Intelligent cooperative protection method, system and equipment for low-altitude falling and medium

Abstract

The invention relates to an intelligent cooperative protection method, system and device for low-altitude falling and a medium. The method comprises the following steps: establishing a reliable link, and obtaining a flight state data packet; based on the system health state code in the flight state data packet, adjusting a warning level of a falling judgment algorithm of the wearable safety equipment and a judgment sensitivity threshold value of the falling judgment algorithm; based on the local sensor data and the real-time flight state data packet of the wearable safety equipment, collaborative falling judgment is carried out, and a trigger decision is generated; and when the trigger decision is high confidence coefficient to confirm falling, a protection instruction is generated to instruct execution of multiple protection and coordination actions such as airbag trigger, sound-light alarm and information return. According to the method, the local motion information of the wearable end and the global state information of the aircraft end are fused, so that the original information island state of the wearable equipment is thoroughly broken, the accuracy and timeliness of recognizing the complex low-altitude falling scene are remarkably improved, and the misinformation and missing report risks are greatly reduced.

Description

Technical Field

[0001] This invention belongs to the field of intelligent safety protection and collaborative control technology, and in particular relates to an intelligent collaborative protection method, system, device and medium for low-altitude fall. Background Technology

[0002] With the booming development of the low-altitude economy and the gradual application of manned aircraft such as electric vertical takeoff and landing (EVTOL) aircraft, the safety of passengers during flight has received increasing attention. To address this, wearable safety devices integrating airbags have emerged. These devices typically incorporate inertial sensors that detect human motion to determine whether to trigger airbag deployment, aiming to provide close protection for passengers in the event of a sudden fall.

[0003] In traditional technologies, these wearable safety devices operate as independent personal protective equipment. Their operation primarily relies on local motion data collected by sensors such as accelerometers and gyroscopes mounted on the device. The device's control algorithm analyzes this local data based on preset thresholds and patterns to determine whether a dangerous situation, such as a fall or drop, has occurred, and then decides whether to activate the airbag.

[0004] However, traditional technologies have significant limitations in the specific and complex application scenario of low-altitude flight. The lack of effective information exchange between the equipment and the aircraft itself creates "information silos," preventing the use of more precise and global status information from the aircraft (such as accurate altitude, speed, and system fault warnings) for decision-making support. This results in three prominent problems: First, relying solely on local motion information makes it difficult to accurately distinguish between normal flight maneuvers and a genuine loss of control during a crash, leading to limited accuracy and a risk of both false alarms and missed alarms. Second, protective actions only begin after the aircraft's own sensors detect the impact, failing to provide early warnings and pre-emptive preparations based on signs of aircraft malfunction, resulting in a reactive response. Third, the entire triggering chain relies on a single electronic sensing and control system, lacking effective redundancy in the event of electronic system failure. Summary of the Invention

[0005] Therefore, it is necessary to provide a low-altitude fall protection intelligent collaborative method that can achieve information collaboration between devices, improve the accuracy and timeliness of judgment, and enhance the overall reliability of the system, in order to address the above-mentioned technical problems.

[0006] Firstly, this application provides a smart collaborative protection method for low-altitude fall, comprising:

[0007] Establish a real-time data communication link between the wearable safety device and the low-altitude aircraft, and periodically receive flight status data packets from the low-altitude aircraft through the real-time data communication link; wherein, the flight status data packets include the aircraft's altitude, speed, attitude and system health status code;

[0008] Based on the system health status code in the flight status data packet, the alert level and the judgment sensitivity threshold of the fall detection algorithm of the wearable safety device are adjusted to obtain the adjusted alert level and the adjusted judgment sensitivity threshold.

[0009] Based on local sensor data and real-time flight status data packets from wearable safety devices, collaborative fall determination is performed to generate trigger decisions;

[0010] When the trigger decision is a high-confidence confirmation of the fall, a protection command is generated. The protection command is used to instruct the execution of relevant protection measures. The protection measures include controlling the gas generator of the wearable safety device to deploy the airbag, activating the local audible and visual alarm, and sending the trigger confirmation information to the low-altitude aircraft via a real-time data communication link.

[0011] Furthermore, based on the system health status code in the flight status data packet, the alert level and the judgment sensitivity threshold of the fall detection algorithm of the wearable safety device are adjusted to obtain the adjusted alert level and the adjusted judgment sensitivity threshold, including:

[0012] Based on the system health status code in the flight status data packet and the list of preset serious fault alarm codes in the wearable safety device, it is determined whether the low-altitude aircraft is in a serious fault state, and the judgment result is obtained.

[0013] If the judgment result is that the current low-altitude aircraft is in a serious malfunction state, the internal collaborative alert status variable of the wearable safety device is set from the normal alert state to the first-level alert state to obtain the adjusted alert level.

[0014] In response to the adjusted alert level, the key detection threshold parameters in the fall determination algorithm are adjusted to obtain the adjusted determination sensitivity thresholds; the adjustment includes reducing the synthetic acceleration modulus threshold from a first conventional value to a second higher sensitivity value and reducing the angular velocity modulus threshold from a third conventional value to a fourth higher sensitivity value.

[0015] Furthermore, based on local sensor data and real-time flight status data packets from wearable safety devices, a collaborative fall determination is performed to generate triggering decisions, including:

[0016] Raw data from the local sensors of the wearable safety device is collected, and the raw data is preprocessed by filtering and compensation to obtain preprocessed data; the local sensors include an inertial measurement unit and a barometer.

[0017] Based on the preprocessed data, a local feature vector sequence is calculated; the local feature vector sequence includes the synthetic acceleration modulus, weightlessness index, and angular velocity modulus.

[0018] Parse real-time flight status data packets to obtain aircraft status evidence, which includes the aircraft's real-time altitude, rate of altitude descent, and fault alarm flags.

[0019] Based on the local feature vector sequence, a local fall suspicion detection result is generated by detecting the composite pattern of continuous weightlessness followed by abnormal angular velocity growth.

[0020] Based on the aircraft status evidence, it is determined whether there is strong anomaly evidence for the low-altitude aircraft, and the result of the determination of the existence of strong anomaly evidence is obtained; among them, strong anomaly evidence is a cliff-like drop in altitude, an altitude drop rate exceeding the normal landing threshold, or an active collision warning signal of the aircraft.

[0021] Based on the local crash suspicion detection results and the judgment results of the existence of strong anomaly evidence of the aircraft, combined with the preset fusion decision rules, a trigger decision is generated.

[0022] Furthermore, based on the local crash suspicion detection results and the determination of the existence of strong anomaly evidence of the aircraft, combined with the preset fusion decision rules, a trigger decision is generated, including:

[0023] If the local crash suspicion assessment result is a suspected crash or a confirmed crash, and the aircraft strong anomaly evidence existence assessment result is strong anomaly evidence existence, a trigger decision with high confidence is generated.

[0024] If the local crash suspicion assessment result is suspected crash or confirmed crash, but the strong anomaly evidence of the aircraft assessment result is no strong anomaly evidence, obtain the current collaborative alert status variable;

[0025] If the collaborative alert status variable is at level one alert, generate a trigger decision with high confidence.

[0026] If the collaborative alert status variable is the normal alert status, the impact signal and the aircraft confirmation anomaly information are continuously monitored within the preset confirmation waiting window to obtain the monitoring results.

[0027] If the monitoring results indicate that a local impact signal or aircraft confirmation anomaly information is detected within the preset confirmation waiting window period, a trigger decision with high confidence is generated.

[0028] If the monitoring result is that no local impact signal or aircraft confirmation anomaly information is detected within the preset confirmation waiting window period, a trigger decision with a confidence level of zero is generated.

[0029] Furthermore, based on the local feature vector sequence, by detecting the composite pattern of continuous weightlessness followed by abnormal angular velocity increase, a local fall suspicion detection result is generated, including:

[0030] Determine whether the weightlessness index in the local feature vector sequence meets the condition of continuous weightlessness to obtain the first judgment result; wherein, the condition of continuous weightlessness is that the weightlessness index is continuously lower than the preset dynamic threshold for a preset duration;

[0031] If the first judgment result is that the weightlessness index meets the condition of continuous weightlessness, then it is judged whether the angular velocity magnitude in the local feature vector sequence meets the condition of abnormal angular velocity growth, and the second judgment result is obtained; wherein, the condition of abnormal angular velocity growth is that it exceeds the preset dynamic angular velocity threshold within the time window after the weightlessness stage.

[0032] If the second judgment result is that the angular velocity modulus meets the abnormal angular velocity growth condition, a local fall suspicion judgment result is generated;

[0033] Based on the local fall suspicion degree judgment result, it is determined whether there is a peak value in the synthetic acceleration magnitude of the local feature vector sequence that exceeds the preset extremely high impact threshold, and a third judgment result is obtained;

[0034] If the third judgment result is that the synthetic acceleration modulus has a peak value exceeding the preset extremely high impact threshold, a local fall suspicion judgment result is generated to confirm the fall.

