0.01A, and temperature 0.01℃.

[0221] When the broadcast transmission link is interrupted (e.g., the broadcast module loses power, radio frequency is abnormal, or the system enters protection mode), the transmitting end performs buffering and retransmission to ensure that the data can still be verified and parsed by the receiving end after recovery.

[0222] The sending end caches the following content to form a cache unit:

[0223] The byte sequence of the payload to be sent (including slot index, data item values, compressed representation, message authentication code, and disclosure key field, if present);

[0224] The authentication context corresponding to delayed disclosure authentication includes: the slot index of the frame, the disclosure slot index that the frame needs to verify, and the key chain commitment parameters used for consistency verification (which are cached if the frame carries a commitment field).

[0225] The cache is stored using a circular buffer or queue structure, with each cache unit having an enqueue timestamp to calculate the interruption duration.

[0226] After the transmitter resumes broadcasting, it retransmits the enhanced frames in ascending order of the time slot index of the buffer unit, and ensures that the disclosure key field is still disclosed according to the correspondence of d time slots delayed:

[0227] If the retransmitted frame contains a message authentication code but its corresponding disclosure key has not yet been retransmitted, the disclosure key field will be inserted into the subsequent retransmitted frame as originally planned.

[0228] If the retransmission spans multiple disclosure points, they will be disclosed one by one in the order of the time slot index to avoid the receiving end being unable to verify the message authentication code due to misalignment of the disclosure key.

[0229] The above strategy ensures that the correspondence between the authentication key and the message authentication code is not disrupted, so that the receiving end can output the battery status information only after obtaining the authentication key and completing the verification.

[0230] When the interrupt duration does not exceed a preset duration threshold, all cache units are cached; when the interrupt duration exceeds the preset duration threshold, only cache units that meet any of the following conditions are cached:

[0231] The cache unit contains a fault bitmap;

[0232] Or the battery risk assessment value corresponding to the cache unit is not less than the preset cache threshold.

[0233] The remaining cache units are discarded to save storage space and prioritize the retention of critical security information.

[0234] S5: The receiving end receives the broadcast remote identification enhancement frame, verifies the message authentication code after obtaining the authentication key, and obtains the battery status information by parsing the extended flag bit when the verification is successful. When the broadcast remote identification enhancement frame contains a compressed representation, the differential data is restored according to the pseudo-random projection rule and the corresponding data item is reconstructed, and the battery alarm or status result is output.

[0235] S51: The receiving end parses the data flag field (5 bytes) to obtain the extended flag bit; it parses the data content field to obtain at least the time slot index, message authentication code, and disclosure key field (if this frame is a disclosure frame). For non-disclosure frames, the receiving end adds the frame to the verification queue according to the time slot index, waiting for the corresponding disclosure key to arrive. For disclosure frames, the receiving end extracts the disclosure key. First, a one-way chain consistency check is performed: the disclosed key is hashed a preset number of times (the number of times is the delay slot number), and it should match the known commitment value; if the check passes, the disclosed key is considered trustworthy. For frames in the verification queue that match the disclosed key, the local HMAC is calculated and compared with the message authentication code; if it passes, the frame payload is considered genuine and valid, and the frame slot index is recorded to resist replay.

[0236] S52: After successful authentication, the receiving end parses the data content fields one by one according to the extended flag bits:

[0237] If a bit is 1, then read and decode according to the byte length and precision rules of that data item;

[0238] If the value is 0, then skip reading that data item.

[0239] In this way, the receiver can determine the set of battery data items actually carried in the current frame without relying on external negotiation.

[0240] S53: When the extended flag indicates that the compressed representation exists, the receiver reads the compressed representation and generates a pseudo-random projection rule consistent with the transmitter based on the time slot index. Based on this, the differential data is recovered and the corresponding data items are reconstructed. Then, the battery alarm or status result is output by combining the fault bitmap and the battery risk assessment value.

[0241] When authentication fails or replay test is successful, the receiver does not output battery status information, but only records audit logs.

[0242] like Figure 2 As shown, another embodiment of the present invention provides a broadcast-based remote identification adaptive transmission system for drone battery status, comprising:

[0243] The battery status acquisition and standardization module is used to acquire battery status information output by the battery management system and perform standardization processing.

[0244] The risk assessment module is used to obtain the battery risk assessment value based on the fault status and time sequence changes in the battery status information, according to the preset risk mapping relationship.

[0245] The payload budget and data item selection module is used to select a set of data items to be sent from the battery status information based on the battery risk assessment value under the single-frame payload budget constraint.

[0246] The payload encoding module is used to encode the set of data items to be sent and generate the payload to be sent.

