Multi-source fusion communication aided navigation positioning method for waterborne vehicle

Abstract

The application discloses a multi-source fusion communication aided navigation positioning method for a waterborne load, and comprises the following steps: acquiring bidirectional communication signals between the waterborne load and at least two communication reference nodes, wherein the bidirectional communication signals contain sending time stamps, receiving time stamps and channel impulse response sequences in the same communication period; performing multi-path component extraction on the channel impulse response sequences to generate multi-path structure observations; constructing time bias state variables based on the bidirectional communication signals, directly embedding the time bias state variables into a time delay expression of the multi-path structure observations, and forming a coupling expression relationship; constructing a space-time joint constraint equation set based on the coupling expression relationships corresponding to the different communication reference nodes; and synchronously obtaining space position information of the waterborne load and time bias state estimation results by jointly solving the space-time joint constraint equation set. The application significantly improves navigation positioning accuracy and reliability.

Description

Technical Field

[0001] This invention relates to the field of underwater acoustic positioning and navigation technology, and in particular to a multi-source fusion communication-assisted navigation and positioning method for underwater vehicles. Background Technology

[0002] Traditional underwater navigation and positioning methods typically treat the underwater vehicle's position calculation and clock synchronization as two separate problems, processed step by step. For example, existing synchronous positioning methods based on two-way ranging first estimate and correct the clock deviation between the underwater vehicle and the communication reference node through two-way communication; then, under the assumption of "synchronization," they use propagation delay measurements to calculate the vehicle's spatial position.

[0003] However, the complex and variable channel environment underwater (such as multipath effects, time-varying fading, and non-line-of-sight propagation) results in significant errors in clock skew estimation. In a step-by-step processing mode, these errors inevitably propagate to subsequent positioning calculations, creating a vicious cycle of "clock skew error → positioning error → clock skew estimation error for the next cycle." Furthermore, traditional methods often treat multipath signals as interference or simply discard them, failing to recognize the rich spatial geometric constraints inherent in the multipath structure itself. This information could be used to enhance positioning constraints and assist in the synchronization calibration of clock skew.

[0004] Therefore, existing step-by-step processing methods struggle to achieve high-precision and high-reliability navigation and positioning in underwater environments with strong multipath propagation and low signal-to-noise ratios. To address these technical problems, this invention proposes a solution that directly embeds time-deviation state variables into multipath structural observations and constructs a spatiotemporal joint constraint equation set for simultaneous solution. Summary of the Invention

[0005] The purpose of this section is to outline some aspects of embodiments of the present invention and to briefly describe some preferred embodiments. Simplifications or omissions may be made in this section, as well as in the abstract and title of this application, to avoid obscuring the purpose of these documents; however, such simplifications or omissions should not be construed as limiting the scope of the invention.

[0006] To solve the above-mentioned technical problems, the present invention provides the following technical solution: a multi-source fusion communication-assisted navigation and positioning method for underwater vehicles, comprising the following steps:

[0007] Acquire bidirectional communication signals between the underwater vehicle and at least two communication reference nodes, wherein the bidirectional communication signals contain a transmission timestamp, a reception timestamp, and a channel impulse response sequence within the same communication cycle;

[0008] Multipath components are extracted from the channel impulse response sequence to generate multipath structure observations that include multipath propagation delay structure and path energy distribution;

[0009] Based on the bidirectional communication signal, a time deviation state variable is constructed, and the time deviation state variable is directly embedded into the time delay expression of the multipath structure observation, so that the multipath propagation time delay structure and the time deviation state variable form a coupled expression relationship.

[0010] A spatiotemporal joint constraint equation set is constructed based on the coupling expression relationship corresponding to different communication reference nodes. The spatiotemporal joint constraint equation set contains both spatial propagation constraint terms and temporal state constraint terms, and the two are bound together through the same multipath structure observation.

[0011] By jointly solving the spatiotemporal constraint equations, the spatial location information and time deviation state estimation results of the underwater vehicle can be obtained simultaneously.

