Unmanned aerial vehicle telemetry link interference and suppression system based on artificial intelligence

Abstract

The invention provides an unmanned aerial vehicle telemetering link interference and suppression system based on artificial intelligence, and relates to the technical field of unmanned aerial vehicle attitude control. Speed disturbance and electric ion density disturbance are preprocessed in a measurement module, and a normalized disturbance index is output in real time in a prediction module; a multivariable nonlinear optimization algorithm is adopted in the calculation and solution module to synchronously determine a transmitting power increment, a key time slot proportion and a link switching strategy, and finally three types of instructions are issued in parallel in the execution module and backup takeover is rapidly completed; multi-factor coupling suppression is achieved, and the application requirement for high-reliability remote measurement of canyon unmanned aerial vehicle formation under a high-voltage transmission line is met.

Description

Technical Field

[0001] The present invention relates to the technical field of unmanned aerial vehicle (UAV) attitude control, and in particular to an artificial intelligence-based UAV telemetry link interference and suppression system. Background Art

[0002] In the canyon terrain beneath high-voltage transmission lines, complex terrain obstruction, electromagnetic interference, and the Doppler effect caused by high-speed movement cause link quality to exhibit strong time-varying and nonlinear attenuation characteristics. To this end, researchers have gradually introduced algorithms such as predictive control and Kalman filtering to dynamically estimate environmental interference and implement real-time adjustments in transmit power and time slot allocation. However, these studies often focus on single parameter optimization or separate the measurement, solution, and execution links. They lack a full-link coupling suppression approach and find it difficult to balance link robustness and resource utilization efficiency under the dual influence of changes in the distribution of high-charge clusters and relative velocity disturbances of the formation. In addition, existing solutions often rely on static threshold prediction or coarse-grained switching strategies when triggering link switching, resulting in high misjudgment rates and extended switching times. They cannot meet the requirements for simultaneous optimization of multiple indicators such as improved availability, reduced bit error rates, and shortened recovery times.

[0003] In the prior art, the publication number CN111694371A is titled "A UAV jamming system and a UAV with jamming capabilities," which are used to simultaneously control the flight of a host UAV and interfere with non-host UAVs, thereby enabling the UAV to fly while also interfering with other UAVs. The UAV jamming system includes a jamming module for generating a jamming signal; a flight control motherboard for receiving the control signal and controlling the flight of the UAV; and a radio frequency module for simultaneously switching the wireless channel for the jamming signal and the wireless channel for the control signal at preset time intervals. This allows, within the same time period, the jamming signal to be transmitted via a first wireless channel using a jamming transmitting antenna, and the control signal to be received via a second wireless channel using a communication receiving antenna, and then sent to the flight control motherboard. The frequency bands corresponding to the first and second wireless channels do not overlap.

[0004] Although various optimization schemes have been proposed for adaptive transmit power adjustment, dynamic time slot resource allocation, and link switching mechanisms, an integrated system that integrates the entire closed-loop link process of "measurement-prediction-calculation-solution-execution" has not yet been formed:

[0005] 1. First, traditional measurement modules focus on signal strength (RSSI) or instantaneous bit error rate (BER) monitoring, ignoring the impact of relative speed differences in UAV formations on Doppler frequency deviation and dynamic link quality degradation;

[0006] 2. Second, although algorithms such as Kalman filtering can predict changes in ion cluster density online, conventional applications often lack preprocessing and normalization of the predicted values, making it impossible to quantify the synergistic effect of the two types of disturbance indicators in subsequent link scoring and bit error rate prediction. Third, existing nonlinear optimization methods are mostly limited to single-objective or dual-objective areas, making it difficult to achieve a multi-variable, multi-constraint joint solution between the three types of control instructions: transmit power increment, critical time slot proportion, and backup switching command.

[0007] 3. Finally, when a threshold is exceeded, the execution module can usually only trigger a single radio frequency backup or time slot switching, and the switching trigger and switching confirmation processes are separated, resulting in excessively long link recovery time and an imbalance in the trade-off between availability and bit error rate.

[0008] The above information disclosed in this Background section is only for enhancement of understanding of the background of the present disclosure and therefore it may contain information that does not form the prior art that is already known to a person of ordinary skill in the art. Summary of the Invention

[0009] In response to the deficiencies in the prior art, the present invention provides an artificial intelligence-based UAV telemetry link interference and suppression system to solve the problems raised in the above background technology.