[0035] Secondly, this application also provides a low-altitude fall intelligent collaborative protection system, comprising:

[0036] The data acquisition module is used to establish a real-time data communication link between the wearable safety device and the low-altitude aircraft, and periodically receive flight status data packets from the low-altitude aircraft through the real-time data communication link; wherein, the flight status data packets include the aircraft's altitude, speed, attitude and system health status code;

[0037] The algorithm adjustment module is used to adjust the alert level and the judgment sensitivity threshold of the fall detection algorithm of the wearable safety device based on the system health status code in the flight status data packet, so as to obtain the adjusted alert level and the adjusted judgment sensitivity threshold.

[0038] The decision generation module is used to perform collaborative crash determination and generate trigger decisions based on local sensor data and real-time flight status data packets from wearable safety devices;

[0039] The protection trigger module is used to generate protection commands when the trigger decision is a high-confidence confirmation of a fall. The protection commands are used to instruct the execution of relevant protection actions. The protection actions include controlling the gas generator of the wearable safety device to deploy the airbag, activating the local audible and visual alarm, and sending trigger confirmation information to the low-altitude aircraft through a real-time data communication link.

[0040] Thirdly, this application also provides a computer device, including a memory and a processor, wherein the memory stores at least one instruction, at least one program, code set or instruction set, and the at least one instruction, the at least one program, the code set or instruction set is loaded and executed by the processor to implement any of the low-altitude fall intelligent collaborative protection methods described in the embodiments of this application.

[0041] Fourthly, this application also provides a computer-readable storage medium storing at least one piece of program code, which is loaded and executed by a processor to implement any of the low-altitude fall intelligent collaborative protection methods described in the embodiments of this application.

[0042] The aforementioned intelligent collaborative protection method, system, device, and medium for low-altitude fall detection acquires flight status data packets via a reliable link. Based on the system health status code in the flight status data packets, it adjusts the alert level and sensitivity threshold of the fall detection algorithm of the wearable safety device. Based on the local sensor data and real-time flight status data packets of the wearable safety device, it performs collaborative fall detection and generates trigger decisions. When the trigger decision confirms a fall with high confidence, it generates protection commands to instruct the execution of multiple protective and collaborative actions, including airbag triggering, audible and visual alarms, and information feedback. By integrating local motion information from the wearable device with global status information from the aircraft, it completely breaks down the original information silos of wearable devices, significantly improving the accuracy and timeliness of identifying complex low-altitude fall scenarios, greatly reducing the risk of false alarms and missed alarms, and enhancing the reliability, adaptability, and overall safety level of the occupant safety protection system in low-altitude manned scenarios. Attached Figure Description

[0043] To more clearly illustrate the technical solutions in the embodiments or related technologies of this application, the accompanying drawings used in the description of the embodiments or related technologies will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0044] Figure 1 This is a flowchart illustrating a low-altitude fall intelligent collaborative protection method in one embodiment;

[0045] Figure 2 This is a flowchart illustrating the steps of xxx in one embodiment;

[0046] Figure 3 This is a schematic diagram of a low-altitude fall protection system in one embodiment. Detailed Implementation

[0047] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this application.

[0048] In one embodiment, a smart collaborative protection method for low-altitude fall is provided. This embodiment illustrates the application of this method to a terminal. It is understood that this method can also be applied to a server, and further to a system including both a terminal and a server, and is implemented through the interaction between the terminal and the server. Figure 1 As shown, in this embodiment, the method includes the following steps:

[0049] Step S101: Establish a real-time data communication link between the wearable safety device and the low-altitude aircraft, and periodically receive flight status data packets from the low-altitude aircraft through the real-time data communication link; wherein, the flight status data packets include the aircraft's altitude, speed, attitude and system health status code.

[0050] Wearable safety devices are portable electronic devices that integrate safety protection functions and can be worn on the body. They can reduce user safety risks and improve emergency response efficiency through technical means. Low-altitude aircraft refer to aircraft that mainly fly in low-altitude airspace and complete specific tasks. Their core characteristics are low flight altitude, relatively slow flight speed, mission scenarios close to the ground or low-altitude environment, and they often have flexible and maneuverable characteristics. Low-altitude airspace usually refers to the area below 1,000 meters above the ground, and in some scenarios it extends to 3,000 meters. Periodicity describes the attribute of things repeating a specific state or behavior according to a certain pattern in the process of development and change. It is a regular cycle.

[0051] For example, a low-latency, highly reliable two-way real-time data communication link is established through a handshake negotiation between a first communication module integrated into a wearable safety device and a second communication module configured on the low-altitude aircraft. This link can be implemented using Bluetooth Low Energy to ensure stable connection in complex electromagnetic environments. The device actively sends data request commands to the aircraft at fixed intervals (e.g., every 100 milliseconds) or passively receives flight status data packets broadcast by the aircraft. These flight status data packets are structured digital information frames, containing at least: high-precision barometric altitude and geometric altitude obtained through the aircraft's multi-source fusion navigation system; components of the three-dimensional velocity vector in the ground coordinate system; pitch, roll, and yaw angles measured by the aircraft's attitude reference system; and a system health status code generated by the aircraft's fault diagnosis unit. This system health status code can use multi-bit binary encoding, with each bit or segment corresponding to a specific subsystem state, such as the power system. Pre-defined severe fault code combinations are included; that is, a specific 0 / 1 arrangement is predefined among all binary combinations, and once this combination is detected, it is directly identified as a severe fault. After receiving the data packet, a Cyclic Redundancy Check (CRC) is performed to ensure data integrity. Then, the data payload is parsed, and physical quantities such as altitude, velocity, and attitude are converted into engineering unit values ​​that can be processed by the internal algorithm.

[0052] The first communication module is the core hardware for wearable safety devices to achieve core secure communication functions. It is also the primary carrier for establishing connections between the device and the outside world and the backend, adapting to the core requirements of wearable devices for miniaturization, low power consumption, and high reliability. The second communication module is the core hardware for information interaction in low-altitude aircraft, adapting to the electromagnetic environment, flight scenarios, and regulatory requirements of low-altitude airspace. Its core requirements are high reliability, low latency, miniaturization, and anti-interference, while also considering lightweight design and low power consumption. Handshake negotiation refers to the process by which the two communicating parties exchange key information according to pre-agreed rules before formally transmitting data, confirming each other's identities, capabilities, and communication parameters, and ultimately reaching an agreement to establish a secure and stable connection. Bluetooth Low Energy (Bluetooth Low Energy)... Bluetooth Energy (BLE) is a lightweight Bluetooth branch launched by the Bluetooth Special Interest Group (SIG). Its core focus is on low power consumption, low complexity, and short range, designed specifically for scenarios requiring long standby times and low data transmission volumes. Structured programming essentially establishes clear and followable rules and frameworks for disordered and complex systems. By defining elements, relationships, and logic, it guides things from chaos to order, reducing the cost of understanding, operation, or management. A multi-source fusion navigation system for aircraft refers to integrating multiple navigation data sources with different principles and characteristics to compensate for the shortcomings of single navigation methods. Ultimately, it provides aircraft with more accurate, reliable, and continuous key navigation information such as position, velocity, and attitude, and is one of the core technologies ensuring safe flight in complex environments. A ground coordinate system is a three-dimensional coordinate system used to describe the position of an object on the Earth's surface or near-Earth space. It is used to establish a fixed ground reference frame, transforming abstract spatial positions into measurable and calculable specific coordinate values. An aircraft attitude reference system (AttitudeReference)... A system (ARS) is a core component of an aircraft's navigation and control system. It measures and outputs the aircraft's attitude information in real time, providing crucial references for flight control, navigation calculations, and pilot guidance, ensuring stable flight at predetermined attitudes such as level, climb, and turn. The aircraft navigation and control system is the core system ensuring the aircraft flies along a predetermined trajectory and completes its flight mission. It consists of two core modules: navigation and control, which work together. The aircraft fault diagnosis unit is a core subsystem ensuring the safe operation of the aircraft. Essentially, it monitors the health status of key systems such as power, avionics, and control in real time through a closed-loop process of "data acquisition-analysis and judgment-early warning / decision-making," promptly identifying faults, locating fault sources, and assisting in subsequent handling. Multi-bit binary encoding is an encoding method that uses sequences of 0s and 1s to represent information such as numbers, characters, and instructions. By expanding the number of bits, it achieves more accurate mapping of richer information. A subsystem is a core concept in systems theory, referring to a local unit within a larger parent system or overall system that has relatively independent functions and exists to achieve the overall goals of the parent system.A power system is the core system that provides the energy and power transmission required for the movement of equipment, machinery, or vehicles. It converts some form of energy into mechanical kinetic energy to drive a target object to achieve movement, rotation, or other forms of motion. Cyclic redundancy check (CRC) is an error detection technique based on polynomial operations, used to verify whether data has been tampered with, lost, or erroneous during transmission or storage.