[0247] The mode switching control module is used to determine the transmission mode as either raw mode or compressed mode according to preset switching conditions.

[0248] The differential generation module is used to determine the differential data item set from the set of data items to be sent when the transmission mode is compression mode, and to generate differential data relative to the previous time for each differential data item in the differential data item set.

[0249] The pseudo-random projection compression module is used to generate a compressed representation of the differential data according to the pseudo-random projection rules determined by the broadcast time slot index, and write the compressed representation into the payload to be transmitted.

[0250] The delayed disclosure authentication module is used to generate delayed disclosure authentication information for the payload to be transmitted. The delayed disclosure authentication information includes a message authentication code and an authentication key disclosed in a preset delay time slot.

[0251] The enhanced frame encapsulation and broadcast transmission module is used to encapsulate the payload to be transmitted into a broadcast remote identification enhanced frame and broadcast it. The data flag field of the broadcast remote identification enhanced frame includes an extended flag bit, which is used to indicate the existence of the value of the corresponding data item in the set of data items to be transmitted in the data content field.

[0252] The receiving verification and parsing module is used to receive broadcast remote identification enhancement frames, verify the message authentication code after obtaining the authentication key, and parse the battery status information according to the extended flag bit when the verification is successful.

[0253] The differential recovery and reconstruction module is used to recover differential data and reconstruct corresponding data items according to pseudo-random projection rules when the broadcast remote identification enhancement frame contains a compressed representation.

[0254] The alarm and status output module is used to output battery alarms or status results based on the battery status information obtained from parsing or reconstruction.

[0255] In summary, by introducing technologies such as adaptive data selection, differential compression, pseudo-random projection, and delayed disclosure authentication during battery status information transmission, this invention greatly improves communication efficiency, reduces bandwidth consumption, ensures data security, and effectively enhances the stability and reliability of the UAV battery management system during long-range flights and long-duration missions.

[0256] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any person skilled in the art can easily conceive of various variations or substitutions within the technical scope disclosed in this application, and these should all be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

Claims

1. A broadcast-based remote identification and adaptive transmission method for UAV battery status, characterized in that, The method includes: Collect and standardize the battery status information output by the battery management system, and obtain the battery risk assessment value based on the fault status and time sequence changes in the battery status information according to the preset risk mapping relationship. Under the single-frame payload budget constraint, a set of data items to be sent is selected from the battery status information based on the battery risk assessment value and encoded to generate a payload to be sent. The transmission mode is determined to be either the original mode or the compressed mode according to the preset switching conditions. In the compressed mode, a set of differential data items is determined from the set of data items to be sent. For each differential data item in the set of differential data items, differential data relative to the previous time is generated. A compressed representation of the differential data is generated according to the pseudo-random projection rule determined by the broadcast slot index and written into the payload to be sent. Generate delayed disclosure authentication information for the payload to be sent, the delayed disclosure authentication information including a message authentication code and an authentication key disclosed in a preset delay time slot; The payload to be transmitted is encapsulated into a broadcast remote identification enhancement frame and broadcast. The data flag field of the broadcast remote identification enhancement frame includes an extended flag bit, which is used to indicate the existence of the value of the corresponding data item in the set of data items to be transmitted in the data content field. The receiving end receives the broadcast remote identification enhancement frame, verifies the message authentication code after obtaining the authentication key, and obtains the battery status information by parsing the extended flag bit when the verification is successful. When the broadcast remote identification enhancement frame contains the compressed representation, the differential data is restored and the corresponding data item is reconstructed according to the pseudo-random projection rule, and the battery alarm or status result is output.

2. The broadcast-based remote identification and adaptive transmission method for UAV battery status according to claim 1, characterized in that, The battery status information includes at least: battery pack voltage, individual cell voltage, battery pack current, state of charge (SOC), state of health (SOH), temperature, cycle count, remaining capacity, and fault bitmap. The preset risk mapping relationship is specifically a parameterized self-calibrated basic risk mapping, including: Calculate the basic risk value based on standardized battery state information. : ; In the formula, For the first Each standardized state data item takes a value. A standardized set of state data items for risk assessment. For risk mapping parameters, A function to normalize the result to [0,1]; A new fault occurrence event is used as a self-calibration trigger signal. The new fault occurrence event is obtained by comparing the fault bit map at the current time with that at the previous time. When there is a fault bit that changes from 0 to 1, it is determined that a new fault occurrence event has occurred. When a new fault event is detected, the risk mapping parameters are updated online using the gradient descent algorithm based on the boundary constraint loss function to obtain the updated base risk value. ; Where the boundary constraint loss function satisfy: ； In the formula, This is an indicator of a new fault occurrence; it is set to 1 when a new fault occurs, and 0 otherwise. This is the lower bound of the fault trigger risk. This represents the upper bound of the risk of no failure. The regularization coefficient is . For the reason The set of parameters that make up the composition.

