An information flow processing method based on a drone countermeasure warning dialogue specification

Abstract

The application discloses a kind of information flow processing methods based on unmanned plane countermeasure warning dialogue specification, comprising: S1, the original flight data stream of unmanned plane is collected, is matched to airspace hierarchical grid and binds corresponding risk level and supporting interactive process template, generates standardization original situation data stream;S2, according to standardization original situation data stream extraction flight characteristic and calculate boundary intrusion probability, combine grid position with boundary intrusion probability to judge violation level, according to violation level matching corresponding interactive process and output standardization warning data stream;S3, the feedback data stream of unmanned plane is collected, combines reaction time delay and motion characteristic to calculate maliciousness score, according to maliciousness score matching differentiated disposal strategy, output disposal strategy data stream;S4, based on disposal strategy data stream monitoring unmanned plane evacuation, full-link interactive data and calculation result are encrypted and archived, output security archival data stream.The application forms the full-link closed loop of warning, feedback, evaluation, disposal.

Description

Technical Field

[0001] This invention relates to the field of drone countermeasures technology, and in particular to an information flow processing method based on drone countermeasure warning dialogue specifications. Background Technology

[0002] In handling drones illegally entering sensitive airspace, warning notifications are a crucial preliminary step in countermeasures, and a key means of distinguishing compliant drones and reducing the risk of accidental damage from mishandling. While existing drone countermeasure warning standards provide unified quantitative constraints on the entire warning interaction chain, current commercial countermeasure equipment lacks a standardized information flow processing solution, exhibiting two major shortcomings: First, the interaction process is based solely on static grading of spatial location, failing to consider the drone's behavioral response to warnings to determine its subjective intrusion intent, resulting in a rigid handling strategy. Existing solutions uniformly send warnings at fixed intervals and wait for feedback for fixed durations. This provides insufficient reaction window for unintentionally intruding drones, easily leading to disputes arising from mishandling; for maliciously intruding drones, repeated warnings delay the opportunity for intervention, failing to achieve differentiated control of "guided but unintentional intrusions and rapid handling of malicious intrusions," and also making it difficult to adapt to the flexible implementation requirements of tiered interaction in the standard.

[0003] Second, there is no unified packaging mechanism for the warning content. The warning information output by the equipment is missing fields and the description is vague. It does not include key traceability fields such as the device ID, prevention and control code, risk level, and quantitative time limit that are required by the standard. It cannot automatically adapt to Chinese and English. The warning instructions received by the operator are unclear, making it difficult to quickly complete the evacuation, reporting, and appeal operations, which further increases the probability of mishandling.

[0004] The applicant proposed improvement measures to address the aforementioned issues. Summary of the Invention

[0005] To address the two major technical problems of existing countermeasures equipment warning processes being static and rigid, unable to distinguish between subjective malice, and lacking a unified and compliant encapsulation mechanism for warning content, this invention proposes an information flow processing method based on the drone countermeasures warning dialogue specification. It introduces a subjective malice quantification assessment mechanism based on behavioral response characteristics and a corresponding standardized warning content encapsulation mechanism. This method not only ensures the reaction window for unintentionally intruding drones and reduces the risk of accidental damage, but also improves the efficiency of handling malicious intrusion scenarios, fully implementing local standard requirements.

[0006] This invention proposes an information flow processing method based on the UAV counter-warning dialogue specification, comprising: S1. Collect raw flight data streams of UAVs, match them to the airspace hierarchical grid according to the real-time coordinates of the UAVs, bind the corresponding risk level and the corresponding interactive process template, and generate standardized raw situation data streams. S2. Extract flight features from the standardized raw situation data stream, calculate the boundary intrusion probability based on the flight features, determine the violation level by combining the grid position and the boundary intrusion probability, match the corresponding interaction process from the interaction process template according to the violation level, and output a standardized warning data stream. S3. Collect feedback data streams from the drone, calculate reaction delay based on the sending time of the standardized warning data stream and the receiving time of the feedback data stream, calculate the malice score by combining the reaction delay and motion characteristics, match differentiated handling strategies based on the malice score, and output the handling strategy data stream. S4. Based on the aforementioned disposal strategy, monitor the drone's withdrawal via data stream, encrypt and archive the entire link interaction data and calculation results, and output a secure archived data stream.

[0007] Furthermore, the flight characteristics include normalized position offset rate, heading variability rate, altitude change rate, normal acceleration variability rate, radial approach rate, and the angle between the heading and the boundary normal.