[0010] The purpose of the present invention is to achieve this by providing an artificial intelligence-based UAV telemetry link interference and suppression system, which is applied to the canyon UAV formation cruising scenario under high-voltage transmission lines, specifically comprising:

[0011] Measurement module: used to measure the relative speed difference between each node UAV in the UAV formation and its neighboring UAVs, and pre-process the relative speed difference to generate a normalized speed disturbance index;

[0012] Prediction module: used to collect the ion cluster density of the node drone at the current moment in real time, and use the Kalman filter algorithm to online predict the ion cluster density prediction value in the future short-term window, and pre-process the ion cluster density prediction value to generate a normalized density disturbance index;

[0013] Calculation module: used to calculate the link score of each node drone based on the speed disturbance index, and calculate the bit error rate prediction value based on the density disturbance index of each node drone at the current moment;

[0014] Solution module: This module uses the link score and bit error rate prediction values ​​as input and uses a multivariable nonlinear optimization algorithm to solve the optimal value of the transmit power increment and critical time slot ratio of each node drone to minimize link quality degradation and bit error rate.

[0015] Execution module: used to adjust the transmission power and critical time slot ratio of each node drone in real time according to the transmission power increment and critical time slot ratio parameters output by the solution module, and trigger the neighboring drone link backup takeover when the link score or bit error rate prediction value exceeds the preset threshold.

[0016] Further: the measurement module includes a measurement unit and a fusion calculation unit;

[0017] The measurement unit includes a GNSS receiver and a three-axis inertial measurement unit, which are used to obtain the GNSS speed of the node drone respectively. and the body acceleration vector ;

[0018] The fusion calculation unit uses a first-order Kalman filter algorithm to calculate the GNSS velocity , the body acceleration vector and IMU integration speed Perform fusion to obtain the instantaneous speed of node drone i ;

[0019] The measurement module is further used to calculate the relative speed difference between any adjacent node drones i and j , and expressed by the two-norm operation;

[0020] Relative speed difference Perform normalization processing to generate normalized speed disturbance index , its value range is limited to the interval [0,1];

[0021] The closer it is to 0, the smaller the relative disturbance of the adjacent node drones i and j is, which can reduce the allocation of redundant link resources;

[0022] The closer it is to 1, the stronger the relative disturbance between the adjacent node drones i and j is, and the power and time slot compensation should be triggered first.

[0023] Further: Setting the speed disturbance index Threshold , used to decide subsequent link scheduling;

[0024] when When , it means that the relative motion disturbance of the adjacent node UAVs i and j is serious;

[0025] when When , it means maintaining the existing link parameters.

[0026] Furthermore: the prediction module includes an onboard micro electrochemical cluster sensor for real-time acquisition of the density of ion clusters around the node drone. ;

[0027] The prediction module uses the ion cluster density measured at the current moment As the initial value of the prediction ;

[0028] The prediction module calculates the future short-term window through the Kalman filter algorithm Predicted value of ion cluster density within ;

[0029] The prediction module predicts the ion cluster density Perform normalization to generate a normalized density disturbance index , its value range is limited to the interval [0,1];

[0030] set up The prediction threshold is , used to determine whether to trigger subsequent link intervention;

[0031] when When , the intervention signal is output;

[0032] when When , the current link parameters are maintained.

[0033] Further: the calculation module includes a link score calculation unit and a bit error rate prediction calculation unit, which are used to calculate the link score according to the speed disturbance index and density perturbation index Generate quantitative evaluation values;

[0034] The link score calculation unit inputs the speed disturbance index of the neighboring UAV ; Bit error rate prediction calculation unit input density disturbance index ;

[0035] The link score calculation unit is used to calculate the link score of each node drone ,in By perturbing the speed of neighboring drones It is obtained by averaging and linearly mapping to the interval (0,1).

[0036] Further: the bit error rate prediction calculation unit is based on the density disturbance index Perform secondary mapping to generate node drone i in the future short time window The bit error rate prediction value ;

[0037] Score the link and the bit error rate prediction value Represented in vector form as .

[0038] Furthermore: the solution module includes an optimization calculation unit and a result output unit, which is used to perform nonlinear optimization based on the input link score and bit error rate prediction value and output the optimal solution; specifically includes:

[0039] Optimized compute units receive vector input via shared memory , output The result output unit then sends it to the power control and time slot scheduling module;

[0040] is the transmit power increment of node UAV i; is the critical time slot proportion of node drone i;

[0041] The optimization calculation unit performs joint optimization on the transmission power increment and critical time slot ratio of all node drones based on nonlinear target optimization to finally obtain and The optimal solution ;

[0042] The result output unit will solve the optimal solution Sent to the communication controller of each node drone i.

[0043] Further: the execution module includes a configuration sending unit and a backup switching unit, which are used to send the transmit power increment and the critical time slot ratio respectively and execute the link backup switching when the trigger condition is met;

[0044] The configuration delivery unit reads the optimal solution from the solution module shared memory ,The backup switching unit monitors link quality and bit error rate alarms.

