A method for navigation tracking and positioning of distributed aerial targets

Abstract

This invention discloses a distributed aerial target navigation, tracking, and positioning method, belonging to the field of distributed aerial target navigation, tracking, and positioning technology, and is used to solve the problem of distributed positioning instability. This invention constructs an aerial target navigation, tracking, and positioning method that combines distributed wireless observation, reconnection chain determination, positioning subdomain screening, and fragment reconstruction correction. By converting the arrival direction, time delay, and frequency shift information acquired by multiple nodes into direction survey lines, distance constraints, and velocity constraints, an observation primitive sequence is generated, and a reconnection chain for a single aerial target is formed accordingly, achieving continuous target tracking in complex low-altitude scenarios. Continuous tracks are then generated through candidate intersection area screening and continuous positioning point verification. Combined with preset route constraints, path holding, yaw correction, and reconnection guidance states are output. Simultaneously, backup nodes are used to reconstruct fragmented sections and update node correction parameters, thereby improving the continuity, stability, and reliability of navigation, tracking, and positioning results.

Description

Technical Field

[0001] This invention relates to the field of distributed air target navigation, tracking, and positioning technology, and more specifically, to a distributed air target navigation, tracking, and positioning method. Background Technology

[0002] In the field of low-altitude airspace perception and navigation tracking, navigation, tracking and positioning technologies for aerial moving targets such as logistics drones, inspection aircraft, emergency detection platforms, flocks of birds and drones in the airspace surrounding airports have been widely used. These technologies typically use computer programs to coordinate wireless observation equipment distributed on building rooftops, communication towers, ground base stations, airport perimeter areas or runway perimeter monitoring points to receive, analyze and correlate signals emitted or reflected by aerial targets, thereby obtaining the target's azimuth, distance, speed and trajectory information. Combined with preset flight channels, mission areas, airport airspace protection ranges and airspace boundaries, continuous perception and location determination of the target's flight process can be achieved.

[0003] However, existing technologies still have limitations in continuous navigation, tracking, and positioning capabilities in complex low-altitude scenarios. When distributed wireless observation nodes operate in urban buildings, cross-river passages, port parks, mountain patrol areas, and airport surrounding airspace, they are easily affected by factors such as building obstruction, ground reflection, link fluctuations, node response differences, and the coexistence of bird flock movements and UAV maneuvers. The azimuth, latency, and frequency shift information output by different nodes are difficult to maintain consistency in both time and space dimensions. Due to the discontinuity, offset, and misalignment of node observation results, the system struggles to establish a stable and unified target reception relationship among multiple nodes. This can easily lead to problems such as bird flocks and UAV targets crossing paths, track jumps, positioning breaks, and navigation status misjudgments. This inconsistency between distributed observation and continuous flight processes makes it difficult for the system to output continuous, reliable aerial target trajectory results that can be used for airport airspace early warning, stable guidance, and accurate tracking, thereby weakening the reliability of navigation, tracking, and positioning in complex airspace. Summary of the Invention

[0004] To overcome the aforementioned deficiencies in the prior art, the following solution is proposed to address the problem of distributed positioning instability in the aforementioned background technology.

[0005] To achieve the above objectives, the present invention provides the following technical solution:

[0006] A distributed aerial target navigation, tracking, and positioning method includes the following steps:

[0007] The system collects aerial target signals received by each distributed wireless observation node, extracts arrival direction, time delay, frequency shift, node attitude and time stamp, and converts them into direction survey lines, range constraints and velocity constraints under a unified spatiotemporal reference to generate an observation primitive sequence.

[0008] Based on the temporal continuity, spatial adjacency, and velocity continuity between observation primitives, a two-way continuity determination is performed to generate a continuity chain corresponding to a single aerial target;

[0009] The direction survey lines and distance constraints in the docking chain are intersected and solved. The target positioning subdomain is determined by combining the velocity constraints and the spatial boundary. A continuous positioning point sequence is generated based on the target positioning subdomain.

[0010] Connecting consecutive positioning point sequences forms a continuous track, and based on the relationship between the continuous track and preset route constraints, path holding state, yaw correction state, and follow-up guidance state are generated.

[0011] When a link is missing or the navigation status is abnormal, the observation primitives of the backup node are called to reconstruct the broken section, and the node correction parameters are updated according to the reconstruction results, and the navigation tracking and positioning results of the air target are output.

[0012] Furthermore, the aerial target signals received by each distributed wireless observation node are collected, and the direction of arrival, time delay, frequency shift, node attitude, and time stamp are extracted, including:

[0013] Direction detection is performed on the air target signals received by each distributed wireless observation node to obtain the arrival direction of the corresponding air target signal;

[0014] Propagation time is measured for the aerial target signals received by each distributed wireless observation node to obtain the time delay of the corresponding aerial target signals;

[0015] Frequency offset detection is performed on the air target signals received by each distributed wireless observation node to obtain the frequency shift of the corresponding air target signal;

[0016] Read the node attitude and time stamp of each distributed wireless observation node at the corresponding acquisition time;

[0017] The arrival direction, time delay, frequency shift, node attitude, and time stamp are associated and written into the same observation record entry.

[0018] Furthermore, the arrival direction, time delay, frequency shift, node attitude, and time stamp are converted into direction survey lines, distance constraints, and velocity constraints under a unified spatiotemporal reference, generating an observation primitive sequence including:

[0019] Time alignment was performed on the time stamps in each observation record entry based on a unified time reference.

[0020] Based on a unified spatial reference, attitude conversion is performed on the node attitudes in each observation record entry, and the arrival direction is converted into a direction survey line.

[0021] The distance constraints for corresponding aerial targets are generated based on the time delay;

[0022] Generate velocity constraints for corresponding aerial targets based on frequency shift;

[0023] Write the direction survey line, distance constraint, and velocity constraint into the corresponding observation record entries, and attach node identifiers and spatiotemporal reference markers;

[0024] The observation records are arranged in chronological order to generate an observation primitive sequence.

[0025] Furthermore, the continuation determination based on the temporal succession, spatial adjacency, and velocity continuity relationships between observation primitives includes:

[0026] The observation primitive sequence is traversed in chronological order.

[0027] Using the current observation primitive as the predecessor primitive, candidate successor observation primitives are searched within a preset time neighborhood.

[0028] The target position projection range is determined based on the direction survey lines, distance constraints, and velocity constraints of the precursor unit.

[0029] Determine the intersection relationship between the direction survey line corresponding to the candidate successor observation primitive and the target position projection range, and generate spatial adjacency determination result and velocity continuation determination result based on the matching result between the distance constraint and velocity constraint corresponding to the candidate successor observation primitive and the target position projection range;

[0030] A continuation determination result is generated based on the time continuation determination result, the spatial adjacency determination result, and the velocity continuation determination result.

[0031] Furthermore, performing bidirectional succession determination to generate a succession chain corresponding to a single aerial target includes:

[0032] Based on the succession determination results, establish a forward succession relationship between the predecessor primitive and the candidate successor observation primitive;

[0033] Using candidate successor observation primitives as backtracking primitives, reverse retrieval is performed in the neighborhood of the preceding time sequence to establish a reverse continuation relationship between candidate successor observation primitives and predecessor primitives.

[0034] When a candidate successor observation primitive and a predecessor primitive simultaneously satisfy both forward succession and reverse succession relationships, it is determined that there is a bidirectional succession relationship between them.

[0035] Observational primitives with bidirectional succession relationships are concatenated in chronological order to generate candidate succession chains;

[0036] The conflict observation primitives in the candidate succession chain are split and removed to generate succession chains corresponding to a single aerial target.

[0037] Furthermore, the intersection solution is performed on the direction survey lines and distance constraints in the docking chain, and the target localization subdomain is determined by combining velocity constraints and spatial boundaries, including:

[0038] Extract the direction survey lines and distance constraints at the same or adjacent times in the continuation chain;

[0039] Perform spatial intersection calculations on the direction survey lines and distance constraints to generate candidate intersection regions;

[0040] Determine the reachable range of the target's motion at corresponding moments based on velocity constraints;

[0041] Determine the overlap between candidate intersection regions and the reachable range of movement, and delete candidate intersection regions located outside the airspace boundary;

[0042] When there are more than two candidate intersection regions, the target positioning subdomain is determined based on the displacement and direction continuity relationships between each candidate intersection region and the target position at the previous moment.

[0043] Furthermore, generating a sequence of continuous positioning points based on the target positioning subdomain includes:

[0044] The center position of the target positioning subdomain at each time point is determined as a candidate positioning point;

[0045] Based on the displacement and orientation relationships between candidate positioning points at adjacent time points, a continuity check is performed on the candidate positioning points;

[0046] When there are multiple candidate positioning points at the current moment, the candidate positioning point that satisfies the displacement continuation condition and the direction continuation condition with the confirmed positioning point at the previous moment is selected as the positioning point at the current moment.

[0047] Arrange the location points at each time point in chronological order to generate a continuous sequence of location points;

[0048] When the displacement change between adjacent positioning points exceeds the preset jump boundary, the positioning point selection is re-executed according to the target positioning subdomain at the corresponding time, and the continuous positioning point sequence is updated.

[0049] Furthermore, connecting consecutive positioning point sequences forms a continuous track, and based on the relationship between the continuous track and preset route constraints, path-keeping state, yaw correction state, and follow-up guidance state are generated, including:

[0050] Connect the positioning points at adjacent times in a continuous positioning point sequence in chronological order to generate a continuous track;

[0051] Extract the heading change, displacement change, and turning change results of adjacent track segments in a continuous track;

[0052] Determine the deviation relationship between the continuous track and the preset route constraints, and generate a path holding state or yaw correction state based on the deviation relationship;

[0053] Based on the direction of extension of the current flight path segment, the subsequent reachable range, and the pointing relationship of preset route constraints, a follow-up guidance status is generated;

[0054] Write the path holding status, yaw correction status, and follow-up guidance status into the corresponding consecutive track entries.