[0008] Furthermore, the calculation of the boundary intrusion probability includes: linearly weighting the flight features by a preset fixed weight to obtain an initial boundary intrusion probability, and then correcting the initial boundary intrusion probability by combining it with the risk coefficient corresponding to the current grid position of the UAV to obtain the final boundary intrusion probability.

[0009] Furthermore, the airspace hierarchical grid is divided into a handling zone, an identification zone, and an early warning zone from the inside out according to the boundary of the identification zone; the correspondence between the risk level and the airspace hierarchical grid is as follows: the handling zone corresponds to level 1 risk, the identification zone corresponds to level 2 risk, the early warning zone corresponds to level 3 risk, and the monitoring zone corresponds to level 4 risk; the method for judging the violation level is as follows: the risk level bound to the current airspace hierarchical grid of the UAV is used as the baseline level. When the boundary intrusion probability is not lower than the first threshold, the baseline level is increased by one level; when the boundary intrusion probability is not higher than the second threshold, the baseline level is decreased by one level; otherwise, the baseline level remains unchanged.

[0010] Furthermore, if the violation level is determined to be a level four risk, a monitoring process will be adopted, which will only monitor the flight status of the drone. If the violation level is determined to be a level three risk, a one-way notification process is adopted, with a total interaction time limit of t1 seconds. The standardized warning data stream includes sending a warning notification, informing the boundary of the prevention and control area, the permitted flight range and the reporting channel. The sending interval adopts a differentiated non-uniform sending strategy based on the boundary intrusion probability. If the violation level is determined to be Level 2 risk, a two-way confirmation process is adopted, with a total interaction time limit of t2 seconds and a fixed sending interval of t3 seconds. The standardized warning data stream includes sending a formal warning notification, inquiring about the drone's identity and flight purpose, clarifying the basis for the violation determination, and quantifying the evacuation route and time limit requirements. The drone is required to provide feedback on receipt confirmation, respond to inquiries, and operational responses. If the risk is determined to be Level 1, an emergency procedure is adopted, and the standardized warning data flow includes simultaneously sending a countermeasure advance notice and initiating countermeasure preparation, clearly defining the countermeasure type, initiation time, and safety prompts; without waiting for the complete feedback process, the synchronous linkage tracking mechanism locks the drone's takeoff location and the operator's area.

[0011] Furthermore, based on the violation level, a corresponding standard template for the interaction process is matched, and the countermeasure device ID, prevention and control area code, globally unique warning number, graded risk level, quantified evacuation coordinates, precise time limit, and feedback channel are filled in; a unified verification maximum number of characters per warning message is set, and the Chinese and English languages ​​of the drone registration are automatically matched, and the location and time instructions are quantified; after the content is filled in, an SHA-256 digital signature is attached to generate a standardized warning data stream.

[0012] Furthermore, the calculation of the malice score includes: Acquire continuous multi-frame flight status data of the UAV after the standardized warning data stream is sent; The reaction time delay, velocity change rate, heading deviation, distance change rate from the boundary of the disposal area, and normal acceleration were extracted as features for malice assessment. After normalizing and mapping each maliciousness assessment feature, the maliciousness score is obtained by weighting and summing the features according to the preset maliciousness weights.

[0013] Furthermore, the differentiated handling strategy is divided according to the range of malice score: when the malice score is not lower than 0.7, countermeasures are initiated; when the malice score is between 0.3 and 0.7, a continuous warning and tracking strategy is implemented; and when the malice score is lower than 0.3, an observation and guidance strategy is implemented.

[0014] Furthermore, S2 also includes a whitelist verification step: matching the drone's identity information with a preset whitelist. If the match is successful, the violation level determination and subsequent handling process are skipped, and a compliant release data stream is output.

[0015] Furthermore, the raw flight data stream of the UAV is publicly available flight status data and broadcast communication message data of the target UAV collected within the legally controlled airspace, excluding the operator's personal privacy and non-public identity information.

[0016] This invention introduces a two-tiered evaluation mechanism of "boundary intrusion probability" and "subjective malice" into the warning interaction process. It distinguishes between two fundamentally different scenarios: unintentional intrusion and malicious intrusion, and matches differentiated handling paths for guided evacuation and rapid countermeasures accordingly. This overcomes the static control deficiencies of fixed intervals and durations in traditional solutions, achieving the hierarchical control goal of "standardized guidance for unintentional intrusions and rapid handling for malicious intrusions." Furthermore, this invention establishes a standardized, compliant, and traceable warning content packaging system, ensuring that every warning message issued to the operator conforms to local standard format requirements, eliminating common problems such as missing fields and vague descriptions. This provides operators with clear, quantifiable, and actionable handling guidelines, effectively reducing the risk of mishandling due to unclear information. Attached Figure Description

[0017] Figure 1 This is a flowchart of an information flow processing method based on the UAV counter-warning dialogue specification proposed in this invention. Detailed Implementation

[0018] The technical solution of the present invention will now be described in detail with reference to the accompanying drawings and embodiments.