[0053] Step S102: Based on the system health status code in the flight status data packet, adjust the alert level and the judgment sensitivity threshold of the fall detection algorithm of the wearable safety device to obtain the adjusted alert level and the adjusted judgment sensitivity threshold.

[0054] The fall detection algorithm is an algorithm that identifies whether a target has fallen, a specific motion state, through sensor data or environmental information. The algorithm manages a set of key detection threshold parameters, which are the quantitative limits for the algorithm to judge whether the motion is abnormal. The most critical parameters are the synthetic acceleration modulus threshold for detecting weightlessness and the angular velocity modulus threshold for identifying tumbling.

[0055] For example, based on the system health status code in the flight status data packet, the alert level and judgment sensitivity threshold of the fall detection algorithm of the wearable safety device are adjusted to obtain the adjusted alert level and judgment sensitivity threshold.

[0056] Step S103: Based on the local sensor data and real-time flight status data packets of the wearable safety device, a collaborative fall determination is performed to generate a trigger decision.

[0057] Among them, collaborative fall detection is designed for fall-related safety incidents, using multi-source data collaborative analysis to achieve accurate fall risk assessment and decision triggering.

[0058] For example, local high-frequency sensor data streams and aircraft status data streams are processed in parallel. After time-stamp-based alignment of these two types of data, a collaborative crash determination is performed to generate a trigger decision. Time alignment essentially synchronizes or matches multiple time-related objects along the time dimension to eliminate time discrepancies between different objects and ensure that their information corresponds consistently at the same point in time.

[0059] Step S104: When the trigger decision is a high-confidence confirmation of the fall, a protection command is generated; wherein, the protection command is used to instruct the execution of relevant protection work; the protection work includes controlling the gas generator of the wearable safety device to deploy the airbag, activating the local audible and visual alarm, and sending trigger confirmation information to the low-altitude aircraft through a real-time data communication link.

[0060] Among them, the gas generator is the core functional component of wearable safety devices. It is used to quickly generate gas in emergency scenarios to drive the device to complete the inflation action and achieve the protection / lifesaving effect. Its design must meet the core requirements of rapid response, small size, and safety and reliability. An airbag is a safety device that inflates and deploys rapidly when a car, aircraft, or other vehicle collides. It is used to cushion the impact between the occupants and hard objects inside the vehicle or aircraft, reducing the damage caused by the collision.

[0061] For example, after generating a high-confidence confirmation of a fall trigger decision, a series of orderly and rapid protective actions are executed. A high-level, specific-pulse-width digital pulse signal is sent to the drive circuit as the primary protective command. This pulse signal turns on the drive circuit, and a powerful current flows through the excitation unit of the gas generating device (such as an electric detonator). In the case of the electric detonator, the current rapidly heats the internal bridge wire and ignites the igniter, which in turn triggers the violent combustion of the gas-generating agent, producing a large amount of harmless gas within milliseconds. The released high-pressure gas is rapidly injected through a pipe into the airbag pre-folded inside the clothing, causing it to expand rapidly and form a buffer protective layer between the occupant's body and external obstacles. Almost simultaneously with the inflation command, an activation command is sent to the audible and visual distress call unit, causing the high-brightness LED lights to operate in an international rescue standard flashing mode and the piezoelectric buzzer to emit a high-decibel intermittent alarm sound for personnel location and distress calls after an accident. In addition, a trigger confirmation message, including the event timestamp, the device's unique identifier, the trigger decision confidence level, and the possible latest approximate location, is immediately encapsulated into a data frame and sent back to the low-altitude aircraft via the communication link. The aircraft can record this information in its flight data recorder and, while its communication system remains operational, send this information as a critical event status to the ground station to provide more accurate joint accident scene information for subsequent rescue and analysis.

[0062] In an electronic system, the drive circuit acts as a bridge connecting the control signal source and the actuator. It converts the weak signal output from the control terminal into a strong electrical signal capable of driving the actuator normally, while also providing electrical isolation and protection between the control terminal and the actuator. This drive circuit typically includes an energy storage capacitor and a power switching transistor to ensure the supply of instantaneous high current even if the main power supply is damaged. An electronic system is an organic whole composed of various electronic components and circuit modules combined according to specific functional requirements, capable of signal acquisition, transmission, processing, storage, control, or energy conversion. A circuit module is an independent circuit unit with specific electrical functions and a standardized design. An electric detonator is a device that utilizes electrical energy... Small pyrotechnic devices that trigger explosions convert electrical energy into heat or shock waves, triggering subsequent explosions, ignition, or work processes; the bridge wire inside an electric detonator is usually made of a thin metal wire with high resistance. When a specific intensity of current is input, the current passing through the bridge wire will rapidly heat up due to the Joule effect (i.e., the current flowing through the resistance generates heat), causing the bridge wire temperature to rise sharply in a short time; a sound and light distress signal unit is a device used to send out distress signals in emergency situations. It can transmit distress information to the surrounding environment or a specific recipient through a combination of sound and light, helping users to be found in distress. It is commonly used in safety protection, emergency rescue, and other scenarios; LED lights (Light-Emitting) A light-emitting diode (LED) is a new type of lighting device based on the principle of semiconductor light emission. When current passes through a semiconductor chip, electrons and holes recombine within the chip, releasing energy in the form of photons, thus producing visible light. The flashing pattern of the international rescue standard conveys a distress signal through regular, easily identifiable light signals, avoiding confusion with natural light sources or non-rescue signals such as lightning or reflections. Its core standard is based on internationally accepted distress signal logic. A piezoelectric buzzer is an electronic component that uses the piezoelectric effect to produce sound. The piezoelectric effect refers to the fact that its key component is a piezoelectric ceramic sheet. This type of material has special physical properties; when an alternating voltage is applied externally, the piezoelectric ceramic will produce mechanical vibrations as the voltage polarity changes (inverse piezoelectric effect). Conversely, if mechanical force is applied to it to make it vibrate, voltage will also be generated (positive piezoelectric effect, the inverse effect is mainly used in buzzers); encapsulation refers to organizing scattered information into frames according to a fixed format to ensure that the receiver can understand and correctly parse it; the flight data recorder, commonly known as the black box, although its shell is mostly orange for easy searching, is the core equipment on the aircraft used to record critical flight information. It is mainly used to trace the cause after an aviation accident and can also assist in daily flight safety analysis; the communication system is a technical system used to realize the effective transmission and exchange of information from the sender to the receiver. Its goal is to break through space limitations and ensure that information is transmitted accurately, efficiently, and reliably; the ground station is the ground infrastructure used for data communication, tracking measurement, and control management with spacecraft.

[0063] In this embodiment, flight status data packets are acquired via a reliable link. Based on the system health status code in the flight status data packets, the alert level and sensitivity threshold of the fall detection algorithm of the wearable safety device are adjusted. Based on the local sensor data of the wearable safety device and the real-time flight status data packets, a collaborative fall detection is performed to generate a trigger decision. When the trigger decision confirms a fall with high confidence, a protection command is generated to instruct the execution of multiple protective and collaborative actions, such as airbag triggering, audible and visual alarms, and information feedback. By fusing local motion information from the wearable device with global status information from the aircraft, the original information silo status of wearable devices can be completely broken, significantly improving the accuracy and timeliness of identifying complex low-altitude fall scenarios, greatly reducing the risk of false alarms and missed alarms. Moreover, by leveraging the sensitivity dynamic adjustment mechanism based on aircraft fault warning, a mode transition from passive impact response to active warning and collaborative protection is achieved, enhancing the reliability, adaptability, and overall safety level of the occupant safety protection system in low-altitude manned scenarios.

[0064] In one embodiment, based on the system health status code in the flight status data packet, the alert level and the sensitivity threshold of the fall detection algorithm of the wearable safety device are adjusted to obtain the adjusted alert level and the adjusted sensitivity threshold, including:

[0065] Step S201: Based on the system health status code in the flight status data packet and the list of preset serious fault alarm codes in the wearable safety device, determine whether the current low-altitude aircraft is in a serious fault state and obtain the judgment result.

[0066] The preset list of critical fault alarm codes is essentially a lookup table or a set of rules that defines which specific status code values ​​or code combinations are classified as critical faults that require the highest level of attention. For example, codes representing complete failure of the multi-rotor power unit, flight control computer crash, or critical alarms related to the integrity of the airframe structure.

[0067] For example, after periodically receiving flight status data packets from the low-altitude aircraft via a communication link, the system health status code field is parsed from them. The parsed real-time system health status codes are then compared item by item or matched according to a pre-defined list of critical fault alarm codes. This comparison process may include exact value matching, range matching, or matching key flag bits after a bitwise AND operation. If the real-time code matches any predefined critical fault mode in the list, a judgment result of "yes" is generated, confirming that the current low-altitude aircraft is in a critical fault state; otherwise, a judgment result of "no" is generated. Here, parsing refers to deconstructing and analyzing the structure, principle, and content of something to clarify its internal logic, essence, or details, making complex or ambiguous objects clear and understandable; rule matching is a logical process based on pre-defined explicit rules / conditions, comparing the target object with the rules to determine whether it meets the matching requirements.