[0012] As a preferred embodiment of the multi-source fusion communication-assisted navigation and positioning method for underwater vehicles described in this invention, the generation of the multipath structure observation includes:

[0013] The channel impulse response sequence is subjected to time-domain matched filtering to obtain a multipath response function, and effective response components that satisfy the dual threshold conditions of amplitude threshold and duration are identified in the multipath response function.

[0014] A multipath propagation delay structure vector is constructed for the effective response components according to their occurrence time order, and the corresponding path energy distribution vector is generated simultaneously.

[0015] The consistency of the multipath propagation delay structure vector is checked in a continuous time window. By comparing the multipath propagation delay structure vectors in adjacent time windows, the topological change difference value between the two is obtained. When the topological change difference value exceeds the preset structural continuity constraint threshold, the corresponding response component is marked as a structural fracture component and suppressed.

[0016] Using the remaining response components after suppression, the multipath propagation delay structure vector and path energy distribution vector are reconstructed, and the reconstructed vectors are used as the final multipath structure observation.

[0017] As a preferred embodiment of the multi-source fusion communication-assisted navigation and positioning method for underwater vehicles described in this invention, the process of constructing and embedding the time deviation state variable specifically includes:

[0018] Based on the bidirectional communication signal between the underwater body and the communication reference node, the timestamp mapping relationship between uplink and downlink communication is established respectively, and the round-trip propagation time is expressed as the superposition of the propagation time component and the time deviation state component.

[0019] The time deviation state component is defined as a state variable that evolves over time, and is described by a state transition model that includes linear drift terms and random disturbance terms.

[0020] The time deviation state variable is embedded into the time delay expression of the multipath structure observation, so that the multipath structure observation is transformed into a coupled observation that depends on the time deviation state variable. The time deviation state variable does not participate in the localization solution as a separate correction parameter, but participates in the joint solution as an endogenous variable of the spatiotemporal joint constraint equation system.

[0021] As a preferred embodiment of the multi-source fusion communication-assisted navigation and positioning method for underwater vehicles described in this invention, the construction process of the spatiotemporal joint constraint equation set specifically includes:

[0022] For different communication reference nodes, local constraint equations are established based on their respective coupling expression relationships, and a unified time deviation state variable is introduced into each local constraint equation so that the local constraint equations of different nodes share the same time state space. Among them, the spatial propagation constraint term and the time state constraint term in each local constraint equation are bound through the same multipath structure observation, rather than being constructed independently and then spliced ​​together.

[0023] The local constraint equations of different communication reference nodes are subjected to consistency coupling processing, so that all local constraint equations satisfy the global consistency condition in the same state variable space, thereby forming an indivisible spatiotemporal joint constraint equation set.

[0024] As a preferred embodiment of the multi-source fusion communication-assisted navigation and positioning method for underwater vehicles described in this invention, the joint solution process specifically includes:

[0025] The spatiotemporal joint constraint equations are transformed into a nonlinear optimization problem with variable constraints, wherein the optimization variables include both the spatial location variable of the underwater vehicle and the temporal deviation state variable.

[0026] A unified cost function based on the observation error of multipath structure is constructed, and spatial error and temporal deviation state error are incorporated into the same optimization objective function for joint minimization.

[0027] During the iterative solution process, the spatial location variable and the time deviation state variable are updated simultaneously, so that the update result of any variable depends on the current estimate of the other variable, thus forming a strongly coupled bivariate solution process.

[0028] As a preferred embodiment of the multi-source fusion communication-assisted navigation and positioning method for underwater vehicles described in this invention, wherein: at least two of the communication reference nodes achieve state sharing through the unified time deviation state variable, including:

[0029] The multipath structure observations generated by each communication reference node are mapped to a unified time-bias state variable space, and cross-node observation correlation is achieved through this shared time-bias state variable.

[0030] When constructing the spatiotemporal joint constraint equation set, state variable consistency constraints are applied to the local constraint equations of each node. This allows the multipath structure observation errors of different nodes to be propagated and corrected through the same time deviation state variable, thereby enabling the multipath structure observations of each communication reference node to be jointly solved in a unified time deviation state variable space, rather than being processed independently.