[0019] like Figure 1 As shown, the present invention proposes an information flow processing method based on the UAV counter-warning dialogue specification, comprising: S1. Collect the raw flight data stream of the UAV, match it to the airspace hierarchical grid according to the real-time coordinates of the UAV, and bind the corresponding risk level and the corresponding interaction process template to generate a standardized raw situation data stream.

[0020] Specifically, multi-sensor data acquisition hardware units (such as radar acquisition circuits, photoelectric image acquisition circuits, and radio spectrum analysis circuits) collect raw flight data streams of UAVs in the controlled airspace. This includes the UAV ID, real-time latitude and longitude coordinates, altitude, flight speed, normal acceleration, heading angle, heading rate of change, broadcast communication messages, and the code of the controlled airspace. During this process, only the UAV's flight status and publicly broadcast information are collected; sensitive data such as the operator's personal privacy and the non-public identity information of the aircraft owner are not collected. The data collection behavior complies with public safety data collection standards, airspace control authorization, and relevant regulations on personal information protection. The data is transmitted to the central processing unit via the SPI bus.

[0021] The raw flight data stream of the UAV is preprocessed to filter out sensor noise. An airspace hierarchical grid is constructed based on the early warning boundary requirements standard. This grid involves acquiring the boundary coordinates of the controlled area and transforming it into a set of grids, specifically divided from the inside out into a response zone, an identification zone, an early warning zone, and a monitoring zone. The planar coordinates of the UAV's projection within the controlled area are matched with the grid set to confirm the UAV's current grid location and its corresponding risk level. The correspondence between risk levels and the airspace hierarchical grid is as follows: the response zone corresponds to Level 1 risk (Level 1 emergency intrusion), the identification zone corresponds to Level 2 risk (Level 2 unauthorized entry), the early warning zone corresponds to Level 3 risk (Level 3 suspected violation), and the monitoring zone corresponds to Level 4 risk.

[0022] The raw flight data stream of the UAV is stored using TLV segmented memory mapping rules. The header is bound to the grid risk level identifier. The message is parsed to extract the registered language type of the UAV and generate a standardized raw situational data stream.

[0023] Standardized raw situational data streams are stored in a memory buffer, providing a basic data source for intrusion probability assessment, hierarchical process scheduling, and compliance alert encapsulation.

[0024] S2. Extract flight features from the standardized raw situational data stream, calculate the boundary intrusion probability based on the flight features, determine the violation level by combining the grid position and the boundary intrusion probability, match the corresponding interaction process from the interaction process template according to the violation level, and output a standardized warning data stream.

[0025] Specifically, the standardized raw situational data stream is input into the Kalman filter module, which performs time-domain filtering and state estimation on position, velocity, heading, and acceleration to eliminate sensor noise and sampling jitter, and outputs a stable continuous flight state sequence; flight features are extracted from the smoothed flight state sequence.

[0026] Flight characteristics include normalized position deviation rate, heading volatility, altitude change rate, normal acceleration volatility, radial approach rate, and the angle between the heading and the boundary normal. Specifically, the position deviation rate is the ratio of the lateral deviation of the UAV's actual position relative to the compliant flight path to the control warning threshold distance, used to measure the degree of deviation from the compliant flight path; a larger ratio indicates a more significant deviation. The heading volatility rate is the ratio of the change in heading angle per unit time to the maximum permissible maneuver angle, used to measure the stability of the flight direction; a larger value indicates more frequent changes in direction and detours. The altitude change rate is the ratio of the short-term difference between ascent and descent altitude to the upper limit of the controlled altitude, used to measure the speed of the UAV's ascent or descent; a larger value indicates more severe altitude fluctuations. The normal acceleration volatility rate is also measured. Acceleration fluctuation rate is the ratio of the lateral acceleration fluctuation value during turning to the maximum safe maneuver acceleration. It is used to measure the severity of maneuvers such as turning and circling. The larger the value, the more abnormal the flight trajectory. Radial approach rate is the ratio of the UAV's approach speed toward the center of the restricted area to the maximum approach speed threshold. It is used to measure the UAV's approach speed toward the controlled area. A positive value indicates a continuous approach to the restricted area. The angle between the heading and the boundary normal is the converted value of the angle between the nose direction and the perpendicular line of the control boundary. It is used to measure the degree of alignment between the nose direction and the control boundary. The smaller the angle, the stronger the intrusion tendency.