[0055] Furthermore, when a segment of the connecting chain is missing or the navigation status is abnormal, the reconstruction of the broken segment by calling the observation primitives of the backup node includes:

[0056] When there are time discontinuities in the connection chain, or when there is an abnormal switch in the path holding state, yaw correction state and connection guidance state corresponding to the continuous track, the break section is identified.

[0057] Search for standby node observation primitives that did not participate in the construction of the current successor chain within the time range corresponding to the fracture segment;

[0058] The reconstruction search range is determined based on the front and rear positioning points of the fracture segment, and the continuation determination is performed on the standby node observation elements within the reconstruction search range.

[0059] The standby node observation primitives that meet the continuation conditions are connected in sequence according to time to generate the reconstruction chain segment corresponding to the fracture section.

[0060] Insert the reconstructed chain segment into the corresponding position of the continuation chain, and update the continuous positioning point sequence and continuous track.

[0061] Furthermore, based on the reconstruction results, the node correction parameters are updated, and the output of the air target navigation, tracking, and positioning results includes:

[0062] By comparing the continuous tracks before and after reconstruction, the abnormal nodes that cause missing links in the connection chain or abnormal navigation status can be identified.

[0063] Extract the direction offset, time delay offset, and frequency shift deviation from the observation primitives corresponding to the abnormal nodes;

[0064] Update the node correction parameters of the corresponding abnormal node based on the direction offset, time delay offset, and frequency shift deviation;

[0065] The updated node correction parameters are used to perform correction processing on subsequently received air target signals;

[0066] The corrected connection chain, continuous positioning point sequence, continuous track, path holding status, yaw correction status, and connection guidance status are output as the results of air target navigation, tracking, and positioning.

[0067] The technical effects and advantages of the distributed aerial target navigation, tracking and positioning method of the present invention are as follows:

[0068] This invention achieves continuous navigation, stable tracking, and reliable positioning of aerial targets in complex low-altitude scenarios by constructing a collaborative processing mechanism of distributed wireless observation, bidirectional connection determination, positioning subdomain screening, and fracture reconstruction correction. It unifies the arrival direction, time delay, and frequency shift information acquired by multiple nodes into direction survey lines, distance constraints, and velocity constraints, which can effectively suppress the impact of building obstruction, signal reflection, and local link fluctuations on positioning results.

[0069] Based on this, the continuity and consistency of a single target track are enhanced by constructing a connecting chain, screening candidate intersection areas, and verifying continuous positioning points, thereby reducing target crossover, positioning jumps, and track breaks.

[0070] Meanwhile, the introduction of path holding, yaw correction, and follow-up guidance status determination, combined with the reconstruction of broken sections by backup nodes and the updating of node correction parameters, improves the adaptability of the navigation, tracking, and positioning system to abnormal operating conditions and the stability of the output results, thereby enhancing the accuracy, robustness, and traceability of distributed airborne target navigation, tracking, and positioning. Attached Figure Description

[0071] Figure 1 This is a flowchart illustrating a distributed aerial target navigation, tracking, and positioning method according to the present invention. Detailed Implementation

[0072] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0073] In order to achieve the above objectives, Figure 1 A schematic diagram of the structure of a distributed aerial target navigation, tracking and positioning method of the present invention is given, which specifically includes the following steps;

[0074] In one implementation, this solution is applied to the navigation, tracking, and positioning of aerial targets in urban low-altitude logistics routes. Taking a cross-river delivery corridor in a certain urban area as an example, the aerial target is a logistics drone performing round-trip transportation tasks along a preset low-altitude channel. High-rise buildings, open areas on the river surface, and communication facility towers are distributed on both sides of the corridor. During flight, the drone is easily affected by building obstruction, river surface reflection, and local link fluctuations. In order to continuously perceive the aerial target, multiple distributed wireless observation nodes are deployed on the roof of warehouse buildings, the top of riverbank light poles, the roof of transfer stations, and communication towers along the route. Each distributed wireless observation node completes the installation position calibration and attitude calibration in advance during deployment and continuously receives broadcast identification signals or response signals transmitted by the aerial target to obtain information related to the aerial target's direction of arrival, propagation time, frequency offset, and corresponding acquisition time.

[0075] The system collects aerial target signals received by each distributed wireless observation node, extracts arrival direction, time delay, frequency shift, node attitude, and time stamp, and converts them into direction survey lines, range constraints, and velocity constraints under a unified spatiotemporal reference to generate an observation primitive sequence. The specific implementation is as follows:

[0076] After receiving the aerial target signal, each distributed wireless observation node first performs bandpass filtering, synchronization acquisition, and effective frame identification on the signal to separate the target signal segments that can be used for measurement. Subsequently, it performs direction detection on the target signal segments to obtain the direction of arrival of the target signal relative to the local coordinate system of the current node. Direction detection can be performed using the phase difference resolution method between array receiving channels or the peak response search method of scanning beam. Taking the aforementioned cross-river delivery corridor as an example, when the logistics drone flies from the warehouse building to the riverbank transfer point, the observation node located on the top of the warehouse building detects that the target signal comes from the southeast, while the observation node located on the top of the riverbank facility detects that the target signal comes from the northwest. Different nodes form the direction of arrival results corresponding to their own positions.

[0077] The aforementioned arrival directions are all stored in the node's local coordinate system, providing the original input for unified spatial conversion.

[0078] When extracting propagation time, each distributed wireless observation node performs propagation time determination on the received target signal to obtain the time delay of the corresponding airborne target signal, specifically including:

[0079] For broadcast signals carrying a transmission time field, the propagation time can be obtained from the difference between the reception time and the transmission time;

[0080] For target signals under the query-response system, the round-trip propagation time can be determined by the query sending time and the response receiving time, and the equivalent one-way propagation time can be obtained after deducting the preset processing delay. Taking the aforementioned logistics drone as an example, different nodes have different delays measured for the same signal segment due to their different spatial distances from the drone. Nodes that are closer to the drone have smaller delays, while nodes that are farther away have larger delays. This difference can directly reflect the spatial constraint relationship between the node and the aerial target.

[0081] When extracting frequency shift, each distributed wireless observation node performs frequency offset detection on the target signal to obtain the frequency shift of the corresponding air target signal. Frequency offset detection can be achieved by comparing the deviation between the received carrier frequency and the target's nominal transmission frequency, or by estimating the phase change rate between consecutive pilot symbols. Since there is relative motion between the air target and different observation nodes, the same air target may exhibit radial velocity trends in different directions relative to different nodes. For example, when a logistics drone flies along the river from west to east, the frequency shift result measured by the east node indicates that the target is approaching the node, while the frequency shift result measured by the west node indicates that the target is moving away from the node. In this scheme, the frequency shift result is used as the basic input for constructing velocity constraints, rather than directly as the final velocity output value.

[0082] After extracting the direction of arrival, time delay and frequency shift, the node attitude and time stamp of each distributed wireless observation node at the corresponding acquisition time are further read. The node attitude includes at least the azimuth, pitch and roll angle information of the node in the unified spatial reference system, which is used to convert the direction of arrival in the local coordinate system to the unified spatial reference system.

[0083] The time stamp includes at least the reception time of the target signal segment, used for subsequent multi-node timing alignment. Node attitude can be determined jointly by the installation calibration results and the output of the inertial measurement unit. The time stamp can be generated by a unified regional timing source. Subsequently, the arrival direction, delay, frequency shift, node attitude, and time stamp corresponding to the same target signal segment are associated and written into the same observation record entry. This observation record entry can adopt a fixed field structure, sequentially including node identifier, signal frame identifier, reception time, node position, node attitude, arrival direction, delay, frequency shift, and quality flag. The quality flag is used to record whether the signal corresponding to this entry has weak, discontinuous, or saturated signals during the measurement process, for subsequent screening.

[0084] The quality marker consists of a valid marker for direction detection, a valid marker for propagation time measurement, a valid marker for frequency offset detection, and a marker for signal continuity. When any of the following occurs: direction detection failure, propagation time measurement failure, or frequency offset detection failure, the observation record entry is marked as a low-confidence record. When the continuous length of a signal segment is lower than the preset minimum segment length or when synchronization capture fails, the observation record entry is marked as a failed record. When generating the observation primitive sequence later, only low-confidence records are retained for verification, and failed records are removed.

[0085] It should be noted that the unified spatial reference system adopts the engineering coordinate system of the corresponding city's low-altitude operation map, which includes east-west coordinate axes, north-south coordinate axes, and altitude coordinate axes. Each distributed wireless observation node records the coordinates of its fixed position during deployment and periodically reads the node attitude correction during operation. The unified time reference adopts the regional master clock. A time difference correction table is established between the node's local clock and the regional master clock. When aligning the time sequence, the time difference correction corresponding to the node identifier is read first, and then a unified conversion is performed on the received time.

[0086] After the observation record entries are formed, the arrival direction, time delay, frequency shift, node attitude, and time stamp are further converted into direction survey lines, distance constraints, and velocity constraints under a unified spatiotemporal reference. Specifically, firstly, the time stamps in each observation record entry are time-aligned according to the unified time reference, and the reception time of each node's local record is uniformly converted to the time axis corresponding to the regional master clock. If a node has a local clock drift, it is compensated through the time difference correction relationship between the node and the master clock. After time alignment, the observation record entries formed by different nodes for the same aerial target at similar times can fall into a unified time series for subsequent time correlation.