[0031] As a preferred embodiment of the multi-source fusion communication-assisted navigation and positioning method for underwater vehicles described in this invention, the screening process for structural continuity constraints on the multipath structural observations specifically includes:

[0032] The sequence consistency of the multipath propagation delay structure vectors generated in multiple consecutive communication cycles is compared, and the differences in topological changes between the structure vectors are calculated.

[0033] When the difference in topological change exceeds the structural continuity constraint threshold, the corresponding multipath response component is marked as a structural fracture component.

[0034] A time consistency weight is introduced into the multipath structural observations retained after screening by structural continuity constraints, so that they participate in the calculation in a weighted form in the subsequent spatiotemporal joint constraint equations.

[0035] As a preferred embodiment of the multi-source fusion communication-assisted navigation and positioning method for underwater vehicles described in this invention, wherein: during the iterative solution process, a reverse constraint mechanism based on multipath structure observations is introduced, including:

[0036] By utilizing the deviation relationship between the multipath propagation delay structure vector and the theoretical propagation time, a reverse correction term for the time deviation state variable is constructed;

[0037] The reverse correction term is incorporated as part of the state update equation, so that the update of the time-biased state variable depends on both the historical time state and the current multipath structure observation.

[0038] The temporal state evolution process is thus simultaneously influenced by both time series constraints and spatial propagation constraints, forming a closed-loop correction for time-biased state variables.

[0039] In a preferred embodiment of the multi-source fusion communication-assisted navigation and positioning method for underwater vehicles described in this invention, before acquiring the bidirectional communication signal, the communication reference node undergoes a filtering process based on link state consistency. This filtering process includes:

[0040] Based on the channel impulse response stability and round-trip propagation time consistency of the communication link, a link state consistency index is constructed.

[0041] When the link state consistency index is lower than a preset threshold, the corresponding communication reference node is excluded from the spatiotemporal joint constraint equation set.

[0042] The link selection result not only affects the input data set, but also directly affects the observability structure of the time deviation state variable, thereby forming a binding relationship between the selected set of communication reference nodes and the time deviation state variable space.

[0043] This invention also provides a multi-source fusion communication-assisted navigation and positioning system for underwater vehicles, applied to the aforementioned multi-source fusion communication-assisted navigation and positioning method for underwater vehicles, comprising:

[0044] The communication signal acquisition module is used to acquire the bidirectional communication signal between the underwater body and the communication reference node;

[0045] The multipath structure observation generation module is used to generate multipath structure observations that include multipath propagation delay structure and path energy distribution, and to serve as propagation structure constraint inputs.

[0046] The time deviation state modeling module is used to construct time deviation state variables and directly embed them into the time delay expression of the multipath structure observations to form a coupled expression relationship.

[0047] The spatiotemporal joint constraint construction module is used to construct a set of spatiotemporal joint constraint equations based on the coupling expression relationships corresponding to different communication reference nodes.

[0048] The joint solution module is used to simultaneously solve the spatial location variables and time deviation state variables of the underwater body within the same optimization framework, and output the spatial location information and time deviation state estimation results.

[0049] The beneficial effects of this invention are:

[0050] 1. This invention directly embeds the time deviation state variable into the time delay expression of the multipath structure observation, and constructs an indivisible spatiotemporal joint constraint equation set that binds the spatial propagation constraint and the time state constraint through the same multipath observation. This achieves synchronous and strongly coupled solution of the underwater vehicle position and clock deviation, fundamentally overcoming the defect of error accumulation caused by the step-by-step processing of positioning and time synchronization in traditional methods, and significantly improving the accuracy and reliability of navigation and positioning.

[0051] 2. This invention effectively suppresses the "structural break" anomalous components caused by sudden interference and non-line-of-sight propagation in time-varying underwater channels by performing dual threshold screening, multipath structure vector construction, and continuous time window consistency verification on the channel impulse response sequence. Furthermore, it introduces time consistency weights to adaptively weight high-quality multipath observations, thereby providing robust and structured input data for spatiotemporal joint solution, enabling the navigation system to maintain stable and accurate positioning output even under harsh hydrological conditions.