[0027] These six features correspond to three judgment dimensions: approach intention, maneuver stability, and flight path compliance. Radial approach rate and the angle between the heading and the boundary normal are the core features, directly quantifying the objective trend of the UAV approaching the identification area. The remaining four are auxiliary features, reflecting the degree to which flight behavior deviates from the conventional flight path. The flight feature data only includes statistical values ​​and does not contain the original trajectory plaintext.

[0028] Specifically, calculating the boundary intrusion probability includes: linearly weighting flight features with preset fixed weights to obtain the initial boundary intrusion probability; then correcting the initial boundary intrusion probability by combining the risk coefficient corresponding to the current grid position of the UAV to obtain the final boundary intrusion probability. Among these, the combined weight of the two core features, radial approach rate and the angle between the heading and the boundary normal, accounts for no less than 0.6. The remaining four are auxiliary correction features. As an example, the weights w1, w2, w3, w4, w5, and w6 for position offset rate, heading fluctuation rate, altitude change rate, normal acceleration fluctuation rate, radial approach rate, and the angle between the heading and the boundary normal are 0.08, 0.07, 0.05, 0.06, 0.42, and 0.32, respectively.

[0029] Furthermore, based on the polygonal boundary of the no-fly zone, an outer boundary grid is defined by extending 500 meters outward. A grid within the identification zone is defined 300 meters inward from the boundary to the core of the restricted area. The remaining innermost area is designated as the core restricted area grid. The closer to the core protection area, the higher the correction coefficient value, amplifying the risk weight associated with the airspace location. As an example, the weight of the warning zone is 0.7, the weight of the identification zone is 0.9, and the weight of the core restricted area is 1.0.

[0030] The method for determining the violation level is as follows: The risk level bound to the current airspace grading grid of the drone is used as the baseline level. When the boundary intrusion probability is not lower than the first threshold, the baseline level is increased by one level; when the boundary intrusion probability is not higher than the second threshold, the baseline level is decreased by one level; otherwise, the baseline level remains unchanged. For example, the first threshold is 0.9, and the second threshold is 0.6. In this invention, the more closely the drone's flight behavior conforms to the characteristics of actively intruding into controlled airspace, the higher the values ​​of the normalized features, and the greater the calculated boundary intrusion probability value. This quantifies the risk level of the drone intruding into the controlled airspace and directly matches it with the corresponding graded handling procedure.

[0031] The specific content of the interaction process and standardized alert data flow corresponding to the risk level is as follows: 1. Level 4 violation: Match the monitoring process, do not interact, only monitor the drone's flight status and observe subsequent flight actions. If the drone crosses the monitoring area and flies into the warning area, then handle it according to the corresponding process of the subsequent control area.

[0032] 2. Level 3 Suspected Violation (Boundary Extension Warning Zone): Matches a one-way notification process, with a total interaction time limit of 10 seconds. The one-way notification process is driven by the device actively pushing warnings, while the feedback channel remains in a passive monitoring state. However, feedback is not a prerequisite for process advancement. If compliance reporting or evacuation confirmation feedback is received in advance, the process can be terminated directly to avoid unnecessary level escalation. If no feedback is received within the time limit, the process will automatically advance according to preset rules and will not be halted due to lack of feedback. A pre-warning template is matched according to the one-way notification process. The standardized warning data stream includes sending a warning notification, informing the boundary of the control area, the permitted flight range, and the reporting channel. Simultaneously, a differentiated non-uniform sending strategy is adopted based on the boundary intrusion probability: if the boundary intrusion probability is below 0.6, the warning notification sending interval is gradually shortened from the initial 5 seconds to give the operator sufficient reaction time; if the boundary intrusion probability is above 0.6, the warning notification sending interval is fixed at 3 seconds. If no feedback is received after reaching the process time limit, a formal warning template is sent. If there is still no response after a second sending, the countermeasure pre-warning preparation stage begins. This non-uniform interval sending strategy can effectively warn intruders, guide unintentional intruders, and quickly deal with malicious intruders, thus balancing the prevention of accidental injury with the efficiency of handling the situation.