[0087] After time alignment is completed, attitude conversion is performed on the node attitudes in each observation record entry based on a unified spatial reference, and the arrival direction is converted into a direction survey line. The unified spatial reference can be the geographic coordinate system or engineering coordinate system corresponding to the regional electronic map. The installation position of each observation node is used as the starting point of the direction survey line, and the arrival direction after node attitude conversion is used as the extension direction of the direction survey line, thus forming a spatial survey line to describe the possible direction of the target. Taking the aforementioned scenario as an example, the warehouse roof node forms a direction survey line extending from the node position to the southeast and upward, and the riverbank node forms a direction survey line extending from the node position to the northwest and upward. The spatial proximity of the two direction survey lines can be used to define the target position.

[0088] When generating distance constraints for corresponding aerial targets based on time delay, the radial distance from the node to the target can be obtained by multiplying the propagation time by the electromagnetic wave propagation speed. In the case of a fixed system processing delay, the propagation time is first deducted for delay and then converted into a distance relationship. If the propagation time comes from the broadcast arrival time difference, a position line constraint corresponding to that time difference can be generated. If the propagation time comes from a one-way or equivalent one-way propagation time, a distance shell constraint with the node as a reference can be generated.

[0089] In one implementation, when the propagation time is the one-way propagation time or the processed equivalent one-way propagation time, the first... The radial distance from a distributed wireless observation node to an aerial target can be determined by the following formula: ,in, Indicates the first Radial distance from each distributed wireless observation node to an aerial target; Indicates the speed of electromagnetic wave propagation; Indicates the first The propagation time measured by a distributed wireless observation node; Indicates the first Fixed processing delay compensation amount corresponding to each distributed wireless observation node.

[0090] Based on the radial distance , with the first The installation locations of the distributed wireless observation nodes serve as reference points, generating distance constraint regions for corresponding aerial targets. For distance constraint regions formed by different nodes, they can be combined with the corresponding directional survey lines for subsequent candidate intersection region solutions.

[0091] In this scheme, the spatial constraint relationship derived from the time delay is uniformly expressed as a distance constraint. It is not limited to a single geometric form, but only requires that it can describe the reachable distance range of the target relative to the observation node. This is compatible with different signal systems and is convenient to use in conjunction with direction survey lines.

[0092] When generating velocity constraints for corresponding aerial targets based on frequency shift, the radial velocity trend and its range of change of the target relative to the current node can be determined based on the frequency shift direction and frequency shift amount. Then, combined with a unified spatial reference, velocity constraints are formed. Velocity constraints are used to limit the change boundary of the target position at adjacent times and provide continuous motion basis for subsequent continuation judgment and location area screening.

[0093] For the same aerial target, the speed constraints at adjacent moments can reflect its tendency to move along the flight corridor, decelerate and turn, or deviate laterally, thus avoiding positioning drift caused by relying solely on the orientation and distance relationship at a single moment.

[0094] After completing the above conversion, the direction survey line, distance constraint, and velocity constraint are written into the corresponding observation record entries, and node identifiers and spatiotemporal reference marks are added to form observation primitives. Each observation primitive contains at least a unified time mark, a unified spatial reference mark, a node identifier, a direction survey line description, a distance constraint description, and a velocity constraint description. Then, the observation record entries are arranged in chronological order to generate an observation primitive sequence.

[0095] Taking the cross-river logistics delivery scenario as an example, when a logistics drone takes off from a warehouse building, flies along the river and approaches the riverbank transfer point, multiple observation nodes distributed on the rooftops and lampposts on both sides of the river continuously output observation primitives. After these observation primitives are sorted in a unified time, they form an observation primitive sequence describing the drone's flight process. This observation primitive sequence provides direct input for subsequent analysis of time continuity, spatial adjacency, and velocity continuity, and can further support the construction of the continuity chain of a single aerial target.

[0096] Meanwhile, to improve the completeness of the implementation, the screening process for observation record entries can be limited. In one optional implementation, the validity of observation record entries is first judged based on quality tags. Only observation record entries that have successfully detected direction, successfully measured propagation time, successfully detected frequency offset, and whose signal continuity meets preset conditions are retained. Then, time alignment and spatial conversion are performed, which can reduce the impact of weak signals, discontinuous signals, and severely distorted signals on the construction of subsequent observation primitive sequences.

[0097] Based on the temporal continuity, spatial adjacency, and velocity continuity among the observation primitives, a two-way continuity determination is performed to generate a continuity chain corresponding to a single aerial target. Specifically, the implementation is as follows:

[0098] After the previous step, an observation element sequence is obtained, which is arranged according to a unified time reference. Each observation element contains a node identifier, a unified time marker, a direction survey line, a distance constraint, and a velocity constraint.

[0099] First, the observation primitive sequence is traversed in chronological order, and the currently processed observation primitive is taken as the predecessor primitive. For each predecessor primitive, a preset time neighborhood is set after its unified timestamp. Within this time neighborhood, candidate successor observation primitives that may be connected to it are searched. The preset time neighborhood can be determined by the system sampling period, the node reporting period, and the expected maneuverability of the air target. Its function is to limit the successor observation primitive to appear within the time interval that can form a continuous flight process. If the timestamp of the candidate successor observation primitive is earlier than that of the predecessor primitive, or although it is later than that of the predecessor primitive but exceeds the preset time neighborhood, the candidate successor observation primitive is directly excluded.

[0100] It should be noted that the preset time neighborhood consists of the forward retrieval start time, the forward retrieval end time, and the boundary tolerance interval. The forward retrieval start time is determined based on the unified time stamp of the predecessor primitive, and the forward retrieval end time is determined based on the maximum reachable flight time corresponding to the predecessor primitive. When the time stamp of a candidate successor observation primitive falls within the boundary tolerance interval, it is marked as a boundary candidate primitive, and its bidirectional continuity is verified first in the subsequent reverse backtracking stage.

[0101] After completing the time retrieval, the target position projection range is determined based on the direction survey line, distance constraint, and velocity constraint of the precursor primitive;

[0102] Specifically, based on the directional survey lines and distance constraints of the precursor primitives, the feasible target location region corresponding to the precursor moment is generated under a unified spatial reference.

[0103] Combining the velocity constraints of the precursor unit and the time interval between the precursor unit and the candidate successor observation unit, the feasible position region is projected forward along the allowed motion direction to obtain the target position projection range near the candidate successor time. This target position projection range can be represented as a spatial strip-shaped region extending forward along the precursor direction survey line, or it can be represented as a spatial envelope region formed by extrapolating the feasible position region of the precursor over time. Its boundary is jointly defined by the precursor velocity constraints and the time interval. The target position projection range is jointly determined by the feasible position region of the precursor, the time interval between the precursor time and the candidate successor time, and the displacement boundary defined by the velocity constraints. The reachable range of motion is jointly constituted by the maximum forward displacement boundary, the lateral offset boundary, and the altitude change boundary of the target position confirmed in the previous time along the track extension direction. When the direction survey line, distance constraint, and velocity constraint corresponding to the candidate successor observation unit all fall within the above spatial projection boundary, it is determined that the position continuity requirement is met.

[0104] Taking the aforementioned cross-river delivery drone as an example, if the precursor element indicates that the drone is flying eastward along the river, the target position projection range will expand from the feasible position area at the precursor moment along the eastward flight channel, rather than expanding to irrelevant areas in the opposite direction of the flight channel or away from the corridor, thus ensuring that the succession determination has a clear spatial orientation.

[0105] For each candidate successor observation primitive, determine the intersection relationship between its orientation survey line and the target position projection range, and further determine the matching results between its range constraint and velocity constraint and the target position projection range, specifically including:

[0106] When there is an intersection, contact, or overlap between the directional survey line and the target position projection range, it indicates that the candidate successor observation element can point to the possible target area extrapolated from the predecessor element in the spatial direction.

[0107] If the direction survey line is completely separated from the target position projection range, it indicates that the candidate successor observation element and the predecessor element are spatially discontinuous. Then, the overlap between the distance constraint of the candidate successor observation element and the target position projection range is determined. When the distance constraint corresponding to the candidate successor observation element covers the part of the target position projection range, it indicates that the candidate successor observation element and the predecessor element are compatible in terms of spatial distance. Furthermore, the velocity constraint of the candidate successor observation element is compared with the velocity change boundary derived from the predecessor element. If the radial velocity trend, velocity change direction, and allowable change amplitude corresponding to the candidate successor observation element all fall within the velocity continuation interval formed by extrapolation of the predecessor velocity constraint, the velocity continuation determination result is satisfied; otherwise, it is determined that it is not satisfied.

[0108] In summary, the spatial adjacency determination result is derived from the direction and distance matching relationship between the candidate successor observation primitive and the target position projection range, while the velocity continuity determination result is derived from the continuity matching relationship between the candidate successor observation primitive and the preceding motion trend.

[0109] In a specific example, the node located on the roof of the warehouse building outputs a predecessor primitive at time one, indicating that the target is located in the low-altitude corridor in the direction of the riverbank and has a velocity trend of eastward along the river. In the subsequent preset time neighborhood, the node located on the top of the lamppost on the riverbank and the node on the top of the commercial building along the line output two candidate successor observation primitives respectively. According to the target position projection range obtained by extrapolating the predecessor primitive, it is found that the directional survey line corresponding to the node on the top of the lamppost on the riverbank passes through the projection range, and its distance constraint overlaps with the projection range. The velocity trend derived by the frequency shift is consistent with the velocity continuation interval of the predecessor primitive. Therefore, this candidate successor observation primitive satisfies the spatial adjacency and velocity continuation conditions. In contrast, although the node on the top of the commercial building is in the preset time neighborhood in time, its directional survey line does not pass through the target position projection range, and its distance constraint is also separated from the range. Therefore, its spatial adjacency judgment result does not meet the requirements. Based on this, the former is retained and the latter is eliminated, thereby avoiding the misconnection of irrelevant observations to the same flight trajectory.

[0110] After generating the time succession determination result, spatial adjacency determination result, and velocity continuation determination result, a continuation determination result is generated according to a preset continuation rule table. The continuation rule table is set in a deterministic logic manner. For example, it is stipulated that when the time succession determination result, the spatial adjacency determination result, and the velocity continuation determination result are satisfied, the continuation determination result is output as valid.