[0052] 3. This invention models the time deviation state variable as an evolutionary state containing linear drift and random disturbance, and embeds it as an endogenous variable into the multipath time delay expression. At the same time, it introduces a reverse constraint mechanism based on multipath structure consistency during the iterative solution process, so that the clock deviation update is subject to dual closed-loop correction by historical time series constraints and spatial propagation constraints. This achieves continuous self-calibration of time synchronization accuracy under long flight time conditions and effectively reduces the dependence on high-precision crystal oscillators. Attached Figure Description

[0053] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort. Wherein:

[0054] Figure 1 This is an overall flowchart of the multi-source fusion communication-assisted navigation and positioning method for underwater vehicles according to the present invention.

[0055] Figure 2 This is a flowchart illustrating the multipath structure observation generation process of the multi-source fusion communication-assisted navigation and positioning method for underwater vehicles according to the present invention.

[0056] Figure 3 This is a flowchart illustrating the construction and embedding of time deviation state variables in the multi-source fusion communication-assisted navigation and positioning method for underwater vehicles according to the present invention.

[0057] Figure 4 This is a flowchart illustrating the spatiotemporal joint constraint equation set construction process of the multi-source fusion communication-assisted navigation and positioning method for underwater vehicles according to the present invention.

[0058] Figure 5 This is a flowchart illustrating the joint solution and reverse constraint process of the multi-source fusion communication-assisted navigation and positioning method for underwater vehicles according to the present invention. Detailed Implementation

[0059] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, the specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings.

[0060] Many specific details are set forth in the following description in order to provide a full understanding of the invention. However, the invention may also be practiced in other ways different from those described herein, and those skilled in the art can make similar extensions without departing from the spirit of the invention. Therefore, the invention is not limited to the specific embodiments disclosed below.

[0061] Secondly, the term "one embodiment" or "embodiment" as used herein refers to a specific feature, structure, or characteristic that may be included in at least one implementation of the present invention. The phrase "in one embodiment" appearing in different places in this specification does not necessarily refer to the same embodiment, nor is it a single or selective embodiment that is mutually exclusive with other embodiments.

[0062] Secondly, the present invention is described in detail with reference to the schematic diagrams. When detailing the embodiments of the present invention, for ease of explanation, the cross-sectional views illustrating the device structure may be partially enlarged, not according to the usual scale. Furthermore, the schematic diagrams are merely examples and should not limit the scope of protection of the present invention. In addition, actual fabrication should include three-dimensional spatial dimensions of length, width, and depth.

[0063] Example 1

[0064] Reference Figure 1-5 The first embodiment of the present invention provides a multi-source fusion communication-assisted navigation and positioning method for underwater vehicles, comprising the following steps:

[0065] S1: Acquire bidirectional communication signals between the underwater vehicle and at least two communication reference nodes. The bidirectional communication signals contain a transmission timestamp, a reception timestamp, and a channel impulse response sequence (i.e., a CIR sequence) within the same communication cycle.

[0066] The underwater vehicles include, but are not limited to, autonomous underwater vehicles (AUVs) or remotely operated underwater vehicles (ROVs). The communication reference node is an underwater beacon or surface buoy with a pre-calibrated absolute position, whose position is pre-calibrated using differential GPS or a long-baseline underwater acoustic positioning system. In one specific implementation, the positioning error is less than 0.1 meters.

[0067] The same communication period refers to a complete time window from when the underwater vehicle initiates its first ranging request until it receives response signals from all participating nodes. For example, when the underwater vehicle communicates with three communication reference nodes using a polling mechanism, if the communication distance is 500 meters and the speed of sound is approximately 1500 meters per second, the round-trip propagation time is approximately 0.67 seconds. Adding the node processing delay (typically 0.1 seconds), the complete interaction time for each node is approximately 0.8 seconds, and the total duration for the three nodes is approximately 2.4 seconds. Therefore, the typical value of the communication period is between 1 and 5 seconds, with the specific value determined by the communication distance, the speed of sound, and the number of nodes.

[0068] In one specific implementation, the bidirectional communication between the underwater vehicle and the communication reference node adopts a half-duplex mode, and the uplink and downlink signals are exchanged sequentially within the same communication cycle according to a preset Time Division Multiple Access (TDMA) protocol.