[0033] 3. Level 2 Unauthorized Entry (Controlled Area Identification Zone): A two-way confirmation process is used, with a total interaction time limit of 15 seconds and a fixed 3-second sending interval. A formal warning template is matched based on the one-way notification process. The standardized warning data flow includes sending a formal warning notification, inquiring about the drone's identity and flight purpose, clarifying the basis for the violation judgment, and quantifying the evacuation route and time limit requirements. The process rules require the drone to provide confirmation of receipt, respond to inquiries, and provide operational responses, supporting three feedback types: compliance reporting, immediate evacuation, and objection appeal. In subsequent steps, if a compliance report is received and verified, the entire process terminates directly; if an immediate evacuation is received, the flight trajectory is continuously monitored, and a subjective malice assessment model is used to verify the authenticity of the evacuation behavior; if an objection appeal is received, manual review is immediately initiated; if no feedback is received within the time limit and the malice judgment exceeds the set threshold, countermeasures are formally initiated.

[0034] 4. Level 1 Emergency Intrusion (Core Restricted Area or High-Speed ​​Approach Disposal Area): Matches the emergency procedure, with a total interaction time limit of 3 seconds and a fixed sending interval of 1 second. The emergency procedure prioritizes core airspace security. Subjective malice assessment is not a prerequisite for initiating countermeasures, avoiding time-consuming assessments that could delay response. Malice score calculation is performed in parallel with the disposal process in the background, without blocking the main process. The results are used for dynamic grading of countermeasure intensity and post-event tracing and archiving. A countermeasure pre-warning template is matched according to the one-way notification process. The standardized warning data stream includes simultaneous sending of countermeasure pre-warning notifications and initiation of countermeasure preparation, clearly defining the countermeasure type, initiation time, and safety prompts. There is no need to wait for the complete feedback process of S3; the synchronous linkage tracking mechanism locks the drone's takeoff location and the operator's area. After countermeasures are initiated, status prompts are continuously sent through the warning channel to ensure uninterrupted communication. The malice score results generated in parallel in the background are synchronously pushed to the S4 verification and archiving stage and included in the full-link disposal log.

[0035] It should be noted that during the interaction, the priority of the communication channel follows the order of dedicated protocol > radio frequency > Bluetooth. When taking countermeasures, two sets of asymmetric orthogonal subcarriers are divided within the same operating frequency band based on the orthogonal frequency division multiplexing architecture. One set is a low-power warning subcarrier group, using QPSK (Quadrature Phase Shift Keying) modulation, carrying warning text and voice data streams; the other set is a dynamically activated interference subcarrier group, using pseudo-random Gaussian white noise modulation, reserved as a counter-interference carrier. The two sets of subcarriers are orthogonal in the frequency domain and configured with independent power control links. When the counter-interference subcarrier is activated, the transmit power of the warning subcarrier is automatically increased by 2dB and an in-band interference cancellation algorithm is enabled to offset the in-band leakage of the countermeasure signal, ensuring that the warning data transmission is not affected when the countermeasure is running at full power, and ensuring that the warning link remains online throughout the entire countermeasure process.

[0036] In the aforementioned interactive process, the delay from the completion of the violation level assessment to the initial warning transmission is no greater than 1 second. For fixed-location drones, omnidirectional transmission is used, with a coverage area of ​​no less than 1.5 times the countermeasure radius. For mobile countermeasure devices, directional and precise transmission is used. The warning employs a dual-channel synchronous system of "text message + voice broadcast." The text warning pushes a standardized warning message to the drone control terminal pop-up display via a dedicated data transmission channel, fully carrying traceable and precise information such as quantified evacuation coordinates, time limits, and feedback channels. The voice warning is simultaneously repeated twice to ensure the operator receives the warning information immediately. If transmission fails, the channel is automatically switched and retransmitted up to three times. Multiple failures automatically trigger manual review. A transmission status data stream is generated, recording the channel, delay, transmission interval, and subcarrier allocation ratio.

[0037] The interaction process, violation level, sending parameters, and grid identifiers are integrated to generate a process scheduling data stream and sent to the central processing unit. The central processing unit receives the process scheduling data stream, reads the built-in four types of standard warning template library (pre-warning, formal warning, countermeasure announcement, and error correction instruction), and matches the corresponding standard interaction process template according to the violation level. It automatically fills in the standard mandatory core fields, including graded risk level, quantified evacuation coordinates, precise time limit, and three types of feedback channels (compliance reporting / immediate evacuation / objection appeal). It sets a unified verification that the number of characters in a single warning message does not exceed a set number (e.g., 120 characters), automatically matches the Chinese and English languages ​​of the drone registration, quantifies the location and time information, and embeds the instruction. After the content is filled in, an SHA-256 digital signature is attached, generating a standardized warning data stream that fully conforms to the local standard format.