[0111] When the time continuity judgment result is satisfied, the spatial adjacency judgment result is satisfied, and the velocity continuity judgment result is in the boundary state, the state to be verified can be output and handed over to the subsequent bidirectional continuity check for further confirmation.

[0112] When the spatial adjacency determination result is not satisfied, regardless of the other determination results, the continuation determination result is directly output as not valid.

[0113] After obtaining the succession determination result, a two-way succession determination is further performed to form a succession chain corresponding to a single aerial target, specifically including:

[0114] First, establish a forward succession relationship between the predecessor primitive and the candidate successor observation primitive based on the succession determination result. Each predecessor primitive and its candidate successor observation primitive that meet the succession conditions can be registered in the forward succession relationship table to form a directed connection from the predecessor primitive to the candidate successor observation primitive.

[0115] Subsequently, using candidate successor observation primitives as backtracking primitives, a reverse search is performed in their preceding time neighborhood to find preceding observation primitives that can form a succession relationship with them, and a reverse succession relationship is established between the candidate successor observation primitive and the preceding primitive. If a preceding primitive and a candidate successor observation primitive have both a forward succession relationship and a reverse succession relationship, then it is determined that there is a bidirectional succession relationship between them. The significance of the bidirectional succession relationship is that the successor observation primitive is reasonable not only from the perspective of forward extrapolation, but also from the perspective of backward backtracking, thereby significantly reducing the probability of false connections caused by local false detection, reflected signals, or short-term noise.

[0116] Taking the aforementioned cross-river logistics drone scenario as an example, the observation primitives at time one (the warehouse rooftop node) and time two (the riverbank light pole node) have established a forward continuity relationship. At this point, using the riverbank light pole node as a backtracking primitive, we search for observation primitives that can serve as a previous source within its preceding time neighborhood. If the backtracking results show that the warehouse rooftop node can reversely explain the riverbank light pole node in terms of spatial position, directional continuity, and velocity changes, then a reverse continuity relationship is established between the two, confirming a bidirectional continuity relationship between the two observation primitives. If another preceding observation primitive, although connected to the riverbank light pole node in the forward determination, has inconsistent directional or distance constraints with the riverbank light pole node in the reverse backtracking, then this pair of primitives will not be confirmed as a bidirectional continuity relationship. After this processing, observation primitives that truly belong to the continuous observation process of the same flight target will be retained, while irrelevant observation primitives that happen to approach each other will be excluded.

[0117] For observation primitives that have formed a bidirectional connection, they are connected in a unified time order to generate candidate connection chains. In specific implementation, each observation primitive can be used as a chain node and the bidirectional connection relationship can be used as a chain edge to construct a connection graph structure. Then, the observation primitive with the earliest time and empty in-degree is used as the chain head node, and the connection is gradually expanded along the bidirectional connection relationship to generate candidate connection chains arranged in time progression.

[0118] If there is more than one bidirectional succession path within the same time window, it indicates that multiple air targets may be flying simultaneously, or that a certain air target has formed multiple competing succession paths in a local area. In this case, conflicting observation primitives in the candidate succession chain are split and eliminated. Conflicting observation primitives refer to observation primitives that simultaneously satisfy bidirectional succession relationships with two different predecessor primitives, or are simultaneously occupied by two different successor primitives. For such observation primitives, their temporal continuity, spatial continuity, and velocity consistency with adjacent chain segments are further compared. Observation primitives that can maintain the continuity of the entire chain are retained in the current candidate succession chain, while observation primitives that disrupt continuity are eliminated from the current candidate succession chain. If necessary, independent chain segments are formed. After splitting and elimination, the succession chain corresponding to a single air target is finally obtained.

[0119] In a more specific scenario, when two logistics drones pass through the same cross-river corridor one after the other, nodes near the riverbank transfer point may observe two sets of observation primitives that are close in direction and distance within a similar time period. If they are connected based solely on a one-way continuity relationship, it is easy to merge the local observation errors of the two drones into the same track. After adopting the above-mentioned two-way continuity and conflict splitting method, it is found that the forward continuity relationship formed by the first drone can only form a stable reverse continuity relationship with its own preceding observation primitives when backtracking. The local observation primitives of the second drone cannot simultaneously satisfy the two-way continuity conditions of the first drone's chain segment. Therefore, this part of the primitives will be split off to form a second continuity chain.

[0120] To further improve the completeness of the implementation, an intermediate result record structure can be set in the process of succession determination and bidirectional succession establishment. The time succession determination result, spatial adjacency determination result, velocity continuation determination result and final succession determination result can be written into the determination result entries respectively. The determination result entries include the predecessor primitive identifier, the candidate successor primitive identifier, the time determination mark, the spatial determination mark, the velocity determination mark and the succession conclusion mark.

[0121] After bidirectional connection is established, forward connection, reverse connection and bidirectional connection can be written into the connection relationship table for subsequent candidate connection chain generation, conflict observation primitive splitting and single air target connection chain output.

[0122] The intersection solution is performed on the direction survey lines and distance constraints in the docking chain. Combined with velocity constraints and spatial boundaries, the target positioning subdomain is determined, and a continuous sequence of positioning points is generated based on the target positioning subdomain. The specific implementation is as follows:

[0123] The actual spatial position of an aerial target is determined by using observations at consecutive moments in a continuous chain, and a continuous sequence of positioning points that can stably characterize the flight trajectory is generated.

[0124] In practice, the direction survey lines and distance constraints at the same or adjacent times are first extracted from the continuation chain. The same time refers to multiple observation primitives that fall into the same sampling period after being aligned with a unified time reference. Adjacent times refer to observation primitives in the current sampling period and the sampling periods before and after that that can jointly reflect the continuous change of the target position.

[0125] It should be noted that the reason why the direction measurement lines and distance constraints at the same or adjacent times are allowed is that in a real deployment environment, although the signal reception times of different observation nodes have been unified to the same time axis, there may still be slight misalignments due to differences in the reporting beats and signal triggering methods of the nodes.

[0126] By extracting effective observations from adjacent time points within the continuum chain, the stability of the intersection solution can be improved without disrupting temporal continuity. The candidate intersection region is described by a closed boundary, which is represented by a sequence of intersection boundary points or a sequence of intersection envelope vertices.

[0127] When the direction survey lines and distance constraints of multiple nodes work together, first determine the single-node candidate intersection region corresponding to each node, and then perform common overlap solution on multiple single-node candidate intersection regions to form a multi-node candidate intersection region;

[0128] The target localization subdomain is represented by a single candidate intersection region retained after filtering by spatial boundaries and motion reachability;

[0129] Taking the aforementioned cross-river delivery scenario as an example, when the logistics drone flies from the warehouse building to the riverbank transfer point, the warehouse rooftop node obtains a direction measurement line and a set of distance constraints during the current sampling period, while the riverbank node obtains another direction measurement line and another set of distance constraints during the same sampling period or the next adjacent sampling period. These two sets of observations can then be used together to solve for the current location.

[0130] After obtaining the direction survey lines and distance constraints, spatial intersection calculations are performed on the direction survey lines and distance constraints to generate candidate intersection regions, specifically including:

[0131] Each directional survey line can be regarded as a spatial constraint axis extending from the installation position of the corresponding node along the target arrival direction. The corresponding distance constraint is then represented as a spatial distance restriction area surrounding the node reference point. When the directional survey line passes through a certain distance constraint area, the spatial range formed by the intersection, contact, or crossing can be regarded as a candidate intersection area. When there are directional survey lines and distance constraints of multiple nodes at the same time, the candidate intersection results formed by each node can be superimposed to find the intersection, so as to narrow down the possible spatial range of the target. The candidate intersection area here is not limited to a single geometric point, but can be a volume or surface region with spatial boundaries, as long as the region meets the conditions jointly defined by the corresponding directional survey line and distance constraint. This processing method can be compatible with direction detection error, propagation time measurement error, and node attitude conversion error, and avoids excessively shrinking the positioning result to a single point in the early stage, thus amplifying the local measurement error.

[0132] To ensure that the convergence results match the actual flight process, the target's reachable range at the corresponding moment is further determined based on the velocity constraints. Specifically, the target position confirmed at the previous moment or the feasible position area in the previous stage of the continuation chain is used as the starting reference. Combined with the time interval between the current moment and the velocity constraints corresponding to the current continuation chain, the spatial range that the target can reach within the time interval is generated. The reachable range reflects the possible range of the target's movement along the main flight direction and also limits its change boundaries in the lateral and altitude directions.

[0133] For example, in the cross-river delivery corridor scenario, logistics drones usually move steadily along the preset flight path, and their reachable range mainly extends along the direction of the corridor, without deviating to the area behind the high-rise buildings far away from the delivery path in a short period of time. This allows the reachable range generated by speed constraints to verify the motion continuity of candidate intersection areas, thereby eliminating false location areas that meet the direction measurement line and distance constraints at a single moment but are difficult to actually reach during the flight process.

[0134] After the candidate intersection area and the reachable range of movement are determined, the overlap relationship between the candidate intersection area and the reachable range of movement is determined, and the candidate intersection area located outside the airspace boundary is deleted. The airspace boundary can be composed of the permitted flight boundary of the low-altitude delivery corridor, the projection boundary of the no-fly building, the safety boundary of the river channel, and the preset height restriction boundary. If a candidate intersection area is compatible with the direction survey line and distance constraints, but falls completely outside the flight corridor or inside the building restricted area, then the candidate intersection area is directly deleted.

[0135] If a candidate intersection region does not overlap with the reachable range, it means that the region cannot be actually reached from the target position at the previous moment within the current time interval, and it is also deleted. After this processing, the remaining candidate intersection regions simultaneously satisfy the conditions of direction observation, distance observation, velocity continuity and spatial constraints, and have higher reliability.