[0045] Further: the configuration issuing unit calls the RF front-end firmware API to set the actual transmission power of the node drone i and the proportion of critical time slots ;

[0046] The backup switching unit scores the link or bit error rate prediction value When the preset threshold is exceeded, a link switching command is generated;

[0047] Setting link score and the bit error rate prediction value The preset thresholds are and ;

[0048] like or , then set the switch event flag; otherwise clear the event flag;

[0049] After the event flag is set, the backup switching unit instructs the neighboring drone j to take over the link in relay mode.

[0050] Furthermore: the response delay of the backup switching unit is inversely proportional to the severity of the trigger threshold;

[0051] The severity of the trigger threshold is the sum of the "relative difference between the link score and the minimum allowed level" and the "relative difference between the bit error rate prediction value and the maximum allowed value";

[0052] The stricter the trigger threshold, the more seriously the link score or bit error rate exceeds the standard, and the shorter the required response delay.

[0053] Compared with the prior art, the present invention has the following beneficial effects:

[0054] The present invention preprocesses velocity disturbances and ion density disturbances in the measurement module, outputs normalized disturbance indicators in real time in the prediction module, and uses a multivariable nonlinear optimization algorithm in the calculation and solution module to synchronously determine the transmission power increment, the proportion of critical time slots, and the link switching strategy. Finally, the execution module issues three types of instructions in parallel and quickly completes the backup takeover. This achieves multi-factor coupling suppression and meets the application requirements of high-reliability telemetry for UAV formations in canyons under high-voltage transmission lines. BRIEF DESCRIPTION OF THE DRAWINGS

[0055] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are merely embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on the provided drawings without paying any creative work.

[0056] Figure 1 It is a schematic diagram of the overall system module of the present invention.

[0057] Figure 2 This is a schematic diagram of the present invention being applied to UAV formation cruising in a canyon under a high-voltage transmission line. DETAILED DESCRIPTION

[0058] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0059] Example 1:

[0060] See also Figure 1 and Figure 2 , the present invention provides a technical solution:

[0061] The AI-based UAV telemetry link interference and suppression system is applied to UAV formation cruising scenarios in canyons under high-voltage transmission lines. Specifically, it includes:

[0062] Measurement module: used to measure the relative speed difference between each node UAV in the UAV formation and its neighboring UAVs, and pre-process the relative speed difference to generate a normalized speed disturbance index;

[0063] Further explanation: the measurement module includes a measurement unit and a fusion calculation unit;

[0064] 1.1) The measurement unit includes a GNSS receiver and a three-axis inertial measurement unit (IMU), which are used to obtain the GNSS speed of the node drone respectively. and the body acceleration vector ; Specifically include:

[0065] GNSS receiver: collects the latitude and longitude and ground speed of the node drone after differential correction , update frequency 1Hz, positioning accuracy ±0.5m.

[0066] Three-axis inertial measurement unit IMU: measures the body acceleration vector , sampling frequency 200Hz, acceleration accuracy ±0.02m / s².

[0067] The GNSS and IMU are connected to the flight control mainboard via the SPI bus, with a data synchronization error of ≤1ms, ensuring data consistency in subsequent fusion calculations.

[0068] 1.2) The fusion calculation unit uses the first-order Kalman filter algorithm to calculate the GNSS velocity , the body acceleration vector and IMU integration speed Perform fusion to obtain the instantaneous speed of node drone i ; The specific calculation logic includes:

[0069] Integration speed The calculation formula is:

[0070]

[0071] Where, is the acceleration amplitude measured by the IMU;

[0072] is the IMU sampling period; in this embodiment Second;

[0073] The Kalman filter fusion steps are as follows:

[0074]

[0075] Where, is the filter gain, which depends on the covariance ratio of GNSS and IMU; is the instantaneous speed of node UAV i;

[0076] when Increases, the GNSS speed is more trusted; when If it decreases, the IMU data will be more trusted;

[0077] The covariance matrix is ​​estimated online using the least squares method to minimize the mean square error of the fusion speed and ensure .

[0078] 1.3) The measurement module is further used to calculate the relative speed difference between any adjacent node drones i and j , expressed through the two-norm operation; the specific expression logic includes:

[0079]

[0080] Where, 、 is the instantaneous velocity after fusion in step 1.2); is the two-norm operation;

[0081] When the relative speed difference Increase from 0 to the maximum theoretical speed difference When , it means the relative motion disturbance between the two UAVs changes from zero to strong;

[0082] Set relative speed difference When the speed increases by 1m / s, the link score decreases. , used for subsequent link optimization.