[0038] Warning signals are sent to the drone and control terminal, and process scheduling data streams, sending status data streams and standardized warning data streams are written to the memory buffer.

[0039] S3. Collect feedback data streams from drones, calculate reaction delay based on the sending time of standardized warning data streams and the receiving time of feedback data streams, calculate malice score by combining reaction delay and motion characteristics, match differentiated handling strategies based on malice score, and output handling strategy data stream.

[0040] Specifically, the system collects feedback data streams transmitted from the drone and the control terminal. These data streams include feedback type, drone identity hash, warning number, and reception timestamp. Simultaneously, it collects the drone's latest flight data in real time, including speed, acceleration, heading, and position changes.

[0041] Upon receiving the feedback data stream, a triple verification process is first performed: CRC data integrity check, identity hash and alert number binding verification, and zero-knowledge digital credential verification. No complete identity plaintext is transmitted throughout the process. Only after successful verification is the response latency calculated. Response latency is the difference between the reception time of the feedback data stream and the transmission time of the standardized alert data stream, measured in seconds.

[0042] Meanwhile, three consecutive frames of flight status data of the drone after the warning was sent were extracted, and four kinematic features were extracted: rate of change of speed, heading deviation, rate of change of distance from the disposal area, and normal acceleration. Together with the reaction delay, these features constitute five dimensions of malice assessment characteristics.

[0043] After normalizing and mapping each maliciousness assessment feature, the scores are weighted and summed according to preset maliciousness weights to obtain the maliciousness score. The calculation formula is as follows:

[0044] In the formula, to This is a normalized mapping function for each maliciousity assessment feature. The output range is [0, 1]. The more a feature matches the malicious intrusion behavior pattern, the higher the output value. to The corresponding feature weights are summed to equal 1. Weight allocation is based on convergent behavior and response speed as the core criteria. As an example, to The values ​​were 0.4, 0.1, 0.1, 0.3, and 0.1 in sequence.

[0045] The normalization rules for each malice assessment feature are as follows: : Reaction delay normalization, When the interval is linearly mapped to [0, 1], Or, if there is no feedback at all, take the maximum value of 1; : The speed change rate is normalized, and the output approaches 1 when the drone is continuously accelerating towards the disposal area, and approaches 0 when the speed is stable or decelerating to withdraw. : Normalization of heading offset: When the heading is ≥90° from the standard evacuation direction, the output approaches 1, and when the heading changes to the evacuation direction, the output approaches 0. : The rate of change of distance from the treatment area is normalized. When the distance from the treatment area continues to decrease (i.e., it is approaching in a positive direction), the output approaches 1. When the distance continues to increase (i.e., it is evacuating), the output approaches 0. Normal acceleration is normalized; the more intense the flight maneuver, the higher the output value; during stable straight flight, the output value approaches 0.

[0046] The malice score was calculated. Then, based on the interval, it is divided into three levels: unintentional intrusion, general violation, and malicious intrusion. Among them: like If the intrusion is determined to be unintentional, the drone will exhibit a short reaction time, a rapid change of course to evacuate, and a stable speed. At this point, an observation and guidance strategy will be implemented, with the observation period appropriately extended to 15 seconds and the warning broadcast frequency reduced to once every 5 seconds. The drone will be guided to evacuate along a designated route without initiating countermeasures.

[0047] like If the violation is deemed a minor breach, and the drone exhibits a reaction delay of 2 to 5 seconds, failing to immediately withdraw but also not venturing further, a continuous warning and tracking strategy will be implemented. A secondary warning will be issued according to standard procedures, and if the drone fails to withdraw within the specified time, a countermeasure warning will be activated.

[0048] like If the intrusion is deemed malicious, and the drone exhibits no feedback or a reaction delay greater than 5 seconds, continuously accelerates towards the disposal area, and shows no tendency to withdraw, the multiple warning procedures will be discontinued, and a countermeasures warning will be triggered directly, with countermeasures quickly initiated.

[0049] This malice level classification and handling mechanism can effectively distinguish between unintentional and malicious intruders while ensuring the accuracy of the judgment, and also take into account data security protection.

[0050] If the identification is determined to be a misidentification, the error correction process will be directly initiated, and a standard error correction instruction will be issued within 5 minutes. A handling scheduling data stream containing the malice level and handling strategy will be generated and sent to the countermeasure execution, trajectory monitoring and log archiving modules.

[0051] S4. Monitor the drone withdrawal based on the data stream of the disposal strategy, encrypt and archive the interaction data and calculation results of the entire link, and output a secure archived data stream.