[0136] In some scenarios, after the above screening, more than two candidate intersection areas may still be retained. The reasons for this usually include the intersection of multiple effective directional survey lines caused by reflection from high-rise buildings, the overlap of local distance constraints caused by reflection from the river surface, and the mirror intersection results formed when multiple nodes are deployed in approximately symmetrical positions. In this case, it is necessary to determine the target positioning subdomain based on the displacement continuity relationship and direction continuity relationship between each candidate intersection area and the target position at the previous moment. The displacement continuity relationship means that the displacement distance and displacement direction of the candidate intersection area relative to the target position at the previous moment should match the known flight rhythm.

[0137] The direction of continuity means that the direction of travel inferred from the direction of the track at the previous moment should be consistent with the direction of the candidate intersection area;

[0138] Taking the aforementioned cross-river delivery corridor as an example, if the target location was in the middle of the river and the track direction was pointing to the transfer point on the east bank at the previous moment, and there are two candidate intersection areas on the north side and the south side of the river at the current moment, the system will prioritize retaining the candidate intersection area extending along the existing track direction as the target positioning subdomain, while excluding the other candidate intersection area that is significantly deviated from the existing track direction. This makes the determination of the target positioning subdomain no longer dependent on single-moment observation, but based on the continuity of the trajectory at multiple moments.

[0139] After determining the target positioning subdomain, a continuous positioning point sequence is further generated based on the target positioning subdomain. Specifically, the center position point of the target positioning subdomain at each time moment is first determined as a candidate positioning point.

[0140] The center location point is taken as the geometric center, envelope center, or centroid of the target positioning subdomain. As long as the location can stably represent the spatial center trend of the target positioning subdomain, it is acceptable. In actual deployment, in order to avoid the offset caused by irregular intersection areas, the candidate positioning point can be generated by taking the envelope center of the boundary points of the target positioning subdomain under a unified spatial coordinate system.

[0141] Taking the aforementioned drone flying over the vicinity of the riverbank transfer point as an example, if the target positioning subdomain is an elliptical spatial region that is slightly elongated along the flight channel, then its center point can be directly used as a candidate positioning point for the aerial target at that moment.

[0142] After obtaining candidate positioning points at adjacent time points, a continuity check is performed on the candidate positioning points based on the displacement and direction change relationships between them. The purpose of the continuity check is to prevent a jump in the continuous positioning point sequence if a valid target positioning subdomain is formed at an individual time but its center position is incompatible with the flight trajectory at the previous and subsequent time points. In specific implementation, the displacement vector, heading change, and velocity change trend between the candidate positioning point at the current time and the confirmed positioning point at the previous time can be compared. When the direction of the displacement vector is consistent with the extension direction of the existing flight path, the displacement length falls within the velocity constraint allowable range, and the heading change falls within the preset turning boundary, the candidate positioning point at the current time is determined to meet the continuity requirements.

[0143] If a candidate location point is located at the center of the target location subdomain, but its direction suddenly reverses relative to the previously confirmed location point, or its displacement length far exceeds the range of change allowed by the current velocity constraint, then the candidate location point will not pass the continuity check.

[0144] When there are multiple candidate positioning points at the current time, the candidate positioning point that satisfies the displacement continuation condition and the direction continuation condition with the confirmed positioning point at the previous time is selected as the positioning point at the current time. The situation of generating multiple candidate positioning points usually occurs when there are multiple target positioning subdomains that can be retained at the current time, or when the current target positioning subdomain forms multiple central candidate positions in a local area due to the intersection of multiple segments.

[0145] For example, in a densely built-up area along the riverbank, the reflection from a building facade causes two local clusters of directional survey lines to intersect. The system can obtain two candidate positioning points. If the displacement direction between one of the candidate positioning points and the previously confirmed positioning point continues the previous track direction, and the displacement length meets the velocity constraint, then the candidate positioning point is determined as the positioning point at the current moment.

[0146] Another candidate location point, which is geometrically feasible but does not conform to the direction of the existing flight trajectory, is not adopted. In this way, it is clear that the selection of the location point depends not only on the current spatial intersection result, but also on the constraint of the continuity of the trajectory at the previous moment, thereby ensuring that the entire trajectory is physically feasible.

[0147] After the location points at each time point are determined, they are arranged in chronological order to generate a continuous location point sequence. The continuous location point sequence is the direct input for the generation of subsequent continuous flight paths. Each element corresponds to the spatial location point of a single aerial target under a unified time mark. The sequence can store the location point identifier, time mark, spatial coordinates, and the associated target location subdomain identifier in an ascending order of time. Taking the aforementioned flight process of a logistics drone departing from the warehouse, flying over the river, and arriving at the riverbank transfer point as an example, the generated continuous location point sequence should be smoothly distributed along the preset delivery corridor, reflecting the stable flight process of the drone from the west bank to the east bank, without irregular jumps on the north and south sides of the river or the back of high-rise buildings.

[0148] In actual operation, at certain times, the candidate positioning point may still shift abnormally due to instantaneous occlusion of the observation node, enhanced reflection, or local electromagnetic interference. To improve the stability of the positioning sequence, when the displacement change between positioning points at adjacent times exceeds the preset jump boundary, the positioning point selection is re-executed according to the target positioning subdomain at the corresponding time, and the continuous positioning point sequence is updated. The preset jump boundary can be determined by the current speed constraint, the upper limit of the aircraft maneuver, and the preset route geometric constraint.

[0149] The continuity check includes at least displacement length check, displacement direction check, and heading change check. When the displacement length between the current candidate positioning point and the confirmed positioning point in the previous time exceeds the displacement boundary, or the displacement direction deviates from the existing track extension direction by more than the direction boundary, or the heading change exceeds the turning boundary, the current candidate positioning point is marked as a jump candidate point. When all candidate positioning points in the current time are marked as jump candidate points, the re-selection of the target positioning subdomain in the current time is triggered.

[0150] In one implementation, the displacement change between adjacent positioning points can be determined by the following formula: The corresponding preset transition boundary can be determined by the following formula: When the following formula is satisfied, the first is determined to be... Timing location triggers jump reselection: ,in, Indicates the first Time location point and the The change in displacement between the positioning points at any given time; Indicates the first The position vector of the location point at any given moment in a unified spatial coordinate system; Indicates the first The position vector of the location point at any given moment in a unified spatial coordinate system; Indicates the first The preset transition boundary corresponding to each moment; Indicates the first The upper limit of the allowable speed corresponding to the speed constraint at any given moment; Indicates the first Time and the The time interval between moments; This represents the position compensation boundary determined by the preset route geometric constraints and maneuver tolerances.

[0151] Based on the above determination results, when the first When a time-based location point triggers a jump and reselection, the target location subdomain at the corresponding time is returned, the displacement and direction continuity relationships of the candidate location points are re-verified, and the continuous location point sequence is updated.

[0152] If, at a certain moment, the displacement of a positioning point relative to the positioning point at the previous moment suddenly exceeds the boundary, it indicates that although the positioning point originates from the center of the current target positioning subdomain, it may be affected by local erroneous intersection results. In this case, return to the target positioning subdomain at that moment, re-check the displacement and direction continuation of all candidate positioning points, and reselect a positioning point that better meets the continuity conditions to replace the original positioning point.

[0153] For example, when a drone approaches a transfer station, a candidate location point at a certain node may shift to the back side of the building due to the reflection of the building's mirror surface. Once the displacement change of the location point relative to the location point at the previous moment exceeds the preset jump boundary, a reselection process is triggered, and another candidate location point that is closer to the direction of the corridor extension is written back into the continuous location point sequence, thereby restoring the continuity of the trajectory.

[0154] It is also possible to set up data structures to record the generation process of the target positioning subdomain and continuous positioning points, specifically including:

[0155] The target location subdomain record entry may include at least a time stamp, subdomain boundary description, source direction survey line identifier, source distance constraint identifier, velocity constraint identifier, and airspace boundary verification identifier;

[0156] The candidate location point record entry may include at least the candidate point coordinates, the corresponding target location subdomain identifier, the displacement continuation determination identifier, and the direction continuation determination identifier;

[0157] A continuous positioning point sequence can record the coordinates of confirmed positioning points, the corresponding time, the corresponding target positioning subdomain, and the jump reselection marker in chronological order.

[0158] Connecting consecutive positioning point sequences forms a continuous track, and based on the relationship between the continuous track and preset route constraints, path-keeping state, yaw correction state, and follow-up guidance state are generated. Specifically, the implementation is as follows:

[0159] First, connect the positioning points of adjacent moments in the continuous positioning point sequence in chronological order to generate a continuous track. The continuous track can be represented as a directed trajectory structure in a spatial coordinate system consisting of multiple adjacent track segments connected end to end. Each track segment corresponds at least to the spatial connection between the positioning point at the previous moment and the positioning point at the current moment and its corresponding time interval.

[0160] To prevent local positioning errors from being directly amplified into jagged tracks, during the connection process, adjacent positioning points with too small time intervals but insignificant spatial changes can be merged. Alternatively, adjacent track segments with normal time intervals but significant local directional jitter can be smoothed. This makes the resulting continuous track more consistent with the physical flight process of the aerial target. Taking a cross-river delivery drone as an example, when the drone takes off from the west bank warehouse and flies along the central axis of the river to the east bank transfer point, the generated continuous track should extend smoothly along the established delivery corridor as a whole, without any obvious reversals or lateral sharp angles that are inconsistent with the flight path at certain moments.

[0161] After the continuous track is formed, the heading change results, displacement change results, and turning change results of adjacent track segments in the continuous track are further extracted. The heading change results are used to characterize the change in the extension direction of the current track segment relative to the previous track segment.

[0162] Displacement change results are used to characterize the magnitude of the change in the target's position in space per unit time;

[0163] The turning change results are used to characterize the turning trend and turning intensity between adjacent track segments.