[0083] 1.4) Relative speed difference Perform normalization processing to generate normalized speed disturbance index , its value range is limited to the interval [0,1]; the specific normalization steps are as follows:

[0084]

[0085] Where, is the maximum theoretical speed difference preset by the system. The value is 20m / s;

[0086] when hour, ;when hour, ;

[0087] The closer it is to 0, the smaller the relative disturbance of the adjacent node drones i and j is, which can reduce the allocation of redundant link resources;

[0088] The closer it is to 1, the stronger the relative disturbance between the adjacent node drones i and j is, and the power and time slot compensation should be triggered first; the higher the demand for triggering link backup and power and time slot "double compensation" is;

[0089] 1.5) Set the speed disturbance index Threshold , used to decide subsequent link scheduling; the specific implementation steps are as follows:

[0090] This embodiment sets It should be noted that It is determined based on the fuzzy analytic hierarchy process (FAHP) or by using the median of the average value of historical data, which will not be elaborated here;

[0091] when When , it means that the relative motion disturbance of the adjacent node UAVs i and j is serious;

[0092] when When , it means maintaining the existing link parameters.

[0093] Prediction module: used to collect the ion cluster density of the node drone at the current moment in real time, and use the Kalman filter algorithm to online predict the ion cluster density prediction value in the future short-term window, and pre-process the ion cluster density prediction value to generate a normalized density disturbance index;

[0094] Further explanation: 2.1) The prediction module includes an onboard micro electrochemical cluster sensor for real-time collection of the ion cluster density around the node drone. ; The specific implementation steps are as follows:

[0095] The measuring range is 0~10 6 ions / cm³, with a typical measurement error of ±5%.

[0096] The micro electrochemical cluster sensor is fixed on the top leading edge of the node drone to ensure the stability of the sampling airflow and avoid the influence of body disturbance.

[0097] Connected to the flight control mainboard via the I²C bus, the sampling frequency is 1Hz, and the data frame synchronization delay is ≤10ms.

[0098] The original measured ion cluster density De-noising is performed using a 3-point sliding average filter to ensure that the data jitter amplitude is ≤2%.

[0099] 2.2) The prediction module uses the current measured ion cluster density As the initial value of the prediction Specific implementation includes:

[0100]

[0101] Where, is the initial value of the predicted state used for Kalman filtering;

[0102] After receiving each new sample data, it is updated immediately , ensuring that the predicted status is consistent with the latest environmental conditions.

[0103] 2.3) The prediction module calculates the future short-term window through the Kalman filter algorithm Predicted value of ion cluster density within ; The specific implementation steps are as follows:

[0104]

[0105] Where, is the Kalman gain, which is adjusted online according to the sensor noise covariance R and the process noise covariance Q;

[0106] when The closer it is to 1, the closer the prediction result is to the real-time measurement; when The closer it is to 0, the smoother the prediction result;

[0107] is the predicted delay. The value is 0.5s;

[0108] 2.4) Using the standard Kalman gain update formula:

[0109]

[0110]

[0111] in is the prior error covariance, is the posterior error covariance.

[0112] Every time you get Afterwards, according to Update the Kalman gain and calculate ;

[0113] 2.5) The prediction module predicts the density of ion clusters Perform normalization to generate a normalized density disturbance index , whose value range is limited to the interval [0,1]; the implementation method is as follows:

[0114]

[0115] Where, Is the upper limit of the sensor range; in this embodiment , The closer it is to 0, the lower the environmental interference; The closer it is to 1, the closer the interference density is to the upper limit of the range.

[0116] set up The prediction threshold is , used to determine whether to trigger subsequent link intervention; this embodiment sets It should be noted that It is determined based on the fuzzy analytic hierarchy process (FAHP) or by using the median of the average value of historical data, which will not be elaborated here;

[0117] when When , the intervention signal is output;

[0118] when When , the current link parameters are maintained.

[0119] Calculation module: used to calculate the link score of each node drone based on the speed disturbance index, and calculate the bit error rate prediction value based on the density disturbance index of each node drone at the current moment;

[0120] Further explanation: The calculation module includes a link score calculation unit and a bit error rate prediction calculation unit, which are used to calculate the link score according to the speed disturbance index. and density perturbation index Generate quantitative evaluation values;

[0121] In this embodiment, a dual-core ARM Cortex-A53 microprocessor is used. The main core is responsible for link score calculation, and the auxiliary core is responsible for bit error rate prediction calculation. The two cores exchange intermediate data through shared memory, and the calculation delay is ≤5ms.

[0122] Two independent tasks are deployed in the flight control real-time operating system (RTOS), with the task priority of link scoring > bit error rate prediction, to ensure that the scoring takes precedence when the speed disturbance is severe.