[0068] Step S202: If the judgment result is that the current low-altitude aircraft is in a serious fault state, the internal collaborative warning state variable of the wearable safety device is set from the normal warning state to the first-level warning state to obtain the adjusted warning level.

[0069] Among them, the internal coordinated alert status variable is usually stored in a register and is used to globally represent the current security alert level. Its default initial value is the normal alert state. A register is a high-speed, small storage unit inside the computer's central processing unit, used to temporarily store the data or instructions that the central processing unit is currently computing or processing.

[0070] For example, after determining that the low-altitude aircraft is currently in a state of serious malfunction, the value of the internal coordinated alert status variable is unconditionally updated to Level 1 alert status based on the input judgment result. This assignment operation is not a simple numerical change, but represents a critical switch in the operating mode, meaning a transition from a default, relatively relaxed monitoring mode to a highly sensitive, proactively prepared emergency standby mode. As an accompanying action for this status switch, its status indicators (such as LEDs) can typically be driven to switch from a steady green light indicating normal operation to a flashing yellow light indicating high alert, thereby providing the wearer with an intuitive visual warning.

[0071] Step S203: In response to the adjusted alert level, adjust the key detection threshold parameters in the fall determination algorithm to obtain the adjusted determination sensitivity threshold; wherein, the adjustment includes reducing the synthetic acceleration modulus threshold from a first conventional value to a second higher sensitivity value and reducing the angular velocity modulus threshold from a third conventional value to a fourth higher sensitivity value.

[0072] Among them, the composite acceleration modulus threshold is used to determine whether the equipment is in a state of significant weightlessness or overweight. This modulus can be calculated from the reading of the triaxial accelerometer and is approximately equal to 1g in a stationary state, which is the acceleration due to gravity. The angular velocity modulus threshold is used to determine whether the equipment is undergoing violent or abnormal rotation. Its value can be calculated from the reading of the triaxial gyroscope.

[0073] For example, the adjusted alert level, i.e., the first-level alert state, is input as a control signal into the crash detection algorithm. This algorithm internally manages a set of key detection threshold parameters. Under normal alert conditions, the synthetic acceleration modulus threshold and angular velocity modulus threshold are set to relatively high first and third normal values, respectively, to effectively filter out common interferences during flight, such as minor turbulence and normal maneuvering vibrations, thus reducing the false alarm rate. Upon entering the first-level alert state, a preset set of parameter configurations is invoked, or a dynamic scaling factor is used to reduce these thresholds to the second and fourth higher sensitivity values, respectively. The preset set of parameter configurations refers to a pre-set complete combination of parameters in addition to the default parameter configuration. These parameters are designed around specific functional objectives and are a supplement or differentiation scheme to the default configuration, eliminating the need for manual setting by the user. The dynamic scaling factor is a parameter or mechanism that dynamically adjusts the scaling ratio according to real-time scenarios or target requirements. Its core is to make the scaling behavior no longer a fixed value but flexibly adaptable to changing conditions; its core logic is to adjust as needed.

[0074] In this embodiment, the system health status code issued by the aircraft is parsed and pattern matched to objectively determine whether the aircraft has entered a serious malfunction. Once a malfunction is confirmed, the process immediately drives the internal collaborative alert status variable to the highest level, treating this state switch as a strong event signal. This signal then triggers the dynamic reconstruction of the key detection thresholds within the local crash determination algorithm, adjusting its sensitivity preset value to a higher level. This significantly reduces the detection delay and missed detection probability of subsequent dangerous motion patterns, fundamentally shifting the response timing of the protection system from post-impact triggering to pre-fault warning and collaborative preparation. This provides crucial algorithm preparation time and decision margin for reliably triggering protective measures at the most critical moment.

[0075] In one embodiment, such as Figure 3 As shown, based on local sensor data and real-time flight status data packets from wearable safety devices, a collaborative fall determination is performed to generate triggering decisions, including:

[0076] Step S301: Collect raw data from the local sensors of the wearable safety device, and perform preprocessing such as filtering and compensation on the raw data to obtain preprocessed data; wherein, the local sensors include an inertial measurement unit and a barometer.

[0077] The Inertial Measurement Unit (IMU) is a sensor module based on inertial principles, including a three-axis accelerometer and a three-axis gyroscope. It is used to measure the motion attitude and motion state of vehicles such as drones, mobile phones, and cars in real time. It does not rely on external signals and can achieve autonomous positioning and attitude perception. The barometer is an instrument that measures atmospheric pressure. It is used to reflect changes in atmospheric pressure by detecting the pressure value of the air column per unit area. The three-axis accelerometer is a sensor that can simultaneously detect acceleration in three orthogonal directions (X, Y, and Z) in space. It is used to convert physical acceleration signals into electrical signals and finally output digital / analog data that can be recognized by the device. The three-axis gyroscope is an inertial sensor that can simultaneously detect the angular rate (rotational motion) of an object in three orthogonal axes (X, Y, and Z). It is used to sense the three rotational movements of the device: roll, pitch, and yaw, which makes up for the limitation of the accelerometer, which can only detect linear motion.

[0078] For example, data can be acquired synchronously via an inertial measurement unit (IMU) and a barometer to obtain unprocessed raw voltage or digital readings. The IMU's triaxial acceleration and angular velocity data are then digitally low-pass filtered to smooth high-frequency noise and retain low-frequency components reflecting macroscopic human motion. Based on built-in temperature sensor readings or a pre-stored temperature-error model, the zero bias and scaling factor of the triaxial accelerometer and gyroscope are compensated in real time to correct measurement deviations caused by changes in ambient temperature. Similarly, temperature compensation and dynamic filtering can be applied to the barometer readings to suppress instantaneous pressure fluctuations caused by cabin airflow disturbances. The final result is pre-processed data: triaxial acceleration values ​​(typically meters per second squared), triaxial angular velocity values ​​(typically radians per second), and pre-calibrated pressure values ​​(typically Pascals). Among them, synchronous data acquisition refers to a data acquisition method in which the acquisition end and the data generation / transmission rhythm are kept in real time. Simply put, the acquisition action is triggered immediately as soon as data is generated, and the acquisition process is synchronized with data generation and flow without significant delay. Digital low-pass filtering refers to a digital signal processing method that retains low-frequency components in digital signals and filters out high-frequency noise / interference. It is implemented entirely through algorithms that calculate discrete digital sampling points. Pre-stored temperature-error model refers to a mathematical model, such as linear or quadratic, that is fitted to the relationship between temperature value and corresponding error value by testing the zero bias and scaling factor of the sensor at different temperatures through experimental calibration in advance. Type; Temperature sensor reading refers to the current operating temperature of the IMU acquired in real time by the temperature sensor, directly relating the relationship between temperature and error; The temperature sensor is a sensing element that converts the physical quantity of temperature into an electrical signal that can be recognized by the circuit, used to realize temperature detection, monitoring and control; Real-time compensation dynamically updates the estimated values ​​of zero bias and scaling factor according to the sensor's operating state, environment (temperature, vibration) and time changes, rather than using fixed calibration parameters, which can effectively suppress the time drift (changing with time), temperature drift (changing with temperature), and motion drift (changing with motion state) of sensor errors, and adapt to the complex scenarios of dynamic movement of the carrier.

[0079] Step S302: Based on the preprocessed data, calculate the local feature vector sequence; wherein the local feature vector sequence includes the synthetic acceleration modulus, weightlessness index and angular velocity modulus.

[0080] For example, based on the preprocessed data, the synthetic acceleration modulus is calculated, which is the square root of the sum of squares of the three-axis acceleration components. This scalar reflects the magnitude of the net external force acting on the wearable device. It approximates the gravitational acceleration g when at rest or in uniform motion, but deviates significantly during free fall or violent collisions. The weightlessness index is calculated, which assesses how close the average value of the synthetic acceleration modulus over a short sliding time window is to the gravitational acceleration g: the mean modulus is continuously calculated within the window; the closer the mean is to 0, the deeper the weightlessness, a key indicator for determining the free fall phase. The angular velocity modulus is calculated, which is the square root of the sum of squares of the three-axis angular velocity components. This scalar directly characterizes the degree of rotation of the wearable device (i.e., the human body) in space; an abnormally high angular velocity modulus usually indicates loss of posture or tumbling. These calculations are performed continuously at the same rate as the acquisition frequency, generating a time-sequential sequence of local feature vectors, where each vector contains a snapshot of the motion characteristics at the current moment. Among them, the short-time sliding window is a technique used for statistics and calculation in streaming data processing. Its core is to define a short time interval of fixed duration (such as 5 seconds or 1 minute) as the calculation window. The window will slide forward continuously at a fixed time step. After each slide, calculations are only performed based on the latest data in the window, and expired data is discarded. It is suitable for short-cycle statistical scenarios with high real-time requirements.

[0081] Step S303: parse the real-time flight status data packet to obtain aircraft status evidence; wherein, the aircraft status evidence includes the aircraft's real-time altitude, altitude descent rate, and fault alarm flags.