3. The broadcast-based remote identification and adaptive transmission method for UAV battery status according to claim 2, characterized in that, The temporal changes in the battery state information are used to calculate the change intensity term, and the change intensity term is obtained through a fault intensity-driven dual-time-scale adaptive statistic, including: The adaptive update coefficient is determined based on the number of bits set in the fault bitmap, satisfying the following: ； In the formula, For short-term update coefficients, For long-term update coefficients, preset ; This is the sensitivity coefficient; This represents the number of fault bits that are set in the fault bitmap at the current moment. This represents the total number of bits in the fault bitmap. As a preset boundary, and satisfying ; Calculate the short-time mean for each standardized state data item. Compared with long-term average During the first sampling, the short-term and long-term means are initialized to the corresponding standardized state data item values; in subsequent sampling, they are recursively updated using the following formula: ； Calculate the intensity of change term : ,in For safety weights, satisfy , The normalized scale threshold; Determine the validation lag factor : ,in The preset number of delay slots; The broadcast period for enhanced frames in broadcast-based remote identification; This indicates the verification waiting delay. Preset reference lag time; Based on fault strength term , Intensity of change Validate the lag factor and the updated base risk value Calculate the battery risk assessment value: ; In the formula, This is the battery risk assessment value, and ; These are preset non-negative coefficients used to adjust the contributions of the fault intensity term, the change intensity term, and the verification lag factor to the battery risk assessment value, respectively.

4. The broadcast-based remote identification and adaptive transmission method for UAV battery status according to claim 3, characterized in that, The preset switching condition is specifically: the battery risk assessment value reaches the preset risk threshold or a fault bit is set in the fault bit map. When the preset switching conditions are met, the transmission mode is determined to be the original mode; otherwise, the transmission mode is determined to be the compressed mode. The payload to be transmitted in the compressed mode includes the original value of the fault bitmap and the compressed representation. The preset risk thresholds include a first risk threshold for entering the original mode and a second risk threshold for returning to the compressed mode, and the first risk threshold is greater than the second risk threshold. When the battery risk assessment value is greater than or equal to the first risk threshold, it enters the original mode and maintains the preset holding window duration. After the holding window duration ends, it switches to the compression mode only when the battery risk assessment value is less than or equal to the second risk threshold and there is no set fault bit in the fault bitmap.

5. The broadcast-based remote identification and adaptive transmission method for UAV battery status according to claim 4, characterized in that, In compression mode, determining the differential data item set from the set of data items to be sent includes: Remove the data items corresponding to the fault bitmap from the set of data items to be sent to obtain a candidate set; Calculate the normalized difference score for each candidate data item in the candidate set. ,satisfy: ; in, For the current moment The standardized values ​​of each candidate data item. For the previous moment The standardized values ​​of each candidate data item. For the first The preset normalization scale threshold corresponding to each candidate data item; The candidate set is sorted from largest to smallest according to the normalized difference score, and candidate data items are selected and added to the difference data item set in the sorted order until further additions would cause the byte length of the compressed representation corresponding to the difference data item set to exceed the byte budget reserved for the compressed representation in the compression mode. The selection stops when further additions are made. The byte budget reserved for the compressed representation is equal to the single frame payload budget minus the fixed field byte length and the remaining byte length after deducting the data item byte length that needs to be written to the payload to be sent with the original value in the compression mode.

6. The broadcast-based remote identification and adaptive transmission method for UAV battery status according to claim 5, characterized in that, The pseudo-random projection rule is generated by a pseudo-random number generator. The seed of the pseudo-random number generator is obtained by hashing the broadcast time slot index, and the input of the hash operation further includes the UAV identifier. The identification information used to generate the seed is written into the payload to be sent and overwritten by the message authentication code, so that after the receiver passes the verification, it generates a consistent pseudo-random projection rule based on the same broadcast slot index and UAV identifier to recover the differential data.