[0123] The link score calculation unit inputs the speed disturbance index of the neighboring UAV ; Bit error rate prediction calculation unit input density disturbance index ;

[0124] The link score calculation unit is used to calculate the link score of each node drone ,in By perturbing the speed of neighboring drones The average is calculated and linearly mapped to the interval (0,1); the specific steps are as follows:

[0125]

[0126] Where, is the set of neighboring drones of node drone i, and the total number of drones in the neighboring drone set is N;

[0127]

[0128] In the formula, when The closer it is to 0, The closer it is to 1;

[0129] when The closer it is to 1, The closer it is to 0;

[0130] The bit error rate prediction calculation unit is based on the density disturbance index Perform secondary mapping to generate node drone i in the future short time window The bit error rate prediction value ; The specific implementation steps are as follows:

[0131] Define the quadratic mapping formula as ;Where, output ;

[0132] when The closer it is to 0, the higher the bit error rate prediction value The closer it is to 0;

[0133] when The closer it is to 1, the higher the bit error rate prediction value The closer it is to 1, the greater the sensitivity to high interference due to secondary mapping.

[0134] The secondary mapping of this embodiment is simple and feasible, highlights the trend of a sharp increase in bit error rate under high-density interference, and has low consumption of hardware computing resources.

[0135] Score the link and the bit error rate prediction value Represented in vector form as .

[0136] Solution module: This module uses the link score and bit error rate prediction values ​​as input and uses a multivariable nonlinear optimization algorithm to solve the optimal value of the transmit power increment and critical time slot ratio of each node drone to minimize link quality degradation and bit error rate.

[0137] Further explanation: The solution module includes an optimization calculation unit and a result output unit, which is used to perform nonlinear optimization based on the input link score and bit error rate prediction value and output the optimal solution; specifically includes:

[0138] The optimization calculation unit of this embodiment uses a DSP processor with a main frequency of 1.2GHz and an integrated floating-point operation unit to ensure that the optimization iteration delay is ≤50ms;

[0139] Deploy an optimization task and an output task in the flight control RTOS. The optimization task has a higher priority than the output task.

[0140] Optimized compute units receive vector input via shared memory , output The result output unit then sends it to the power control and time slot scheduling module;

[0141] is the transmit power increment of node UAV i; is the critical time slot proportion of node drone i;

[0142] The optimization calculation unit performs joint optimization on the transmission power increment and critical time slot ratio of all node drones based on nonlinear target optimization to finally obtain and The optimal solution ; The specific logic includes:

[0143] The nonlinear objective optimization function is:

[0144]

[0145] Where: is the maximum transmit power increment allowed by the system; in this embodiment ; , , , is the non-negative weight coefficient of the corresponding parameter, and the example value is =0.5, =0.3, =0.1, =0.1.

[0146] If the link score Decrease, then Increase, drive or Increase;

[0147] like Increase, then Increase, also drives resource allocation;

[0148] like or Increase, then the respective penalty terms 、 Increase and restrict over-allocation.

[0149] The optimization calculation unit uses the Sequential Quadratic Programming (SQP) algorithm to solve the above nonlinear programming problem to balance solution accuracy and real-time performance. The specific implementation steps include:

[0150] All and The initial value is set to 0.1× the maximum value;

[0151] The SQP iteration content is as follows:

[0152] Linearize the objective and constraints at the current point; solve the quadratic subproblem to obtain the step size and direction;

[0153] Update the variables until the objective function converges or the number of iterations reaches a maximum of 20;

[0154] When the objective function decreases Converges early;

[0155] The optimization calculation unit imposes the following constraints on each variable during the solution process:

[0156]

[0157]

[0158]

[0159] For the transmit power increment constraint: the projection operation is used to ensure that after each iteration Maintain in [0,20]dBm;

[0160] For the critical time slot proportion constraint: add a linear inequality to the quadratic subproblem ;

[0161] The result output unit will solve the optimal solution The implementation method is as follows: Convert to floating point number, Convert to percentage;

[0162] The CAN bus is sent at a rate of 10 Hz, and each frame contains the data of a single-node drone i. ;

[0163] After receiving the message, each node drone returns a confirmation message to the optimization calculation unit. If the communication fails, the last parameters are retained.

[0164] Execution module: used to adjust the transmission power and critical time slot ratio of each node drone in real time according to the transmission power increment and critical time slot ratio parameters output by the solution module, and trigger the neighboring drone link backup takeover when the link score or bit error rate prediction value exceeds the preset threshold.

[0165] Further explanation: The execution module includes a configuration sending unit and a backup switching unit, which are used to send the transmit power increment and the critical time slot ratio respectively and perform link backup switching when the trigger condition is met; the implementation method includes:

[0166] The configuration issuing unit and the backup switching unit both run on the ARMCortex-M4 real-time controller;

[0167] In the flight control RTOS, it is divided into two task chains - configuration task and switching task, and the tasks communicate through event flags;

[0168] The configuration delivery unit reads the optimal solution from the solution module shared memory ,The backup switching unit monitors link quality and bit error rate alarms;

[0169] Furthermore, the configuration issuing unit calls the RF front-end firmware API to set the actual transmission power of the node drone i and the proportion of critical time slots ; The implementation method is:

[0170]

[0171] Where, is the baseline transmit power. The value is 10dBm;

[0172] Directly Passed into the firmware API and converted into a percentage duty cycle;

[0173] The API is called immediately after each optimization is completed and repeatedly issued at a rate of 100 Hz to ensure that the firmware parameters are synchronized with the algorithm output.