[0082] The real-time altitude of an aircraft is usually derived from high-precision navigation calculations that integrate data from the Global Navigation Satellite System (GNSS) and barometers; the three-dimensional velocity vector is used to calculate the velocity component in the vertical direction; the Global Navigation Satellite System is a general term for satellite navigation and positioning systems that can provide all-weather, high-precision position, velocity, and time information to various users worldwide, and does not refer to a single system.

[0083] For example, real-time flight status data packets transmitted via a communication link can be monitored and received at a frequency lower than that of local sensors. These data packets have a predefined binary frame structure, including a frame header, data payload, and a Cyclic Redundancy Check (CRC) code. Upon receiving the data packet, its CRC is first verified to ensure error-free data transmission. Then, the data payload is parsed according to a known protocol format to extract key aircraft status evidence, including the aircraft's real-time altitude, three-dimensional velocity vector, and fault alarm flags directly set by the flight control system. Based on the altitude value parsed from the previous cycle and its corresponding timestamp, when a new altitude value is parsed, the current altitude value is subtracted from the cached altitude value, and then divided by the actual time interval between the two data packet arrivals to obtain the instantaneous altitude drop rate, which is typically negative; its absolute value is used for comparison. The parsed real-time altitude, the calculated altitude drop rate, and the fault alarm flags (such as collision alarms) extracted directly from the data packets are combined to generate aircraft status evidence for collaborative decision-making. Verification refers to checking and confirming the authenticity, validity, compliance, and accuracy of things through specific methods and based on established standards to verify whether they meet preset requirements. It is a process of eliminating falsehoods and ensuring reliability. The known protocol format refers to the data rules agreed upon in advance by the two communicating parties, including the data composition structure, field order, field length, encoding method, separators / identifiers, etc., which is equivalent to the common language of communication between the two parties. The flight control system is the "brain + nerve center" of the aircraft. It is used to stabilize the attitude of the aircraft, control the flight trajectory, and make the aircraft fly according to instructions through perception, calculation, and execution, while ensuring flight safety. It is compatible with various aircraft such as drones and spacecraft.

[0084] Step S304: Based on the local feature vector sequence, generate a local fall suspicion detection result by detecting the composite pattern of continuous weightlessness followed by abnormal angular velocity growth.

[0085] Among them, the composite mode of continuous weightlessness followed by abnormal angular velocity growth is a combination of mechanical / kinematic states superimposed with a weightless environment and a sudden, unexpected rapid increase in angular velocity. The core is to first be in a continuous weightless (zero gravity / microgravity) unsupported state, and then suddenly superimposed with rotational acceleration around a certain axis. It is commonly seen in aerospace and manned spacecraft attitude loss or simulated weightlessness test scenarios.

[0086] For example, based on a continuous sequence of local feature vectors, a local fall suspicion detection result can be generated by detecting a composite pattern of continuous weightlessness followed by an abnormal increase in angular velocity.

[0087] Step S305: Based on the aircraft status evidence, determine whether there is strong anomaly evidence for the low-altitude aircraft, and obtain the result of the determination of the existence of strong anomaly evidence for the aircraft; among which, strong anomaly evidence is a cliff-like drop in altitude, an altitude drop rate exceeding the normal landing threshold, or an active collision warning signal from the aircraft.

[0088] Among them, the normal landing threshold refers to the upper limit benchmark value of the descent rate set in advance according to the aircraft type, design standards, and landing approach operation specifications. It corresponds to the safe maximum descent rate during the normal flight phase (with landing approach as the core) and is a key preset indicator for determining whether the aircraft's descent state is within the normal operating range.

[0089] For example, based on the obtained aircraft status evidence, logical judgments are performed to generate a result determining the existence of strong anomaly evidence. The logical judgments primarily rely on three paths: First, determining whether there is a precipitous drop in altitude. This depends not only on the altitude drop rate calculated in a single instance, but also on analyzing the altitude sequence of multiple consecutive data packets. If a monotonous and rapid decrease in altitude is found within a short period, and the total decrease exceeds a preset threshold for significant change, then this condition is considered met. Second, determining whether the altitude drop rate exceeds a normal landing threshold. This is done by comparing the calculated instantaneous altitude drop rate with the normal landing threshold; if the absolute value of the former is consistently greater than the latter, then this condition is considered met. Third, directly checking the fault alarm flags parsed from the aircraft data packets to see if they contain a clear active collision alarm signal. This is a high-priority alarm directly triggered by aircraft sensors (such as a three-axis accelerometer) or the aircraft vision system. If any one of the above three judgment paths is true, strong anomaly evidence is determined, and a positive judgment result is output; otherwise, a negative judgment result is output. Logical judgment, based on established logical rules, objective facts, or preconditions, uses reasoning, analysis, induction, deduction, and other modes of thinking to make definite or probable judgments about the relationships, properties, and outcomes of things. A preset threshold for significant changes refers to a pre-set critical value used to determine whether the changes observed in the research / analysis are statistically / practically significant changes, rather than random fluctuations. Short time is a time interval defined according to the scenario. An aircraft vision system refers to an optical and image processing system installed on aircraft (such as drones, airplanes, spacecraft, etc.) to perceive, identify, and understand the surrounding environment.

[0090] Step S306: Based on the local crash suspicion detection results and the aircraft strong anomaly evidence existence judgment results, combined with the preset fusion decision rules, a trigger decision is generated.

[0091] The pre-defined fusion decision rules are typically implemented in the form of a conditional decision tree. For example, one core rule is: if the local detection result is "confirmed crash", then regardless of the aircraft evidence, a high-confidence confirmation crash trigger decision is generated. Another core rule is: if the local detection result is "suspected crash", and the aircraft strong anomaly evidence existence judgment result is "existence", then a high-confidence confirmation crash trigger decision is also generated. Furthermore, if the local result is only "suspected" and the aircraft evidence is temporarily "non-existent", but the current state is at the first level of alert, then a high-confidence decision may also be generated or a very short secondary confirmation waiting period may be entered.

[0092] For example, based on the local fall suspicion detection results, such as "none", "suspected", "confirmed", and the results of the strong anomaly evidence of the aircraft, such as "existence" or "non-existence", combined with other contextual parameters such as the current collaborative alert level, a definite trigger decision is finally generated by executing a preset fusion decision rule. This decision not only includes a binary conclusion of whether to trigger, but also usually has a confidence level, such as high, medium, or low, to characterize the sufficiency and reliability of the information on which this judgment is based.

[0093] In this embodiment, a temporal feature sequence reflecting microscopic changes in human motion is constructed; global status messages from the aircraft are analyzed to extract key evidence such as altitude change rate and alarm flags. Local fall suspicion detection based on temporal patterns and strong aircraft anomaly evidence judgment based on physical thresholds are performed separately. Through preset fusion decision rules, locally perceived suspicions and aircraft-provided anomaly evidence are logically correlated and comprehensively evaluated to generate a final trigger decision. This approach allows local motion characteristics and global flight status to be placed within the same decision framework for mutual verification and supplementation, thereby enhancing the perception and accuracy of real fall events in complex and uncertain low-altitude flight environments, effectively suppressing false alarms caused by local motion interference, and significantly reducing missed detections of concealed or coupled fall patterns.

[0094] In one embodiment, based on the local crash suspicion detection result and the aircraft strong anomaly evidence existence judgment result, combined with a preset fusion decision rule, a trigger decision is generated, including:

[0095] Step S401: If the local crash suspicion judgment result is suspected crash or confirmed crash, and the aircraft strong anomaly evidence existence judgment result is strong anomaly evidence existence, generate a trigger decision with high confidence.

[0096] For example, the system is based on a local crash suspicion assessment result (which may have values ​​of none, suspected crash, or confirmed crash) and a binary yes / no assessment result regarding the existence of strong anomaly evidence of the aircraft. When both the local result (suspected or confirmed) and the strong anomaly evidence of the aircraft are true, the highest priority decision logic is initiated. This means that without any additional delay or confirmation, a trigger decision with a high confidence level is immediately generated, concluding that triggering is necessary. This decision is a structured data object, including a trigger command, a timestamp, and a high confidence flag. The technical principle behind executing high-priority decision-making logic is that two Boolean conditions being true simultaneously means that abnormal alarms are simultaneously obtained from two independent and physically different data sources: the microscopic movement of the human body and the macroscopic state of the aircraft. This constitutes strong cross-validation. For example, a local sensor detects a human body in a state of weightlessness and tumbling (suspected / confirmed), while the aircraft data simultaneously reports a precipitous drop in altitude or has sent a collision alarm. This double confirmation is used to largely eliminate the possibility of false alarms from a single sensor or local interference, clearly pointing to a real, serious fall accident involving the entire "human-machine" system. The "human-machine" system refers to the organic whole formed by the interaction of information, energy, and matter between humans and machines to achieve a specific goal.

[0097] Step S402: If the local crash suspicion judgment result is suspected crash or confirmed crash, but the strong anomaly evidence existence judgment result is strong anomaly evidence does not exist, obtain the current collaborative alert status variable.