[0069] In one specific implementation, the transmission and reception timestamps are recorded by the local high-precision crystal oscillator driven clock circuits of the underwater device and the communication reference node, respectively, with a timestamp resolution at the microsecond level. Taking a single bidirectional communication between the underwater device and a certain reference node as an example: the underwater device at its local clock time... Send a ranging request signal, which contains The value; the communication reference node at the local clock time Received the request and recorded it. The reference node at the local clock time A response signal was sent, which contained... and The water-based download body is at the local clock time. Received a response signal and recorded it. The above. , , , These are the sending timestamp and the receiving timestamp.

[0070] in, , , , This will be used in subsequent steps to construct the time deviation state variable. . This represents the relative time deviation between the local clock of the underwater device and the local clock of the communication reference node, which is used as a state variable to be estimated in the spatiotemporal joint solution (see step S3 for details).

[0071] In one exemplary implementation, the channel impulse response sequence is obtained by embedding a known broadband training sequence (such as a linear frequency modulated signal or a pseudo-random code) into the communication signal. After receiving the signal, the underwater receiver and the communication reference node perform time-domain matched filtering or frequency-domain correlation operations with the received signal using a locally stored copy, thereby estimating the channel impulse response sequence.

[0072] Specifically, matched filtering is implemented through the following sliding correlation operation:

[0073] ;

[0074] in For the received baseband complex signal (with time) change); A locally stored copy of the training sequence. Its conjugate; The duration of the training sequence; This is the estimated channel impulse response.

[0075] Calculation result It's about latency. The complex function is discretized into a complex vector in the digital domain. Each element Each discrete time delay tap represents the channel complex gain at that time delay position. This vector is the channel impulse response sequence (CIR sequence).

[0076] In a preferred embodiment, to ensure that signals from at least two communication reference nodes can be acquired conflict-free within the same communication cycle, the system employs either a polling or parallel code division multiple access (CDMA) mechanism. When using the polling mechanism, the underwater device sequentially communicates bidirectionally with node 1, node 2, ..., node N. The complete round-trip time of each node is allocated to an independent time slot, and all time slots constitute a superframe, which is one communication cycle. When using the CDMA mechanism, each reference node responds simultaneously using mutually orthogonal spreading codes. The underwater device extracts signals from multiple nodes in parallel through despreading, thereby completing bidirectional communication with multiple nodes within a shorter time window, which is also defined as one communication cycle.

[0077] In a preferred embodiment, to ensure the validity of multipath structure observations in subsequent steps, the CIR sequence is preprocessed after acquiring the channel impulse response sequence and before link filtering: zero-frequency components are removed (i.e., the mean of the sequence is subtracted), a Hamming window is applied to suppress sidelobes, and energy is normalized (making the total energy of the sequence equal to 1). The preprocessed CIR sequence is used for link state consistency filtering and multipath component extraction in step S2.

[0078] Furthermore, before acquiring the bidirectional communication signals (i.e., before formally using the data from the communication reference nodes to construct the spatiotemporal joint constraint equations), the communication reference nodes undergo a screening process based on link state consistency to ensure that all nodes participating in subsequent solutions are high-quality nodes. The screening process specifically includes the following sub-steps:

[0079] S11: Construct link state consistency metrics Specifically, a link state consistency index is constructed based on the channel impulse response stability and round-trip propagation time consistency of the communication link. .

[0080] Among them, the channel impulse response stability (denoted as) ), which characterizes the similarity of multipath delay structures in CIR sequences over multiple consecutive communication cycles.

[0081] In one specific implementation, quantification is achieved by calculating the Pearson correlation coefficient between the preprocessed CIR sequence of the current period and the CIR sequence of the previous communication period. For example, suppose both CIR sequences have a length of [missing information]. The formula for calculating the Pearson correlation coefficient is:

[0082] ;

[0083] in, and The two CIR sequences are respectively in the first... The complex amplitude modulus of a delay tap, and The mean is used. When the sequence lengths are inconsistent, they are aligned to the same length through interpolation or truncation. The range of values ​​is The closer the value is to 1, the more stable the channel structure.