7. The broadcast-based remote identification and adaptive transmission method for UAV battery status according to claim 6, characterized in that, The set of differential data items is divided into a critical differential data subset and a non-critical differential data subset, and different measurement resources are allocated to the two subsets to generate the compressed representation, wherein: The critical differential data subset includes at least one or more of the battery pack voltage differential, battery pack current differential, and temperature differential, and the non-critical differential data subset consists of data items in the differential data item set other than the critical differential data subset. Under the constraint of the byte budget reserved in the compressed representation, the total number of available measurements for generating the measurement value sequence in this frame is determined based on the number of encoded bytes of a single measurement value and the number of bytes of the reconstruction parameter field; wherein the total number of available measurements is the maximum number of measurements that ensures the byte length of the compressed representation does not exceed the byte budget reserved in the compressed representation, and the total number of available measurements is obtained by subtracting the number of bytes of the reconstruction parameter field from the byte budget reserved in the compressed representation, dividing by the number of encoded bytes of a single measurement value, and taking the integer part; Based on the battery risk assessment value, the proportion of key differential data subsets in the differential data item set, and the preset stratification coefficient, the number of key measurements is determined and the remaining number of measurements is determined as non-key measurements. The number of key measurements does not decrease as the battery risk assessment value increases, does not decrease as the proportion of key differential data subsets increases, and the number of key measurements is not greater than the total number of available measurements. Key measurement matrices corresponding to key measurement numbers and non-key measurement matrices corresponding to non-key measurement numbers are generated according to the pseudo-random projection rules. A linear projection is performed on the difference data vectors corresponding to the key difference data subsets to obtain a key measurement value sequence, and a linear projection is performed on the difference data vectors corresponding to the non-key difference data subsets to obtain a non-key measurement value sequence. The reconstruction parameter field, the key measurement value sequence, and the non-key measurement value sequence together constitute the compressed representation. The reconstruction parameter field includes at least the key and non-key measurement numbers and is used to instruct the receiving end to generate a consistent measurement matrix according to the corresponding measurement numbers to recover the difference data.

8. The broadcast-based remote identification and adaptive transmission method for UAV battery status according to claim 1, characterized in that, The broadcast-style remote identification enhancement frame includes a data type field, a version number field, a data length field, a data flag field, and a data content field; wherein the data flag field is extended from a basic length to an extended length, the basic length being 3 bytes and the extended length being 5 bytes, and the newly added 4th and 5th bytes constitute the extended flag bit.

9. The broadcast-based remote identification and adaptive transmission method for UAV battery status according to claim 1, characterized in that, When broadcast transmission is interrupted, the payload to be transmitted and the authentication context corresponding to the delayed disclosure authentication information are cached; after broadcast is resumed, the broadcast-style remote identification enhancement frame is retransmitted, and the correspondence between the authentication key and the message authentication code is not destroyed, so that the receiving end can parse and output the battery status information after obtaining the authentication key and completing the verification; and when the interruption duration exceeds the preset duration threshold, only the payload to be transmitted containing the fault bitmap or the battery risk assessment value is not less than the preset cache threshold is cached.

10. A broadcast-based remote identification adaptive transmission system for UAV battery status, applied to the broadcast-based remote identification adaptive transmission method for UAV battery status as described in any one of claims 1-9, characterized in that, The system includes: The battery status acquisition and standardization module is used to acquire battery status information output by the battery management system and perform standardization processing. The risk assessment module is used to obtain a battery risk assessment value based on the fault status and time sequence changes in the battery status information according to a preset risk mapping relationship. The payload budget and data item selection module is used to select a set of data items to be sent from the battery status information based on the battery risk assessment value under the single-frame payload budget constraint. The payload encoding module is used to encode the set of data items to be sent and generate the payload to be sent. The mode switching control module is used to determine the transmission mode as either raw mode or compressed mode according to preset switching conditions. The differential generation module is used to determine a differential data item set from the set of data items to be sent when the transmission mode is compression mode, and to generate differential data relative to the previous time for each differential data item in the differential data item set. The pseudo-random projection compression module is used to generate a compressed representation of the differential data according to the pseudo-random projection rules determined by the broadcast time slot index, and write the compressed representation into the payload to be transmitted; The delayed disclosure authentication module is used to generate delayed disclosure authentication information for the payload to be sent. The delayed disclosure authentication information includes a message authentication code and an authentication key disclosed in a preset delay time slot. The enhanced frame encapsulation and broadcast transmission module is used to encapsulate the payload to be transmitted into a broadcast remote identification enhanced frame and broadcast it, wherein the data flag field of the broadcast remote identification enhanced frame includes an extended flag bit, which is used to indicate the existence of the value of the corresponding data item in the set of data items to be transmitted in the data content field. The receiving verification and parsing module is used to receive the broadcast remote identification enhancement frame, verify the message authentication code after obtaining the authentication key, and parse the battery status information according to the extended flag bit when the verification is successful. The differential recovery and reconstruction module is used to recover differential data and reconstruct corresponding data items according to the pseudo-random projection rule when the broadcast remote identification enhancement frame contains the compressed representation; The alarm and status output module is used to output battery alarms or status results based on the battery status information obtained from parsing or reconstruction.

Images