[0174] Furthermore, the backup switching unit is used to score the link or bit error rate prediction value When the preset threshold is exceeded, a link switching command is generated; the implementation method is:

[0175] Setting link score and the bit error rate prediction value The preset thresholds are and ; This embodiment ;

[0176] like or , then set the switching event flag; otherwise, clear the event flag; specifically use the RTOS event group API to detect the flag and ensure that the detection cycle is ≤10ms.

[0177] Furthermore, after the event flag is set, the backup switching unit instructs the neighboring UAV j to take over the link in relay mode and completes the switching within 50ms. The implementation method is as follows:

[0178] The switch command format is ;

[0179] Sent within 0–10ms after receiving the switch event flag ;

[0180] Within 10ms–40ms, the neighboring drone j starts the relay firmware module; within 40ms–50ms, the link access is confirmed and the control signal routing is switched;

[0181] The neighboring drone j sends within 50ms , otherwise try again.

[0182] Furthermore, the response delay of the backup switching unit is inversely proportional to the severity of the trigger threshold;

[0183] The severity of the trigger threshold is the sum of the "relative difference between the link score and the minimum allowed level" and the "relative difference between the bit error rate prediction value and the maximum allowed value".

[0184] The stricter the trigger threshold, the more seriously the link score or bit error rate exceeds the standard, and the shorter the required response delay.

[0185] The specific implementation is as follows:

[0186] The severity of the trigger threshold is defined as follows:

[0187]

[0188] In the formula, if , then the first term >0;

[0189] like , then the second term >0;

[0190] Trigger threshold severity ;

[0191] The response delay function is as follows:

[0192]

[0193] Where: is the backup switching response delay; is the maximum allowed response delay; is the response sensitivity adjustment coefficient; is the degree to which the trigger threshold is exceeded;

[0194] This embodiment The value is 50ms; The initial value is 1;

[0195] when The closer it is to 0, The closer ;when The bigger it is, The smaller the value, the faster the link switching.

[0196] By mapping the response delay function, the system can adaptively adjust the backup switching delay according to the severity of the exceeding standard, ensuring the minimum resource consumption when the standard is slightly exceeded and giving priority to ensuring communication reliability when the standard is seriously exceeded; the response sensitivity adjustment coefficient Controls the steepness of the mapping curve and optimizes the configuration based on the hardware switching capability and energy budget.

[0197] Example 2:

[0198] Six drones of the same model (A–F) were deployed in a canyon power transmission line area. Each drone integrated the measurement module, solution module, and execution module firmware. The measurement module collected link scores in real time. and the bit error rate prediction value ; The solution module outputs one-time based on sequential quadratic programming:

[0199] Transmit power increment ; Critical time slot ratio ;

[0200] Backup switch command : or .

[0201] The execution module includes a configuration sending unit and a backup switching unit, running on an ARM Cortex-M4 controller and using the flight control RTOS task chain for parallel processing. It calls the RF front-end API and repeatedly sends it via the CAN bus at 100Hz, with an API response delay of ≤1ms and a synchronization error of ≤2ms. The switching process can be completed within 50ms after joint debugging. →Relay Start→ shake hands.

[0202] Before the flight, the six aircraft were to conduct vertical cruising at a height of 30m with a formation spacing of 50m, and to cross a 200m cliff to simulate a high ionization and motion disturbance environment. .

[0203] Verification of linkage command issuance:

[0204] Send randomly generated dBm vs. , and is issued when the trigger conditions are met .

[0205] Verify the correctness of API calls, CAN frame synchronization, and switching commands.

[0206] During normal cruise, ground equipment is used to dynamically reduce the signal or inject ionization interference to make or Approaching or exceeding the limit respectively.

[0207] Record each triggering moment, adjacent drone relay start and Receipt time, calculate the overall backup switching response delay .

[0208] Repeat the whole process 10 times, take the average value, and observe the response delay distribution under different over-limit levels; , and The coupling changes: when When the increase is 10%, the average When the response is increased by 0.05 and decreased by 5%, the switching probability increases by 15%. Adjusted to 1.2.

[0209] Table 1 Feasibility study of execution modules:

[0210]

[0211] By comparing with the traditional single RF switching solution, the following average indicators are obtained:

[0212] For link availability:

[0213] Traditional solution: 98.7%; this solution: 99.3%, an increase of 0.6%.