[0084] Among them, the consistency of round-trip propagation time (denoted as ) The round-trip time (RTT) characterizes the stability of the round-trip propagation time obtained from multiple consecutive two-way ranging measurements. Specifically, the round-trip propagation time is first defined. for: This expression eliminates the influence of clock skew between the underwater module and the reference node. Then, the most recent K times (preferably, K ≥ 5) are calculated. variance of measured values , Defined as the reciprocal of the variance and normalized to Interval: Or through Calculation, where This is the preset reference variance. The closer to 1, the more likely it is to be a number. The more stable the measurement, the better the time synchronization stability of the link.

[0085] S12: Construct a comprehensive link state consistency metric Specifically, a comprehensive link state consistency indicator is constructed through weighted fusion: ,in, and These are preset weighting coefficients.

[0086] In a preferred embodiment, the weighting coefficient is taken as: , This reflects a relative emphasis on the consistency of round-trip transmission time, because The stability is directly related to the time deviation state variable The observability of the data. This value is determined based on the relative statistical characteristics of channel variation and clock drift in a typical underwater environment, and can be obtained through pool testing or simulation calibration before deployment.

[0087] S13: Threshold Comparison and Node Exclusion. Specifically, when the link state consistency index... Below the preset threshold If the node is currently in an unreliable state (which may be caused by drastic channel changes, node failure, or strong interference), it is excluded from the set of participating nodes in the current positioning cycle. That is, the node does not participate in the construction of the spatiotemporal joint constraint equation set in the subsequent S4 step.

[0088] In a preferred embodiment, a preset threshold is used. The determination method is as follows: In the initial stage of system deployment, a large amount of link status data is collected under typical working conditions, and statistics are compiled during normal operation. The distribution characteristics of the values ​​will Set it to 0.5 times its mean or the first quartile. For example, if... Normalized to Interval, then It can be preset to 0.6. When When the threshold is less than 0.6, the link status is determined to not meet the consistency requirements. This threshold can be manually or adaptively adjusted according to seasonal changes in the actual marine environment or water turbidity.

[0089] It should be noted that the above link selection results not only affect the input data set, but also directly affect the time deviation state variable. The observable structure. Specifically, time-biased state variables. The observability depends on the spatial distribution geometry of the selected communication reference nodes. If the selected communication reference nodes are too concentrated in space (e.g., all located on the same side of the carrier), then... The coupling relationship with carrier location variables may degenerate into unobservable behavior; however, through a link filtering mechanism, the system can proactively select a set of nodes with reasonable spatial distribution and good link quality, thereby enabling the selected node set to... A physical binding relationship is formed between the observable spaces. This binding relationship ensures that the time-biased state variables can be uniquely and stably estimated in the subsequent joint solution of S5.

[0090] After the above link filtering process, the bidirectional communication signals (including transmission timestamps, reception timestamps, and preprocessed CIR sequences) corresponding to the retained communication reference nodes will be used as valid inputs and proceed to step S2 to generate multipath structure observations.

[0091] It should be noted that this step, by acquiring the bidirectional communication signals between the underwater vehicle and at least two communication reference nodes, and combining link state consistency screening, eliminates unreliable nodes, providing high-quality and highly stable timestamp and channel impulse response data for subsequent steps, thus laying a reliable data foundation for multipath structure extraction and spatiotemporal joint solution.

[0092] S2. Extract multipath components from the channel impulse response sequence to generate multipath structure observations that include multipath propagation delay structure and path energy distribution.

[0093] Specifically, generating multipath structure observations includes:

[0094] The channel impulse response sequence is subjected to time-domain matched filtering to obtain the multipath response function, and the effective response components that satisfy the dual threshold conditions of amplitude threshold and duration are identified in the multipath response function.

[0095] Construct a multipath propagation delay structure vector for the effective response components in chronological order of their occurrence, and simultaneously generate the corresponding path energy distribution vector;

[0096] A continuous time window consistency check is performed on the multipath propagation delay structure vector. By comparing the multipath propagation delay structure vectors in adjacent time windows, the topological change difference value between the two is obtained. When the topological change difference value exceeds the preset structural continuity constraint threshold, the corresponding response component is marked as a structural fracture component and suppressed.