[0164] In one feasible approach, the heading vector of the current track segment can be determined first based on the spatial coordinates of adjacent positioning points. Then, the heading change, displacement change, and turning change results can be obtained through the changes in the angle, length, and projection between the heading vectors. For low-altitude logistics routes, if the UAV is in a stable cruise phase across a river, the heading change results between adjacent track segments should be kept within a small range, the displacement change results should match the cruise speed, and the turning change results should be close to zero.

[0165] If the UAV approaches the East Coast transfer station and prepares to enter the landing guidance segment, the changes in heading and turning between adjacent track segments will gradually increase, reflecting that it is performing a predetermined turning entry maneuver.

[0166] Preset route constraints are used to limit the flight channels, directional trends and turning rules that air targets are allowed to follow in the current mission scenario. Preset route constraints can be composed of the delivery corridor centerline, corridor boundary line, altitude layer boundary, approach turning window, mission target point direction and restricted area boundary. That is, when the corresponding track segment of a continuous track is located within the preset route boundary and the heading change result is consistent with the extension direction of the corridor centerline, a path holding state is generated.

[0167] When the corresponding track segment of a continuous track is still within the allowable flight boundary, but its lateral deviation continues to increase or the course change result deviates from the preset direction, a yaw correction state is generated.

[0168] When the extension direction of the current flight path segment meets the continuity relationship with the preset turning window or the target arrival area, a continuity guidance state is generated.

[0169] When the path holding state, yaw correction state, and follow-up guidance state do not meet the transition rules between the preceding and following states within a continuous sampling window, a navigation state anomaly flag is generated.

[0170] In cross-river delivery scenarios, preset route constraints can be specifically defined as:

[0171] After taking off from the west bank warehouse, the aerial target should fly eastward along the pre-set low-altitude corridor over the river. Before approaching the riverbank transfer point, it should remain within the main corridor area and complete the deflection into the top docking platform of the transfer station within the pre-set turning window. The continuous flight path should be compared with the above-mentioned pre-set route constraints to determine whether the current continuous flight path remains within the predetermined corridor, whether there is any lateral deviation, whether the turn is premature or delayed, and whether the subsequent flight path extension direction still points to the legitimate mission target area.

[0172] When determining the deviation relationship between a continuous flight path and preset route constraints, analysis can be performed from three aspects: lateral deviation, longitudinal extension deviation, and turning timing deviation, including:

[0173] Lateral deviation is used to describe the change in lateral distance of the current continuous track relative to the centerline of the route;

[0174] Longitudinal extension deviation is used to describe whether the current continuous track's direction of travel still extends along the preset corridor main axis;

[0175] The turning timing deviation is used to describe whether the current turning action occurs within the predetermined turning window. If the current continuous track is always within the preset route boundary, the heading change result is consistent with the corridor direction, and there is no lateral offset beyond the preset boundary, then the current state is determined to be the path holding state.

[0176] If the current continuous flight path has deviated from the preset route centerline, or if the heading change results show that it is deviating from the predetermined direction but has not yet left the permitted flight boundary, then the current state is determined to be a yaw correction state. Taking the aforementioned scenario as an example, when the UAV is affected by local crosswinds and deviates to the south of the river surface during the crossing process, but is still within the channel boundary, its state can be determined to be a yaw correction state based on the deviation relationship between the continuous flight path and the corridor centerline, rather than being directly regarded as lost or disengaged from the mission.

[0177] After generating the path-keeping state and yaw correction state, the follow-up guidance state is further generated based on the extension direction of the current track segment, the subsequent reachable range, and the pointing relationship of the preset route constraints. Specifically, this includes:

[0178] The subsequent reachable range refers to the spatial area that an aerial target may reach within several subsequent sampling periods under the constraints of the current flight path direction and current motion trend. This subsequent reachable range is matched with the target channel, turning window and target arrival area in the preset flight path constraints to determine whether the aerial target should continue to extend its original course, perform fine-tuning and regression, or enter a specific turning access channel. Taking the cross-river delivery scenario as an example, when the drone is located in the middle of the river and its continuous flight path is consistent with the center line of the corridor, the continuous guidance state can be characterized as continuing to move forward along the current corridor main axis.

[0179] When the drone approaches the East Coast transfer station, and the current flight path extends towards the access window above the transfer station, and the subsequent reachable range covers the access window, the follow-up guidance status can be characterized as entering towards the target docking platform.

[0180] If the UAV's flight path deviates to the south side of the corridor but the subsequent reachable range still overlaps with the main corridor, the follow-up guidance status can be characterized as a return to the northern boundary. Thus, the follow-up guidance status not only reflects whether there is a deviation at present, but also clearly indicates the follow-up relationship between the current flight path and the subsequent legal route.

[0181] The path-keeping status, yaw correction status, and follow-through guidance status can be written into corresponding consecutive track entries, making the navigation status retrievable. Specifically, this includes:

[0182] Each continuous track entry may include at least the starting position identifier, the ending position identifier, the heading change result, the displacement change result, the turning change result, the path holding status identifier, the yaw correction status identifier, and the follow-up guidance status identifier. For a track segment corresponding to a certain moment, if it meets the path holding condition, the path holding status identifier is set to valid and the yaw correction status identifier is set to invalid.

[0183] If the yaw correction conditions are met, the yaw correction status flag is set to valid, and the recommended return direction is recorded in the follow-up guidance status.

[0184] If it is in the import stage before entering the window, the subsequent route segment identifier or target area identifier that should be connected is recorded in the continuation guidance state. Through this itemized writing method, the continuous track is no longer just the result of connecting position points, but becomes a composite data object that simultaneously contains track structure information and navigation status information, providing a direct basis for subsequent judgment of missing segments in the continuation chain, identification of navigation status anomalies, and execution of broken segment reconstruction.

[0185] In a more specific operational example, when the logistics drone takes off from the west bank warehouse and flies along the central axis of the river, it generates a continuous track based on the continuous positioning point sequence, and detects that the heading changes of each track segment are consistent with the preset route centerline, and the lateral deviation is kept within the allowable range. Therefore, multiple entries corresponding to the continuous track are continuously marked as path-keeping status.

[0186] As the drone approached the east bank, due to localized gusts of wind between the high-rise buildings on the riverbank, its continuous flight path began to deviate towards the north side of the corridor. It was found that the deviation between the flight path and the centerline increased but still did not exceed the corridor boundary. Therefore, several subsequent flight path segments were marked as yaw correction states, and the guidance direction towards the centerline of the Tropic of Cancer South Corridor was written into the subsequent guidance states.

[0187] After the drone continues to move forward and enters the turning access window above the transfer station, the connection guidance status is updated to enter the transfer platform based on the consistency between the current trajectory extension direction and the target platform direction. Through the above process, a clear, coherent and executable logical chain is formed between the continuous trajectory, deviation relationship and guidance status.

[0188] It should be noted that constraint records are added to the process of generating deviation relationships and guidance states. Deviation relationship records can include at least the track segment identifier, centerline deviation, boundary proximity, steering window matching result, and deviation determination result. Continuing guidance state records can include at least the current track segment identifier, subsequent reachable range description, target route segment identifier, continuing direction identifier, and state generation time. Through this data organization method, the input objects, analysis basis, determination logic, and output results in the continuous track generation, deviation relationship analysis, and navigation state writing processes all have a clear structure.

[0189] When a link is missing or the navigation status is abnormal, the observation primitives of the backup node are invoked to reconstruct the broken section, and the node correction parameters are updated based on the reconstruction results. The navigation, tracking, and positioning results of the airborne target are then output. The specific implementation is as follows:

[0190] This step involves using the observation primitives of backup nodes that are not involved in the construction of the current continuation chain to fill in the missing segments when there are missing segments in the continuation chain or discontinuous changes in the navigation state. After the completion of the completion, the node that caused the break or state mismatch is identified and the correction parameters of the node are updated to improve the stability of continuous tracking and continuous positioning.

[0191] In practice, the continuity of the connecting chain is checked first, and the timing consistency of the path holding status, yaw correction status and connecting guidance status associated with the continuous track is checked. A missing segment of the connecting chain means that the time interval between two adjacent confirmed chain nodes in the connecting chain exceeds the preset continuous time interval, and there is no confirmed connecting observation primitive within the time interval.

[0192] Navigation state anomaly refers to a switch between path-keeping state, yaw correction state, and follow-up guidance state that does not conform to the trajectory evolution law within adjacent sampling intervals. For example, in a cross-river delivery scenario, when a drone is flying stably along the main axis of the corridor, its state should be continuously maintained as path-keeping state.

[0193] If the system successively provides path-holding status, follow-up guidance status, and yaw correction status within a very short period of time and then returns to path-holding status, and the corresponding continuous track does not show any obvious turning or lateral deviation, it can be determined that there is an abnormal switching of navigation status. For example, if there is a clear gap between the time mark of the previous chain node and the time mark of the next chain node in the follow-up chain, and the continuous positioning point sequence within the gap is interrupted, it can be determined that the time range corresponds to a broken segment. Thus, reconstruction triggering conditions can be identified at both the track level and the chain level.

[0194] After determining the fracture section, the backup node observation elements that did not participate in the construction of the current connection chain are further searched within the time range corresponding to the fracture section. The backup node refers to the distributed wireless observation node that was not selected into the current connection chain during the previous connection determination process, but actually completed signal reception and generated observation elements within the corresponding time range. Since the current connection chain usually prioritizes the observation elements with high temporal continuity and high spatial matching degree when it is constructed, there may still be available observation elements from adjacent rooftops, riverside towers or peripheral nodes of relay stations that have not been adopted in the same time period.