[0052] The full-link interaction record refers to all data streams from S1 to S4. These data streams, along with the calculation results of boundary intrusion probability and malice score, are stored using AES-256 encryption. A retrieval index is established by drone ID, warning number, and timestamp. Logs are retained for no less than 180 days, and the overall false alarm rate is controlled within 0.5%.

[0053] Securely archived data streams are pushed to the command and management terminal, completing a closed loop of information flow for warnings, feedback, malicious assessment, and handling.

[0054] In a preferred embodiment, the calculation of the boundary intrusion probability and the malice score can adopt an edge-end collaborative architecture, without uploading the original data to the cloud, thus balancing the accuracy of the judgment and the data security.

[0055] Compared with the prior art, the present invention has the following technical advantages: 1. Achieved differentiated hierarchical control based on behavioral response. This invention introduces a two-tiered evaluation mechanism of "boundary intrusion probability" and "subjective malice" into the warning interaction process. The former dynamically adjusts the sending interval and initial process type based on the drone's spatial position and movement trend before the warning is sent, while the latter quantitatively determines the intrusion intent based on the operator's feedback behavior and the drone's response characteristics after the warning is sent. The two work together to enable the system to distinguish between two fundamentally different scenarios: unintentional intrusion and malicious intrusion. Differentiated handling paths are then matched for guided evacuation and rapid countermeasures, overcoming the static control defects of fixed intervals and fixed durations in traditional solutions. This achieves the hierarchical control objective of "standardized guidance for unintentional intruders and rapid handling for malicious intruders."

[0056] 2. A standardized, compliant, and traceable warning content packaging system has been established. This invention incorporates four types of standard templates: pre-warnings, formal warnings, countermeasure announcements, and error correction instructions. Based on process scheduling results, all mandatory traceability fields are automatically filled in, and a digital signature is added after unified verification. Automatic switching between Chinese and English is supported. This mechanism ensures that every warning message issued to the operator complies with local standard format requirements, eliminating common problems such as missing fields and vague descriptions. It provides operators with clear, quantifiable, and actionable handling guidelines, effectively reducing the risk of mishandling due to unclear information.

[0057] 3. Ensure the continuous online status of the warning link throughout the entire countermeasure process. This invention divides the same operating frequency band into orthogonal warning subcarrier groups and dynamically activating / disabling interference subcarrier groups, with each group configured with an independent power control link. When the countermeasure interference subcarrier is activated, the system automatically increases the transmission power of the warning subcarrier and enables an in-band interference cancellation algorithm to offset in-band leakage of the countermeasure signal. This ensures that warning text and voice data can still be stably transmitted through an independent channel during full-power operation of the countermeasure equipment, enabling parallel execution of countermeasure actions and warning guidance, and avoiding communication link interruptions caused by countermeasure activation.

[0058] 4. Forming a closed-loop system encompassing alerts, feedback, assessment, and response. This invention starts with standardized raw situational data streams, sequentially completing grid binding, intrusion probability assessment, initial process scheduling, alert content encapsulation and transmission, feedback reception and verification, subjective malicious intent quantification, and tiered response decisions. Finally, the entire interaction record and assessment results are encrypted and archived. Each step is interconnected and progressively builds upon the previous one, forming a closed-loop information flow processing chain. The execution status of any step is traceable, and the response results are reproducible, providing complete data support for post-event auditing and strategy optimization.

[0059] The preferred embodiments of the present invention have been described in detail above. It should be understood that those skilled in the art can make numerous modifications and variations based on the concept of the present invention without creative effort. Therefore, all technical solutions that can be obtained by those skilled in the art based on the concept of the present invention through logical analysis, reasoning, or limited experimentation on the basis of existing technology should be within the scope of protection defined by the claims.

Claims

1. An information flow processing method based on the UAV countermeasure warning dialogue specification, characterized in that, include: S1. Collect raw flight data streams of UAVs, match them to the airspace hierarchical grid according to the real-time coordinates of the UAVs, bind the corresponding risk level and the corresponding interactive process template, and generate standardized raw situation data streams. S2. Extract flight features from the standardized raw situation data stream, calculate the boundary intrusion probability based on the flight features, determine the violation level by combining the grid position and the boundary intrusion probability, match the corresponding interaction process from the interaction process template according to the violation level, and output a standardized warning data stream. S3. Collect feedback data streams from the drone, calculate reaction delay based on the sending time of the standardized warning data stream and the receiving time of the feedback data stream, calculate the malice score by combining the reaction delay and motion characteristics, match differentiated handling strategies based on the malice score, and output the handling strategy data stream. S4. Based on the aforementioned disposal strategy, monitor the drone's withdrawal via data stream, encrypt and archive the entire link interaction data and calculation results, and output a secure archived data stream.