[0098] For example, when the fusion condition is not met—that is, when the local crash suspicion judgment result is suspected or confirmed as true, but the strong anomaly evidence existence judgment result of the aircraft is false—this situation indicates that its own sensors detected a strong abnormal motion pattern, consistent with the local characteristics of a crash. However, within the current reception period, the data from the aircraft has not yet reflected the corresponding global catastrophic anomaly. A high-confidence conclusion cannot be immediately drawn because there is a possibility that the local sensors have been subjected to severe but not crash-related special interference, such as severe convulsions caused by a sudden illness of the occupants in the cabin, or accidental impact to the equipment. A broader range of contextual information can be obtained by querying a global software variable maintained internally, namely the current collaborative alert status variable, which is a prospective state assessment based on the early health status of the aircraft. Here, "maintenance" is a partial verb meaning to continuously care for, repair, and protect something to maintain its original state, function, or rights; it can also refer to taking actions to ensure that order, relationships, etc., are not disrupted.

[0099] Step S403: If the collaborative alert state variable is at level 1 alert, generate a trigger decision with high confidence.

[0100] The Level 1 alert status itself is a strong risk warning signal. It is triggered by a serious fault alarm sent by the aircraft earlier (such as power failure or flight control anomaly). It indicates that the aircraft is in a high-risk condition and the probability of crashing is extremely high. In this context, when the local sensors detect a motion pattern that is suspected to be crashing, even if the real-time telemetry data of the aircraft has not yet captured decisive evidence of crashing, it may be due to sensor limitations, data fusion delay, or the fault mode not yet being fully manifested. It is also likely that the local anomaly will be interpreted as a high-risk fault evolving into an actual crash.

[0101] For example, if the current collaborative alert status variable is at level one alert, then the detection results of local suspected or confirmed falls are given higher weight, sufficient to trigger protective actions. Based on this, a trigger decision with high confidence is generated and output.

[0102] Step S404: If the coordinated alert status variable is the normal alert status, continuously monitor the impact signal and the aircraft confirmation anomaly information within the preset confirmation waiting window to obtain the monitoring results.

[0103] The preset confirmation waiting window is essentially a waiting period with a countdown (e.g., 10 seconds). During this period, you need to wait for the final confirmation instruction to confirm execution or cancel execution. For example, when deleting files on a mobile phone, a pop-up window will automatically delete the file after 10 seconds. Clicking cancel will terminate the window. The core is "timer + waiting for confirmation". The normal alert status means that no serious fault warnings of the aircraft have been received before this local anomaly occurred, and the aircraft is in a relatively safe operating baseline by default.

[0104] For example, when the judgment condition is not met, i.e., the current collaborative alert status variable is in a normal alert state, facing a situation with only local unilateral suspected or confirmed evidence and a lack of aircraft collaborative verification, the system chooses not to trigger immediately, but instead initiates a risk control mechanism by opening a preset confirmation waiting window with a very short duration. During this window, two continuous monitoring tasks are executed in parallel: first, high-frequency monitoring of local sensor data streams, specifically detecting instantaneous acceleration peaks exceeding a preset extremely high impact threshold, i.e., local impact signals, which correspond to the most direct physical requirement of airbag protection, i.e., a collision; second, continuous monitoring of data packets from the aircraft communication link to check whether any new aircraft confirmation anomaly information is received during the window, such as newly delivered warning signs indicating collision or loss of control, or data confirming a sudden drop in altitude. Among them, the risk control mechanism is a systematic and process-oriented management system established by an organization to identify, assess, respond to and monitor various risks and ensure the achievement of business / operational goals. Through a closed-loop design of proactive prevention, process control and post-event remediation, risks are controlled within an acceptable range to avoid or reduce the losses they cause. The preset extremely high impact threshold refers to the acceleration red line set in advance based on the tolerance and safety standards of the protected object. This threshold is much higher than the tolerance value of conventional impacts. Breaking through it means reaching the extreme impact level.

[0105] Step S405: If the monitoring result is that a local impact signal or aircraft confirmation anomaly information is detected within the preset confirmation waiting window period, a trigger decision with high confidence is generated.

[0106] For example, if, before the window expires, monitoring results indicate the detection of a local impact signal indicative of a severe collision, or the receipt of confirmed anomaly information from the aircraft, it means that the situation has further deteriorated or received critical verification during the brief delayed observation period. Both the appearance of a local impact signal and the belated but definitive confirmation of the anomaly from the aircraft strongly corroborate the severity of the initial suspected local signal, ruling out the possibility of it being a transient interference. Therefore, once such a signal is detected, the waiting window should be immediately terminated, and a high-confidence trigger decision should be generated.

[0107] Step S406: If the monitoring result is that no local impact signal or aircraft confirmation anomaly information is detected within the preset confirmation waiting window period, a trigger decision with a confidence level of zero is generated.

[0108] For example, if the preset confirmation waiting window expires and the condition is not triggered—that is, if no local impact signal is detected and no aircraft confirmation of anomaly is received during the entire window—this monitoring result will be summarized as "not detected." This situation may correspond to a false alarm, such as a local anomaly caused by a brief, violent non-fall maneuver or sensor transient malfunction, and this anomaly does not develop into an actual collision, and the aircraft status remains normal. A trigger decision with zero confidence is generated. This decision is essentially a no-trigger command, or a decision object with no event identifier. Based on this, the internal suspected status flags will be cleared, and the system will return to the normal continuous monitoring cycle.

[0109] In this embodiment, by judging the synchronization of local and aircraft evidence, millisecond-level zero-delay triggering is achieved when dual confirmation is obtained. When the evidence is not synchronized, a priori warning level is introduced as a key arbitration factor, allowing for rapid response based on strong unilateral evidence in high-risk warning situations. If the situation is in a normal baseline state and faces insufficient evidence, an extremely short dynamic confirmation window is initiated as a buffer. Within this window, direct evidence of impact or final confirmation by the aircraft is actively sought to complete the decision-making loop; otherwise, it is considered a false alarm and the system safely exits. This approach can minimize the risk of missed alarms while suppressing the probability of false triggering to an extremely low level in complex and ever-changing real low-altitude environments, thus constructing an intelligent safety decision-making capability that combines agility, robustness, and high reliability.

[0110] In one embodiment, based on a local feature vector sequence, a local fall suspicion detection result is generated by detecting a composite pattern of sustained weightlessness followed by an abnormal increase in angular velocity, including:

[0111] Step S501: Determine whether the weightlessness index in the local feature vector sequence meets the continuous weightlessness condition, and obtain the first judgment result; wherein, the continuous weightlessness condition is that the weightlessness index is continuously lower than the preset dynamic threshold for a preset duration.

[0112] The weightlessness index is calculated by averaging the synthetic acceleration modulus over a sliding time window and comparing it with the gravitational acceleration g. The closer the value is to 0, the deeper the weightlessness. The dynamic threshold is selected from a preset parameter table based on the current cooperative alert state variable (normal or Level 1 alert). The threshold is lower in the Level 1 alert state to improve sensitivity. The preset parameter table is a set of parameters that are defined and stored in advance. The thresholds for the normal state and the Level 1 alert state are already set in the table. There is no need for real-time calculation. The corresponding values ​​can be retrieved directly according to the current state. The preset duration is a fixed time length that is set in advance to distinguish between short-term weightlessness fluctuations (such as slight turbulence of the aircraft) and a truly dangerous continuous free fall trend.

[0113] For example, a series of current and historical weightlessness index values ​​are compared with a preset dynamic threshold, and the duration of the weightlessness index continuously falling below the dynamic threshold is monitored to see if it reaches a preset duration. If both the value falling below the threshold and the duration meeting the standard are met, the condition of continuous weightlessness is determined to be true, and a first judgment result of "true" is obtained; otherwise, a first judgment result of "false" is obtained.

[0114] Step S502: If the first judgment result is that the weightlessness index meets the continuous weightlessness condition, determine whether the angular velocity magnitude in the local feature vector sequence meets the abnormal angular velocity growth condition, and obtain the second judgment result; wherein, the abnormal angular velocity growth condition is that it exceeds the preset dynamic angular velocity threshold within the time window after the weightlessness stage.

[0115] Among them, the preset dynamic angular velocity threshold and the weightlessness threshold (preset dynamic threshold) are similar, and are also dynamic, adjusting with the alert level, with the threshold being lower at high alert levels.

[0116] For example, when the first judgment result is true, i.e., after confirming the continuous weightlessness phase, the detection of typical subsequent phases during the fall, i.e., attitude loss of control, is performed. This is based on the angular velocity modulus data stream in the local feature vector sequence, and the start time or time period information of the weightlessness phase identified by the previous steps. Immediately after the start of the weightlessness phase, a time window is opened. The duration of this window is predefined to limit a reasonable time range for finding abnormal rotation after weightlessness occurs. Within this time window, the real-time calculated angular velocity modulus is continuously monitored and compared with another preset dynamic angular velocity threshold. It is checked whether any angular velocity modulus sample value exceeds this dynamic threshold within the time window. If an angular velocity modulus exceeding the limit is detected within the specified window period, it is determined that the abnormal angular velocity growth condition is met, and a second judgment result of "true" is obtained; if all angular velocity values ​​remain below the threshold within the window period, it is determined that the condition is not met, and a second judgment result of "false" is obtained.