[0195] Using the start and end times of the fractured section as the search boundary, all candidate primitives within that time range are extracted from the backup observation primitive buffer area and reordered according to a unified time benchmark to form a backup observation primitive set corresponding to the fractured section. Taking the aforementioned scenario as an example, when the drone passes near the high-rise buildings along the riverbank, the original primary node may experience an interruption in observation due to building obstruction, but the nodes located on the light poles or communication towers outside the relay station may still intermittently receive the target signal. These observation primitives that are not used by the current connection chain constitute backup inputs that can be reconstructed.

[0196] To ensure continuity between the reconstructed chain segment and the flight path before and after the fracture segment, the reconstruction search range is determined based on the front and rear positioning points of the fracture segment. Specifically, the front positioning point of the fracture segment is used as the starting reference point for reconstruction, and the rear positioning point of the fracture segment is used as the ending reference point for reconstruction. Combining the extension direction of the flight path segment corresponding to the front positioning point, the backtracking direction of the flight path segment corresponding to the rear positioning point, and the duration of the fracture, a spatial search envelope connecting the front and rear is generated. This search envelope limits the spatial range that the target may pass through within the fracture segment, while also considering the flight corridor boundary, mission channel direction, and historical speed continuity trend. For cross-river logistics drone scenarios, this reconstruction search range is usually represented as a narrow spatial band extending along the preset delivery corridor, rather than extending to the area behind tall buildings or restricted airspace that is significantly deviated from the flight path. After setting this, even if there are many backup node observation primitives, subsequent continuation judgments are only made within the spatial range consistent with the flight logic of the fracture segment, thereby reducing the possibility of mistakenly accessing irrelevant observations.

[0197] After obtaining the reconstruction search range, a succession determination is performed on the observation primitives of the backup nodes, specifically including:

[0198] The succession determination is consistent with the determination logic in the previous main succession chain construction stage. It is still based on the temporal succession relationship, spatial adjacency relationship and velocity continuity relationship. However, its determination reference object is expanded from the original main chain predecessor primitive to the front positioning point of the fracture section, the rear positioning point of the fracture section and its corresponding track extension boundary. Specifically, it first determines whether the time mark of the backup observation primitive falls into the valid search window within the fracture section, then determines whether the direction survey line and distance constraint corresponding to the backup observation primitive cross the reconstruction search range, and finally determines whether the velocity constraint corresponding to the backup observation primitive is compatible with the motion trend of the track before and after the fracture section.

[0199] Only when all three judgment results are met simultaneously is the backup observation element considered a valid link node that can be used for reconstruction. For example, when a drone approaches the east bank from the middle of the river, the original connection chain is interrupted for a certain period of time due to the obstruction of tall buildings on the riverbank. However, the nodes at the top of the light poles outside the transfer station and the nodes at the top of the communication towers along the line generate several backup observation elements during that period of time. After the connection judgment, it is found that the directional survey lines of these backup observation elements all pass through the spatial search zone connecting the positioning point at the front end of the fracture and the positioning point at the rear end of the fracture. Their distance constraints and velocity constraints are also compatible with the flight trends of the preceding and following tracks, so they are selected for the fracture reconstruction process.

[0200] After selecting backup node observation primitives that meet the continuation conditions, they are concatenated in chronological order to generate the reconstruction chain segment corresponding to the fracture section. The reconstruction chain segment can be organized using the same data structure as the original continuation chain, that is, the backup observation primitives are used as chain nodes, and the continuation relationship established through reconstruction judgment is used as chain edges to form a directed temporal chain within the fracture section. If there are multiple possible reconstruction paths within the fracture section, the system further compares the degree of trajectory connection between each path and the positioning point at the front end of the fracture and the positioning point at the rear end of the fracture, and prioritizes retaining the path with continuous displacement, continuous direction, and smooth turning as the final reconstruction chain segment.

[0201] Subsequently, the reconstructed chain segment is inserted into the corresponding position of the continuation chain to form an updated complete continuation chain. After the continuation chain is updated, the target positioning subdomain solution and positioning point selection within the broken segment are re-executed to generate a complete sequence of continuous positioning points. These are then reconnected to form an updated continuous track. In this way, the missing spatial positions and track segments within the broken segment are restored, and subsequent navigation status can be re-determined based on the updated continuous track.

[0202] In a more specific operational example, a logistics drone takes off from the west bank warehouse and flies eastward along the cross-river corridor. When it enters the vicinity of the high-rise buildings on the east bank, the main riverbank rooftop node loses stable observation due to short-term obstruction, causing the connection chain to be interrupted within several sampling periods. At the same time, the transfer station perimeter light pole node and communication tower node, which were not originally used by the main chain, continue to receive the target signal. After detecting the existence of a time discontinuity in the connection chain, the observation primitives generated by these two backup nodes within the time range of the break are extracted, and a reconstruction search range extending along the corridor is constructed based on the location points at the front and rear ends of the break. After judgment, the three observation primitives of the transfer station perimeter light pole node and the two observation primitives of the communication tower node meet the reconstruction conditions. These five backup observation primitives are connected in series to form a reconstruction chain segment and inserted into the original connection chain interruption position.

[0203] The updated sequence of continuous positioning points has re-covered the missing track in front of the high-rise buildings on the east bank, and the continuous track has been restored to a complete form that extends smoothly from the middle of the river to above the transfer station. This shows that the reconstruction process of the fractured section is a verifiable reconstruction of the interrupted chain segment using the real observation results of the backup nodes.

[0204] After the reconstruction of the fractured section is completed, the node correction parameters are further updated based on the reconstruction results. Specifically, the continuous tracks before and after reconstruction are compared to identify the abnormal nodes that cause missing links or navigation mismatches. Abnormal nodes are not limited to nodes with hardware failures, but refer to nodes whose observation results deviate continuously from the continuous tracks confirmed by reconstruction within a certain period of time. The tracks before reconstruction, the tracks after reconstruction, and the original observation elements of each node can be compared one by one. If the direction survey line corresponding to a node deviates from the position direction corresponding to the track after reconstruction for a long time, or the distance constraint generated by its time delay continuously falls outside the target position after reconstruction, or the velocity trend generated by its frequency shift does not match the direction of the track after reconstruction, then the node is identified as an abnormal node.

[0205] After the abnormal node is identified, the direction offset, time delay offset and frequency shift deviation are extracted from the observation primitive corresponding to the abnormal node. The direction offset can be obtained by the angle difference between the direction survey line corresponding to the abnormal node and the target direction of the reconstructed continuous track under the same time mark.

[0206] The time delay offset can be calculated from the difference between the distance constraint corresponding to the abnormal node and the actual spatial distance from the reconstructed target location to the node;

[0207] The frequency shift deviation can be obtained from the difference between the velocity constraint corresponding to the abnormal node and the radial velocity component of the reconstructed continuous track. Taking the rooftop node near the high-rise buildings on the east bank as an example, if the direction detection continues to deflect to the north due to the reflection of the metal curtain wall for a certain period of time, its direction offset will continue to be large.

[0208] If the propagation time of the same node is measured to be longer due to local link clock drift, its delay offset will also increase accordingly. These offsets and deviations provide a direct basis for subsequent node corrections.

[0209] Based on the extracted directional offset, time delay offset, and frequency shift deviation, the system updates the node correction parameters of the corresponding abnormal node. The node correction parameters may include at least directional correction parameters, time delay correction parameters, and frequency shift correction parameters. The directional correction parameters, time delay correction parameters, and frequency shift correction parameters are stored in a one-to-one correspondence with the node identifier. When a node is identified as an abnormal node in multiple consecutive reconstruction segments, the directional offset, time delay offset, and frequency shift deviation in the multiple reconstruction segments corresponding to that node are read to determine the updated parameters of that node. Parameter compensation is then performed after the node completes direction detection, propagation time determination, and frequency offset detection, and before generating direction survey lines, distance constraints, and velocity constraints.

[0210] The orientation correction parameter is used to correct the arrival direction or orientation survey line output by the node, so that it is more consistent with the true orientation under the unified spatial reference.

[0211] The time delay correction parameter is used to correct the propagation time measurement results, so that the distance constraint generated by the node is closer to the real spatial distance;

[0212] Frequency shift correction parameters are used to correct the frequency offset detection results of the node, so that the derived velocity constraints are more in line with the actual radial motion trend of the target. The update method can be to directly write back the deviation measured in the current reconstruction section, or to combine the historical deviations in multiple reconstruction sections to form a long-term correction amount for the node. For the aforementioned Jiang'an rooftop node, if the system finds that its direction measurement line is continuously deviated to the north relative to the actual flight track in multiple flight sorties, the deviation amount can be written into the direction correction parameter table, and angle compensation can be performed before the subsequent direction detection results are converted into direction measurement lines.

[0213] After the update is completed, the new node correction parameters are used to perform correction processing on the subsequently received air target signals. Specifically, after the direction of arrival is obtained by direction detection, direction compensation is first performed according to the direction correction parameters, and then a unified spatial reference conversion is performed.

[0214] After obtaining the time delay by measuring the propagation time, time compensation is first performed according to the time delay correction parameters, and then distance constraints are generated.

[0215] After the frequency shift is detected, the deviation is corrected according to the frequency shift correction parameters, and then the velocity constraint is generated.

[0216] After the above processing, the spatiotemporal consistency between the observation results of the original abnormal node and other nodes in subsequent operation will be significantly improved, thereby reducing the probability of recurrence of chain interruption, positioning point jump or navigation state mismatch.

[0217] In this embodiment, the final output of the air target navigation tracking and positioning result is not a single coordinate value, but includes a corrected continuation chain, a continuous positioning point sequence, a continuous track, and the path holding state, yaw correction state and continuation guidance state corresponding to the continuous track. The continuation chain is used to characterize the continuous observation connection relationship of the air target between distributed observation nodes.

[0218] A continuous sequence of location points is used to characterize the discrete position sequence of an aerial target on a unified time axis; a continuous flight path is used to characterize the continuous flight trajectory of an aerial target.