2. The information flow processing method based on the UAV counter-warning dialogue specification according to claim 1, characterized in that, The flight characteristics include normalized position offset rate, heading variability, altitude change rate, normal acceleration variability, radial approach rate, and the angle between the heading and the boundary normal.

3. The information flow processing method based on the UAV counter-warning dialogue specification according to claim 2, characterized in that, The calculation of the boundary intrusion probability includes: linearly weighting the flight features by preset fixed weights to obtain an initial boundary intrusion probability, and then correcting the initial boundary intrusion probability by combining the risk coefficient corresponding to the current grid position of the UAV to obtain the final boundary intrusion probability.

4. The information flow processing method based on the UAV counter-warning dialogue specification according to claim 3, characterized in that, The airspace hierarchical grid is divided into a disposal zone, an identification zone, an early warning zone, and a monitoring zone from the inside out according to the boundary of the identification zone; the correspondence between the risk level and the airspace hierarchical grid is as follows: the disposal zone corresponds to level 1 risk, the identification zone corresponds to level 2 risk, the early warning zone corresponds to level 3 risk, and the monitoring zone corresponds to level 4 risk. The method for determining the violation level is as follows: the risk level of the current airspace classification grid bound to the UAV is used as the baseline level. When the boundary intrusion probability is not lower than the first threshold, the baseline level is increased by one level; when the boundary intrusion probability is not higher than the second threshold, the baseline level is decreased by one level; otherwise, the baseline level remains unchanged.

5. The information flow processing method based on the UAV countermeasure warning dialogue specification according to claim 4, characterized in that, If the violation level is determined to be a level four risk, a monitoring process will be adopted, which will only monitor the drone's flight status. If the violation level is determined to be a level three risk, a one-way notification process is adopted, with a total interaction time limit of t1 seconds. The standardized warning data stream includes sending a warning notification, informing the boundary of the prevention and control area, the permitted flight range and the reporting channel. The sending interval adopts a differentiated non-uniform sending strategy based on the boundary intrusion probability. If the violation level is determined to be Level 2 risk, a two-way confirmation process is adopted, with a total interaction time limit of t2 seconds and a fixed sending interval of t3 seconds. The standardized warning data stream includes sending a formal warning notification, inquiring about the drone's identity and flight purpose, clarifying the basis for the violation judgment, and quantifying the evacuation route and time limit requirements. The drone is required to provide confirmation, answer the questions, and respond to the operation. If the risk is determined to be Level 1, an emergency procedure is adopted, and the standardized warning data flow includes the simultaneous sending of a countermeasure advance notice and the initiation of countermeasure preparation, clearly defining the countermeasure type, initiation time, and security prompts. There is no need to wait for a complete feedback process; the synchronous linkage tracking mechanism locks the drone's takeoff location and the operator's location.

6. The information flow processing method based on the UAV countermeasure warning dialogue specification according to claim 4, characterized in that, The calculation of the malice score includes: Acquire continuous multi-frame flight status data of the UAV after the standardized warning data stream is sent; The reaction time delay, velocity change rate, heading deviation, distance change rate from the boundary of the disposal area, and normal acceleration were extracted as features for malice assessment. After normalizing and mapping each maliciousness assessment feature, the maliciousness score is obtained by weighting and summing the features according to the preset maliciousness weights.

7. The information flow processing method based on the UAV countermeasure warning dialogue specification according to claim 6, characterized in that, The differentiated handling strategy is divided according to the malice score range: when the malice score is not lower than 0.7, countermeasures are initiated; when the malice score is between 0.3 and 0.7, a continuous warning and tracking strategy is implemented; and when the malice score is lower than 0.3, an observation and guidance strategy is implemented.

8. The information flow processing method based on the UAV counter-warning dialogue specification according to claim 1, characterized in that, The S2 also includes a whitelist verification step: matching the drone's identity information with a preset whitelist. If the match is successful, the violation level determination and subsequent handling process are skipped, and a compliant release data stream is output.

9. The information flow processing method based on the UAV countermeasure warning dialogue specification according to claim 1, characterized in that, The raw flight data stream of the UAV is the publicly available flight status data and broadcast communication message data of the target UAV collected within the legally controlled airspace, excluding the operator's personal privacy and non-public identity information.

Images