[0117] Step S503: If the second judgment result is that the angular velocity modulus meets the abnormal angular velocity growth condition, generate a local fall suspicion judgment result.

[0118] For example, if the second judgment result is true, it means that two key signs—persistent weightlessness and the subsequent abnormal increase in angular velocity—have been detected sequentially. According to the physical model of low-altitude falls, this "weightlessness followed by tumbling" sequence is a very strong local indication signal of a fall event, significantly different from most normal flight maneuvers or ordinary disturbances. That is, a local fall suspicion judgment result is generated as a suspected fall. The physical model of low-altitude falls is based on classical mechanics' free fall + impact impulse. Due to the characteristics of low altitude, air resistance can be approximately ignored, and there is no obvious acceleration phase in the fall process. The core focus is on the velocity upon impact and the impact force / deceleration process after impact; it is a simplified model of uniformly accelerated linear motion.

[0119] Step S504: Based on the local fall suspicion judgment result, determine whether there is a peak value in the synthetic acceleration modulus of the local feature vector sequence that exceeds the preset extremely high impact threshold, and obtain the third judgment result.

[0120] For example, based on the synthetic acceleration modulus data stream in the local feature vector sequence and the context information of the current suspected fall state, the system detects whether there is an instantaneous peak value exceeding a preset extremely high impact threshold. This extremely high impact threshold is a very high fixed or quasi-static threshold, much higher than the threshold used to detect weightlessness or general vibrations. Its purpose is to specifically identify extremely severe acceleration impacts that typically only occur during high-speed collisions with the ground or other hard objects. The real-time acceleration modulus is compared point-by-point with this threshold. If, during the period of the suspected fall state, the acceleration modulus at any sampling point exceeds this extremely high threshold, a third judgment result of "true" is obtained, indicating that decisive impact evidence has been detected; otherwise, it is false.

[0121] Step S505: If the third judgment result is that the synthetic acceleration modulus has a peak value exceeding the preset extremely high impact threshold, generate a local fall suspicion judgment result confirming the fall.

[0122] For example, if the third judgment result is true, it means that in a state of high alert for a suspected fall, a peak value of the synthetic acceleration modulus, used to characterize a violent collision, has been further captured. The appearance of this impact peak value is temporally related to the previous suspected pattern of "weightlessness followed by tumbling," forming a coherent chain of a complete fall event: "initial weightlessness -> attitude loss -> final impact." Based on this complete local chain of evidence, it is highly certain that a fall event has occurred and a collision has taken place. The local fall suspicion judgment result will be immediately updated, upgrading from suspected fall to confirmed fall.

[0123] In this embodiment, by simulating and identifying a time-series event chain that conforms to physical laws, from continuous weightlessness to abnormal angular velocity growth, and then to extremely high impact acceleration, false alarms caused by single instantaneous anomalies can be effectively filtered out, improving the specificity of local detection and enhancing the autonomy, robustness, and hierarchical decision-making of the entire protection system in complex and uncertain scenarios.

[0124] It should be understood that although the steps in the flowcharts of the embodiments described above are shown sequentially according to the arrows, these steps are not necessarily executed in the order indicated by the arrows. Unless explicitly stated herein, there is no strict order restriction on the execution of these steps, and they can be executed in other orders. Moreover, at least some steps in the flowcharts of the embodiments described above may include multiple steps or multiple stages. These steps or stages are not necessarily completed at the same time, but can be executed at different times. The execution order of these steps or stages is not necessarily sequential, but can be performed alternately or in turn with other steps or at least some of the steps or stages of other steps.

[0125] Based on the same inventive concept, this application also provides a low-altitude fall intelligent collaborative protection system for implementing the aforementioned low-altitude fall intelligent collaborative protection method. The solution provided by this system is similar to the implementation scheme described in the above method; therefore, the specific limitations of one or more embodiments of the low-altitude fall intelligent collaborative protection system provided below can be found in the limitations of the low-altitude fall intelligent collaborative protection method described above, and will not be repeated here.

[0126] In one exemplary embodiment, such as Figure 3 As shown, a low-altitude fall protection intelligent collaborative protection system 300 is provided, comprising:

[0127] The data acquisition module 301 is used to establish a real-time data communication link between the wearable safety device and the low-altitude aircraft, and periodically receive flight status data packets from the low-altitude aircraft through the real-time data communication link; wherein, the flight status data packets include aircraft altitude, speed, attitude and system health status code;

[0128] The algorithm adjustment module 302 is used to adjust the warning level and the judgment sensitivity threshold of the fall determination algorithm of the wearable safety device based on the system health status code in the flight status data packet, so as to obtain the adjusted warning level and the adjusted judgment sensitivity threshold.

[0129] The decision generation module 303 is used to perform collaborative fall determination and generate trigger decisions based on local sensor data and real-time flight status data packets from wearable safety devices.

[0130] The protection trigger module 304 is used to generate a protection command when the trigger decision is a high-confidence confirmation of a fall. The protection command is used to instruct the execution of relevant protection work. The protection work includes controlling the gas generator of the wearable safety device to deploy the airbag, activating the local audible and visual alarm, and sending trigger confirmation information to the low-altitude aircraft through a real-time data communication link.

[0131] In one embodiment, the algorithm adjustment module 302 is further configured to:

[0132] Based on the system health status code in the flight status data packet and the list of preset serious fault alarm codes in the wearable safety device, it is determined whether the low-altitude aircraft is in a serious fault state, and the judgment result is obtained.

[0133] If the judgment result is that the current low-altitude aircraft is in a serious malfunction state, the internal collaborative alert status variable of the wearable safety device is set from the normal alert state to the first-level alert state to obtain the adjusted alert level.

[0134] In response to the adjusted alert level, the key detection threshold parameters in the fall determination algorithm are adjusted to obtain the adjusted determination sensitivity thresholds; the adjustment includes reducing the synthetic acceleration modulus threshold from a first conventional value to a second higher sensitivity value and reducing the angular velocity modulus threshold from a third conventional value to a fourth higher sensitivity value.

[0135] In one embodiment, the decision generation module 303 is further configured to:

[0136] Raw data from the local sensors of the wearable safety device is collected, and the raw data is preprocessed by filtering and compensation to obtain preprocessed data; the local sensors include an inertial measurement unit and a barometer.

[0137] Based on the preprocessed data, a local feature vector sequence is calculated; the local feature vector sequence includes the synthetic acceleration modulus, weightlessness index, and angular velocity modulus.

[0138] Parse real-time flight status data packets to obtain aircraft status evidence, which includes the aircraft's real-time altitude, rate of altitude descent, and fault alarm flags.

[0139] Based on the local feature vector sequence, a local fall suspicion detection result is generated by detecting the composite pattern of continuous weightlessness followed by abnormal angular velocity growth.

[0140] Based on the aircraft status evidence, it is determined whether there is strong anomaly evidence for the low-altitude aircraft, and the result of the determination of the existence of strong anomaly evidence is obtained; among them, strong anomaly evidence is a cliff-like drop in altitude, an altitude drop rate exceeding the normal landing threshold, or an active collision warning signal of the aircraft.

[0141] Based on the local crash suspicion detection results and the judgment results of the existence of strong anomaly evidence of the aircraft, combined with the preset fusion decision rules, a trigger decision is generated.

[0142] In one embodiment, the decision generation module 303 is further configured to:

[0143] If the local crash suspicion assessment result is a suspected crash or a confirmed crash, and the aircraft strong anomaly evidence existence assessment result is strong anomaly evidence existence, a trigger decision with high confidence is generated.

[0144] If the local crash suspicion assessment result is suspected crash or confirmed crash, but the strong anomaly evidence of the aircraft assessment result is no strong anomaly evidence, obtain the current collaborative alert status variable;

[0145] If the collaborative alert status variable is at level one alert, generate a trigger decision with high confidence.

[0146] If the collaborative alert status variable is the normal alert status, the impact signal and the aircraft confirmation anomaly information are continuously monitored within the preset confirmation waiting window to obtain the monitoring results.

[0147] If the monitoring results indicate that a local impact signal or aircraft confirmation anomaly information is detected within the preset confirmation waiting window period, a trigger decision with high confidence is generated.

[0148] If the monitoring result is that no local impact signal or aircraft confirmation anomaly information is detected within the preset confirmation waiting window period, a trigger decision with a confidence level of zero is generated.