[0219] The path holding status, yaw correction status, and follow-up guidance status are used to characterize the current navigation relationship between the air target and the preset route and the subsequent guidance direction;

[0220] By simultaneously outputting the above information, it is possible not only to provide the current location of the aerial target, but also to provide information on how the target evolved from the previous flight path, whether it has deviated from the preset route, and how it should continue its flight, thus forming a complete, traceable, and continuously updated navigation, tracking, and positioning result.

[0221] Furthermore, a dedicated data organization structure can be set up for fracture segment reconstruction and node correction. The fracture segment record entry can include at least the fracture start time, fracture end time, fracture front end location point identifier, fracture rear end location point identifier, spare node set identifier, and reconstruction search range description; the reconstruction chain segment record entry can include at least the spare observation primitive identifier set, chain segment start and end time, chain segment continuity mark, and chain segment insertion position mark; the node correction parameter record entry can include at least the node identifier, direction correction parameter, time delay correction parameter, frequency shift correction parameter, parameter update time, and parameter source segment identifier.

[0222] The above embodiments can be implemented, in whole or in part, by software, hardware, firmware, or any other combination thereof. When implemented using software, the above embodiments can be implemented, in whole or in part, in the form of a computer program product.

[0223] Those skilled in the art will recognize that the modules and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0224] In addition, the functional modules in the various embodiments of this application can be integrated into one processing module, or each module can exist physically separately, or two or more modules can be integrated into one module.

[0225] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

[0226] In conclusion, the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A distributed aerial target navigation, tracking, and positioning method, characterized in that: The specific steps include: The system collects aerial target signals received by each distributed wireless observation node, extracts arrival direction, time delay, frequency shift, node attitude and time stamp, and converts them into direction survey lines, range constraints and velocity constraints under a unified spatiotemporal reference to generate an observation primitive sequence. Based on the temporal continuity, spatial adjacency, and velocity continuity between observation primitives, a two-way continuity determination is performed to generate a continuity chain corresponding to a single aerial target; The direction survey lines and distance constraints in the docking chain are intersected and solved. The target positioning subdomain is determined by combining the velocity constraints and the spatial boundary. A continuous positioning point sequence is generated based on the target positioning subdomain. Connecting consecutive positioning point sequences forms a continuous track, and based on the relationship between the continuous track and preset route constraints, path holding state, yaw correction state, and follow-up guidance state are generated. When a link is missing or the navigation status is abnormal, the observation primitives of the backup node are called to reconstruct the broken section, and the node correction parameters are updated according to the reconstruction results, and the navigation tracking and positioning results of the air target are output.

2. The distributed aerial target navigation, tracking, and positioning method according to claim 1, characterized in that: Collect aerial target signals received by each distributed wireless observation node, and extract arrival direction, time delay, frequency shift, node attitude, and time stamp, including: Direction detection is performed on the air target signals received by each distributed wireless observation node to obtain the arrival direction of the corresponding air target signal; Propagation time is measured for the aerial target signals received by each distributed wireless observation node to obtain the time delay of the corresponding aerial target signals; Frequency offset detection is performed on the air target signals received by each distributed wireless observation node to obtain the frequency shift of the corresponding air target signal; Read the node attitude and time stamp of each distributed wireless observation node at the corresponding acquisition time; The arrival direction, time delay, frequency shift, node attitude, and time stamp are associated and written into the same observation record entry.

3. The distributed aerial target navigation, tracking, and positioning method according to claim 2, characterized in that: The arrival direction, time delay, frequency shift, node attitude, and time stamp are converted into direction survey lines, range constraints, and velocity constraints under a unified spatiotemporal reference, generating an observation primitive sequence including: Time alignment was performed on the time stamps in each observation record entry based on a unified time reference. Based on a unified spatial reference, attitude conversion is performed on the node attitudes in each observation record entry, and the arrival direction is converted into a direction survey line. The distance constraints for corresponding aerial targets are generated based on the time delay; Generate velocity constraints for corresponding aerial targets based on frequency shift; Write the direction survey line, distance constraint, and velocity constraint into the corresponding observation record entries, and attach node identifiers and spatiotemporal reference markers; The observation records are arranged in chronological order to generate an observation primitive sequence.

4. The distributed aerial target navigation, tracking, and positioning method according to claim 3, characterized in that: The continuation determination is performed based on the temporal succession, spatial adjacency, and velocity continuity relationships between observed primitives, including: The observation primitive sequence is traversed in chronological order. Using the current observation primitive as the predecessor primitive, candidate successor observation primitives are searched within a preset time neighborhood. The target position projection range is determined based on the direction survey lines, distance constraints, and velocity constraints of the precursor unit. Determine the intersection relationship between the direction survey line corresponding to the candidate successor observation primitive and the target position projection range, and generate spatial adjacency determination result and velocity continuation determination result based on the matching result between the distance constraint and velocity constraint corresponding to the candidate successor observation primitive and the target position projection range; A continuation determination result is generated based on the time continuation determination result, the spatial adjacency determination result, and the velocity continuation determination result.

5. A distributed aerial target navigation, tracking, and positioning method according to claim 4, characterized in that: Performing bidirectional succession determination to generate a succession chain corresponding to a single aerial target includes: Based on the succession determination results, establish a forward succession relationship between the predecessor primitive and the candidate successor observation primitive; Using candidate successor observation primitives as backtracking primitives, reverse retrieval is performed in the neighborhood of the preceding time sequence to establish a reverse continuation relationship between candidate successor observation primitives and predecessor primitives. When a candidate successor observation primitive and a predecessor primitive simultaneously satisfy both forward succession and reverse succession relationships, it is determined that there is a bidirectional succession relationship between them. Observational primitives with bidirectional succession relationships are concatenated in chronological order to generate candidate succession chains; The conflict observation primitives in the candidate succession chain are split and removed to generate succession chains corresponding to a single aerial target.

6. A distributed aerial target navigation, tracking, and positioning method according to claim 5, characterized in that: The intersection solution is performed on the direction survey lines and distance constraints in the docking chain. The target localization subdomain is determined by combining velocity constraints and spatial boundaries, including: Extract the direction survey lines and distance constraints at the same or adjacent times in the continuation chain; Perform spatial intersection calculations on the direction survey lines and distance constraints to generate candidate intersection regions; Determine the reachable range of the target's motion at corresponding moments based on velocity constraints; Determine the overlap between candidate intersection regions and the reachable range of movement, and delete candidate intersection regions located outside the airspace boundary; When there are more than two candidate intersection regions, the target positioning subdomain is determined based on the displacement and direction continuity relationships between each candidate intersection region and the target position at the previous moment.

7. A distributed aerial target navigation, tracking, and positioning method according to claim 6, characterized in that: Generating a sequence of continuous location points based on the target location subdomain includes: The center position of the target positioning subdomain at each time point is determined as a candidate positioning point; Based on the displacement and orientation relationships between candidate positioning points at adjacent time points, a continuity check is performed on the candidate positioning points; When there are multiple candidate positioning points at the current moment, the candidate positioning point that satisfies the displacement continuation condition and the direction continuation condition with the confirmed positioning point at the previous moment is selected as the positioning point at the current moment. Arrange the location points at each time point in chronological order to generate a continuous sequence of location points; When the displacement change between adjacent positioning points exceeds the preset jump boundary, the positioning point selection is re-executed according to the target positioning subdomain at the corresponding time, and the continuous positioning point sequence is updated.

8. A distributed aerial target navigation, tracking, and positioning method according to claim 7, characterized in that: Connecting consecutive positioning point sequences forms a continuous track, and based on the relationship between the continuous track and preset route constraints, path-keeping state, yaw correction state, and follow-up guidance state are generated, including: Connect the positioning points at adjacent times in a continuous positioning point sequence in chronological order to generate a continuous track; Extract the heading change, displacement change, and turning change results of adjacent track segments in a continuous track; Determine the deviation relationship between the continuous track and the preset route constraints, and generate a path holding state or yaw correction state based on the deviation relationship; Based on the direction of extension of the current flight path segment, the subsequent reachable range, and the pointing relationship of preset route constraints, a follow-up guidance status is generated; Write the path holding status, yaw correction status, and follow-up guidance status into the corresponding consecutive track entries.

9. A distributed aerial target navigation, tracking, and positioning method according to claim 8, characterized in that: When a break in the link is missing or the navigation status is abnormal, the reconstruction of the broken section by calling the observation primitives of the backup node includes: When there are time discontinuities in the connection chain, or when there is an abnormal switch in the path holding state, yaw correction state and connection guidance state corresponding to the continuous track, the break section is identified. Search for standby node observation primitives that did not participate in the construction of the current successor chain within the time range corresponding to the fracture segment; The reconstruction search range is determined based on the front and rear positioning points of the fracture segment, and the continuation determination is performed on the standby node observation elements within the reconstruction search range. The standby node observation primitives that meet the continuation conditions are connected in sequence according to time to generate the reconstruction chain segment corresponding to the fracture section. Insert the reconstructed chain segment into the corresponding position of the continuation chain, and update the continuous positioning point sequence and continuous track.

10. A distributed aerial target navigation, tracking, and positioning method according to claim 9, characterized in that: The node correction parameters are updated based on the reconstruction results, and the output of the air target navigation, tracking and positioning results includes: By comparing the continuous tracks before and after reconstruction, the abnormal nodes that cause missing links in the connection chain or abnormal navigation status can be identified. Extract the direction offset, time delay offset, and frequency shift deviation from the observation primitives corresponding to the abnormal nodes; Update the node correction parameters of the corresponding abnormal node based on the direction offset, time delay offset, and frequency shift deviation; The updated node correction parameters are used to perform correction processing on subsequently received air target signals; The corrected connection chain, continuous positioning point sequence, continuous track, path holding status, yaw correction status, and connection guidance status are output as the results of air target navigation, tracking, and positioning.

Images