[0214] For bit error rate:

[0215] Traditional solution: 0.0021; this solution: 0.00087, a decrease of 59%.

[0216] For backup switching delay:

[0217] Traditional solution: 60ms; this solution: 41.2ms, a 31% reduction.

[0218] The above improvements exceed the design goals of availability ≥0.5%, BER reduction ≥50%, and recovery time reduction ≥30%, fully demonstrating that the present invention has significant innovation and practicality compared with existing solutions through the coordinated optimization of power, time slots and switching instructions.

[0219] The selected technical route of "linked link backup and power / time slot adaptive scheduling based on formation relative velocity coupling and atmospheric ion cluster density prediction" has the following advantages:

[0220] Dual compensation for motion disturbance and environmental interference: taking into account the two major sources of interference, formation speed difference and ion cluster density;

[0221] Prediction-driven advance backup: Kalman filtering is used to predict short-term ion density, achieving second-level advance backup links.

[0222] Linked power + time slot + link switching: One optimization model simultaneously generates three types of control instructions to ensure link stability.

[0223] It should be noted that: Figure 2 The number "1" in the figure represents the schematic diagram of three types of control instructions; the number "2" represents the node drone.

[0224] It should be noted that: All calculation formulas in this application document use regression analysis including but not limited to machine learning algorithms to deeply analyze the relevant parameters collected and identify their natural trends and relationships. Use professional software, such as Python's Scikit-learn library or R language, to automatically generate mathematical models that match the data. Then, objectively evaluate the performance of the model through methods such as cross-validation, and combine continuous feedback and optimization to ensure that the created formula truly reflects the inherent laws of the data, thereby ensuring its effectiveness and accuracy. In all calculation formulas in this application, the parameters in each formula are dimensionlessly processed in a consistent range to ensure that different physical quantities are compared on the same scale; dimensionless technical means include but are not limited to Min-Max Normalization and Z-Score normalization;

[0225] The technical solution of the present invention, or the part that contributes to the prior art, can be embodied in the form of a software product. The computer software product can be stored in a computer-readable storage medium, such as a computer floppy disk, read-only memory (ROM), random access memory (RAM), flash memory (FLASH), hard disk or optical disk, etc., and includes a number of instructions for enabling a computer device (which can be a personal computer, server, or network device, etc.) to execute the methods of various embodiments of the present invention.

[0226] The logic and / or steps represented in the flowcharts or otherwise described herein, for example, can be considered as a sequenced list of executable instructions for implementing the logical functions, and can be embodied in any computer-readable medium for use by, or in conjunction with, an instruction execution system, apparatus, or device (e.g., a computer-based system, a system including a processor, or other system that can fetch and execute instructions from an instruction execution system, apparatus, or device). For purposes of this specification, a "computer-readable medium" can be any device that can contain, store, communicate, propagate, or transport a program for use by, or in conjunction with, an instruction execution system, apparatus, or device.

[0227] The above embodiments are only intended to help understand the method and core concept of the present invention. It should be noted that, without departing from the principles of the present invention, a number of improvements and modifications may be made to the present invention by those skilled in the art, and such improvements and modifications also fall within the scope of protection of the claims of the present invention.

Claims

1. An artificial intelligence-based UAV telemetry link interference and suppression system, applied to the canyon UAV formation cruising scenario under high-voltage transmission lines, characterized by: Specifically include: Measurement module: used to measure the relative speed difference between each node UAV in the UAV formation and its neighboring UAVs, and pre-process the relative speed difference to generate a normalized speed disturbance index; Prediction module: used to collect the ion cluster density of the node drone at the current moment in real time, and use the Kalman filter algorithm to online predict the ion cluster density prediction value in the future short-term window, and pre-process the ion cluster density prediction value to generate a normalized density disturbance index; Calculation module: used to calculate the link score of each node drone based on the speed disturbance index, and calculate the bit error rate prediction value based on the density disturbance index of each node drone at the current moment; Solution module: This module uses the link score and bit error rate prediction values ​​as input and uses a multivariable nonlinear optimization algorithm to solve the optimal value of the transmit power increment and critical time slot ratio of each node drone to minimize link quality degradation and bit error rate. Execution module: used to adjust the transmission power and critical time slot ratio of each node drone in real time according to the transmission power increment and critical time slot ratio parameters output by the solution module, and trigger the neighboring drone link backup takeover when the link score or bit error rate prediction value exceeds the preset threshold.