[0149] In one embodiment, the decision generation module 303 is further configured to:

[0150] Determine whether the weightlessness index in the local feature vector sequence meets the condition of continuous weightlessness to obtain the first judgment result; wherein, the condition of continuous weightlessness is that the weightlessness index is continuously lower than the preset dynamic threshold for a preset duration;

[0151] If the first judgment result is that the weightlessness index meets the condition of continuous weightlessness, then it is judged whether the angular velocity magnitude in the local feature vector sequence meets the condition of abnormal angular velocity growth, and the second judgment result is obtained; wherein, the condition of abnormal angular velocity growth is that it exceeds the preset dynamic angular velocity threshold within the time window after the weightlessness stage.

[0152] If the second judgment result is that the angular velocity modulus meets the abnormal angular velocity growth condition, a local fall suspicion judgment result is generated;

[0153] Based on the local fall suspicion degree judgment result, it is determined whether there is a peak value in the synthetic acceleration magnitude of the local feature vector sequence that exceeds the preset extremely high impact threshold, and a third judgment result is obtained;

[0154] If the third judgment result is that the synthetic acceleration modulus has a peak value exceeding the preset extremely high impact threshold, a local fall suspicion judgment result is generated to confirm the fall.

[0155] In one embodiment, a computer device is provided, including a memory and a processor, the memory storing a computer program, the processor executing the computer program to implement the steps of the intelligent collaborative protection method for low-altitude fall as described above.

[0156] In one embodiment, a computer-readable storage medium is provided having a computer program stored thereon, which, when executed by a processor, implements the steps in the above method embodiments.

[0157] For the device embodiments, since they basically correspond to the method embodiments, the relevant parts can be referred to in the description of the method embodiments. The device embodiments described above are merely illustrative. The components described as separate parts may or may not be physically separate, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this disclosure according to actual needs. Those skilled in the art can understand and implement this without creative effort.

[0158] The above-described embodiments are merely illustrative of several implementation methods of the embodiments of this application, and their descriptions are relatively specific and detailed. However, they should not be construed as limiting the scope of the patent application. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of the embodiments of this application, and these modifications and improvements all fall within the protection scope of the embodiments of this application.

Claims

1. A smart collaborative protection method for low-altitude fall, characterized in that, The method includes: A real-time data communication link is established between the wearable safety device and the low-altitude aircraft, and flight status data packets are periodically received from the low-altitude aircraft through the real-time data communication link; wherein, the flight status data packets include the aircraft's altitude, speed, attitude and system health status code; Based on the system health status code in the flight status data packet, the alert level and the judgment sensitivity threshold of the fall determination algorithm of the wearable safety device are adjusted to obtain the adjusted alert level and the adjusted judgment sensitivity threshold. Based on the local sensor data of the wearable safety device and the real-time flight status data packet, a collaborative fall determination is performed to generate a trigger decision; When the trigger decision is a high-confidence confirmation of a fall, a protection command is generated; wherein, the protection command is used to instruct the execution of relevant protection work; the protection work includes controlling the gas generator of the wearable safety device to deploy the airbag, activating the local audible and visual alarm, and sending trigger confirmation information to the low-altitude aircraft through the real-time data communication link.

2. The method according to claim 1, characterized in that, The step of adjusting the alert level and the sensitivity threshold of the fall detection algorithm of the wearable safety device based on the system health status code in the flight status data packet, to obtain the adjusted alert level and the adjusted sensitivity threshold, includes: Based on the system health status code in the flight status data packet and the list of preset serious fault alarm codes in the wearable safety device, it is determined whether the current low-altitude aircraft is in a serious fault state, and the determination result is obtained. If the judgment result indicates that the current low-altitude aircraft is in the serious fault state, the internal collaborative alert status variable of the wearable safety device is set from the normal alert state to the first-level alert state to obtain the adjusted alert level. In response to the adjusted alert level, the key detection threshold parameters in the fall determination algorithm are adjusted to obtain the adjusted determination sensitivity threshold; wherein, the adjustment includes reducing the synthetic acceleration modulus threshold from a first conventional value to a second higher sensitivity value and reducing the angular velocity modulus threshold from a third conventional value to a fourth higher sensitivity value.

3. The method according to claim 1, characterized in that, The method of collaboratively determining a fall based on local sensor data from the wearable safety device and real-time flight status data packets, and generating a trigger decision, includes: The wearable safety device collects raw data from its local sensors and performs filtering and compensation preprocessing on the raw data to obtain preprocessed data; wherein the local sensors include an inertial measurement unit and a barometer; Based on the preprocessed data, a local feature vector sequence is calculated; wherein, the local feature vector sequence includes the synthetic acceleration modulus, the weightlessness index, and the angular velocity modulus; The real-time flight status data packet is parsed to obtain aircraft status evidence; wherein, the aircraft status evidence includes the aircraft's real-time altitude, altitude descent rate, and fault alarm flags; Based on the local feature vector sequence, a local fall suspicion detection result is generated by detecting the composite pattern of continuous weightlessness followed by abnormal angular velocity growth. Based on the aircraft status evidence, it is determined whether there is strong anomaly evidence for the low-altitude aircraft, and the result of the determination of the existence of strong anomaly evidence is obtained; wherein, the strong anomaly evidence is a cliff-like drop in altitude, an altitude drop rate exceeding the normal landing threshold, or an active collision warning signal of the aircraft. Based on the local crash suspicion detection results and the strong anomaly evidence existence judgment results of the aircraft, combined with the preset fusion decision rules, the trigger decision is generated.

4. The method according to claim 3, characterized in that, The triggering decision is generated based on the local crash suspicion detection result and the aircraft strong anomaly evidence existence judgment result, combined with a preset fusion decision rule, including: If the local crash suspicion judgment result is a suspected crash or a confirmed crash, and the strong anomaly evidence existence judgment result is that the strong anomaly evidence exists, a trigger decision with a high confidence level is generated; If the local crash suspicion judgment result is a suspected crash or a confirmed crash, but the strong anomaly evidence existence judgment result is that the strong anomaly evidence does not exist, obtain the current collaborative alert status variable; If the collaborative alert state variable is in a level 1 alert state, generate the triggering decision with high confidence. If the collaborative alert status variable is a normal alert status, the impact signal and the aircraft confirmation anomaly information are continuously monitored within the preset confirmation waiting window to obtain the monitoring results. If the monitoring result indicates that the local impact signal or the aircraft confirms abnormal information is detected within the preset confirmation waiting window period, a trigger decision with high confidence is generated. If the monitoring result indicates that no local impact signal or aircraft confirmation anomaly information is detected within the preset confirmation waiting window period, the trigger decision with a confidence level of zero is generated.

5. The method according to claim 3, characterized in that, The method of generating a local fall suspicion detection result based on the local feature vector sequence by detecting a composite pattern of continuous weightlessness followed by an abnormal increase in angular velocity includes: Determine whether the weightlessness index in the local feature vector sequence meets the condition of continuous weightlessness to obtain a first determination result; wherein, the condition of continuous weightlessness is that the weightlessness index is continuously lower than a preset dynamic threshold for a preset duration; If the first judgment result is that the weightlessness index meets the continuous weightlessness condition, then it is determined whether the angular velocity magnitude in the local feature vector sequence meets the abnormal angular velocity growth condition, and a second judgment result is obtained; wherein, the abnormal angular velocity growth condition is that the preset dynamic angular velocity threshold is exceeded within the time window after the weightlessness stage. If the second judgment result is that the angular velocity modulus satisfies the abnormal angular velocity growth condition, a local fall suspicion judgment result is generated; Based on the local fall suspicion degree judgment result of the suspected fall, it is determined whether there is a peak value in the synthetic acceleration modulus in the local feature vector sequence that exceeds a preset extremely high impact threshold, and a third judgment result is obtained; If the third judgment result is that the synthetic acceleration modulus has a peak value exceeding a preset extremely high impact threshold, a local fall suspicion judgment result is generated to confirm the fall.

6. A low-altitude fall protection intelligent collaborative protection system, characterized in that, The system includes: The data acquisition module is used to establish a real-time data communication link between the wearable safety device and the low-altitude aircraft, and to periodically receive flight status data packets from the low-altitude aircraft through the real-time data communication link; wherein, the flight status data packets include aircraft altitude, speed, attitude and system health status code; The algorithm adjustment module is used to adjust the alert level and the judgment sensitivity threshold of the fall determination algorithm of the wearable safety device based on the system health status code in the flight status data packet, so as to obtain the adjusted alert level and the adjusted judgment sensitivity threshold. The decision generation module is used to perform collaborative fall determination and generate trigger decisions based on the local sensor data of the wearable safety device and the real-time flight status data packet; The protection trigger module is used to generate a protection command when the trigger decision is a high-confidence confirmation of a fall; wherein the protection command is used to instruct the execution of relevant protection work; the protection work includes controlling the gas generator of the wearable safety device to deploy the airbag, activating the local audible and visual alarm, and sending trigger confirmation information to the low-altitude aircraft through the real-time data communication link.

7. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that, When the processor executes the computer program, it implements the steps of the method according to any one of claims 1 to 5.

8. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, it implements the steps of the method according to any one of claims 1 to 5.

Images