2. The artificial intelligence-based UAV telemetry link interference and suppression system according to claim 1 is characterized by: The measurement module includes a measurement unit and a fusion calculation unit; The measurement unit includes a GNSS receiver and a three-axis inertial measurement unit, which are used to obtain the GNSS speed of the node drone respectively. and the body acceleration vector ; The fusion calculation unit uses a first-order Kalman filter algorithm to calculate the GNSS velocity , the body acceleration vector and IMU integration speed Perform fusion to obtain the instantaneous speed of node drone i ; The measurement module is further used to calculate the relative speed difference between any adjacent node drones i and j , and expressed by the two-norm operation; Relative speed difference Perform normalization processing to generate normalized speed disturbance index , its value range is limited to the interval [0,1]; The closer it is to 0, the smaller the relative disturbance of the adjacent node drones i and j is, which can reduce the allocation of redundant link resources; The closer it is to 1, the stronger the relative disturbance between the adjacent node drones i and j is, and the power and time slot compensation should be triggered first.

3. The artificial intelligence-based UAV telemetry link interference and suppression system according to claim 2 is characterized by: Set the speed disturbance indicator Threshold , used to decide subsequent link scheduling; when When , it means that the relative motion disturbance of the adjacent node UAVs i and j is serious; when When , it means maintaining the existing link parameters.

4. The artificial intelligence-based UAV telemetry link interference and suppression system according to claim 3 is characterized by: The prediction module includes an onboard micro electrochemical cluster sensor for real-time acquisition of the ion cluster density around the node drone. ; The prediction module uses the ion cluster density measured at the current moment As the initial value of the prediction ; The prediction module calculates the future short-term window through the Kalman filter algorithm Predicted value of ion cluster density within ; The prediction module predicts the ion cluster density Perform normalization to generate a normalized density disturbance index , its value range is limited to the interval [0,1]; set up The prediction threshold is , used to determine whether to trigger subsequent link intervention; when When , the intervention signal is output; when When , the current link parameters are maintained.

5. The artificial intelligence-based UAV telemetry link interference and suppression system according to claim 4 is characterized in that: The calculation module includes a link score calculation unit and a bit error rate prediction calculation unit, which are used to calculate the link score according to the speed disturbance index. and density perturbation index Generate quantitative evaluation values; The link score calculation unit inputs the speed disturbance index of the neighboring UAV ; Bit error rate prediction calculation unit input density disturbance index ; The link score calculation unit is used to calculate the link score of each node drone ,in By perturbing the speed of neighboring drones It is obtained by averaging and linearly mapping to the interval (0,1).

6. The artificial intelligence-based UAV telemetry link interference and suppression system according to claim 5 is characterized by: The bit error rate prediction calculation unit is based on the density disturbance index Perform secondary mapping to generate node drone i in the future short time window The bit error rate prediction value ; Score the link and the bit error rate prediction value Represented in vector form as .

7. The artificial intelligence-based UAV telemetry link interference and suppression system according to claim 6 is characterized by: The solution module includes an optimization calculation unit and a result output unit, which is used to perform nonlinear optimization based on the input link score and bit error rate prediction value and output the optimal solution; specifically, it includes: Optimized compute units receive vector input via shared memory , output The result output unit then sends it to the power control and time slot scheduling module; is the transmit power increment of node UAV i; is the critical time slot proportion of node drone i; The optimization calculation unit performs joint optimization on the transmission power increment and critical time slot ratio of all node drones based on nonlinear target optimization to finally obtain and The optimal solution ; The result output unit will solve the optimal solution Sent to the communication controller of each node drone i.

8. The artificial intelligence-based UAV telemetry link interference and suppression system according to claim 7, characterized in that: The execution module includes a configuration sending unit and a backup switching unit, which are used to send the transmit power increment and the critical time slot ratio respectively and execute the link backup switching when the trigger condition is met; The configuration delivery unit reads the optimal solution from the solution module shared memory ,The backup switching unit monitors link quality and bit error rate alarms.

9. The artificial intelligence-based UAV telemetry link interference and suppression system according to claim 8, characterized in that: The configuration issuing unit calls the RF front-end firmware API to set the actual transmit power of the node drone i and the proportion of critical time slots ; The backup switching unit scores the link or bit error rate prediction value When the preset threshold is exceeded, a link switching command is generated; Setting link score and the bit error rate prediction value The preset thresholds are and ; like or , then set the switch event flag; otherwise clear the event flag; After the event flag is set, the backup switching unit instructs the neighboring drone j to take over the link in relay mode.

10. The artificial intelligence-based UAV telemetry link interference and suppression system according to claim 9, characterized in that: The response delay of the backup switching unit is inversely proportional to the severity of the trigger threshold; The severity of the trigger threshold is the sum of the "relative difference between the link score and the minimum allowed level" and the "relative difference between the bit error rate prediction value and the maximum allowed value"; The stricter the trigger threshold, the more seriously the link score or bit error rate exceeds the standard, and the shorter the required response delay.

Images