[0097] Using the remaining response components after suppression, the multipath propagation delay structure vector and path energy distribution vector are reconstructed, and the reconstructed vectors are used as the final multipath structure observation.

[0098] Specifically, the multipath response function refers to the complex CIR sequence obtained in S1. The amplitude-delay curve obtained by taking the modulus (or power value) of the complex amplitude at each discrete delay tap is denoted as . ,in relative time delay ( , (This refers to the sampling period). or This curve visually reflects the change in signal strength with arrival time for different propagation paths.

[0099] In one specific implementation, the amplitude threshold is set to the peak amplitude of the multipath response function. times, The value range is 0.1 to 0.3, with a typical value of 0.2; the duration threshold is set to 2 to 5 times the system sampling period, with a typical value of 3 times the sampling period, meaning that the half-power width of the effective response component needs to continuously cover at least 3 sampling points to avoid misjudging narrow pulse noise as effective multipath.

[0100] In one specific implementation, a multipath propagation delay structure vector is constructed for the effective response components according to their arrival time order, and the corresponding path energy distribution vector is generated simultaneously. Specifically, this includes: extracting N effective multipath components and sorting them according to their arrival time in ascending order. The first arriving response component is defined as the receiving path component (usually corresponding to the direct path between the underwater vehicle and the communication reference node, or the first arriving refracted / diffracted path when the direct path is blocked), and its arrival time is denoted as... In subsequent processing, this is usually normalized to 0 and used as a reference for relative time delay. The first... The arriving response component is defined as the first... The radial component, whose arrival time is denoted as . Then the first The time difference of arrival of the radial relative to the first radial is The multipath propagation delay structure vector constructed in this way is denoted as... (in Or you can use it directly. and order The corresponding path energy distribution vector is denoted as... ,in For the first The energy of a path can be obtained by squaring or integrating the CIR complex amplitude modulus corresponding to that path.

[0101] In one specific implementation, the topology change difference value is calculated using dynamic time-bending (DTW) distance or Euclidean distance. For example, suppose two adjacent communication cycles (windows) and window The multipath propagation delay structure vectors are respectively and If both dimensions are the same (i.e., the number of multipaths has not changed drastically), then the normalized Euclidean distance is calculated: If the dimensions are different, align them using DTW first, then calculate the distance. Exceeding the preset structural continuity constraint threshold When this occurs, it is determined that the multipath structure of the current window has broken. It is important to note that when... When =0, we can let =0, meaning there is no effective multipath component, which is considered as no fracture.

[0102] In a preferred embodiment, the structural continuity constraint threshold The typical value range is 0.1 to 0.3, with 0.15 being preferred. This threshold... The data can be obtained through offline collection of typical underwater environment data and statistical calibration: calculate the DTW / Euclidean distance distribution under continuous and stable channels, and take 2 to 3 times the mean as the threshold.

[0103] Specifically, after suppressing the structural fracture component, the remaining response components typically correspond to physically stable multipaths (such as surface reflection paths and seabed reflection paths), whose temporal delay structure and energy distribution exhibit continuity and consistency over a short period. The reconstructed vector is denoted as... and This serves as the input for the subsequent spatiotemporal joint constraint equations.

[0104] Furthermore, the screening process for structural continuity constraints in multipath structure observations specifically includes:

[0105] The sequence consistency of the multipath propagation delay structure vectors generated in multiple consecutive communication cycles is compared, and the differences in topological changes between the structure vectors are calculated.

[0106] When the difference in topological change exceeds the structural continuity constraint threshold, the corresponding multipath response component is marked as a structural fracture component.

[0107] A time consistency weight is introduced into the multipath structural observations retained after screening by structural continuity constraints, so that they participate in the calculation in a weighted form in the subsequent spatiotemporal joint constraint equations.

[0108] In a preferred embodiment, multiple consecutive communication cycles refer to the preceding cycles, including the current cycle. One cycle, The typical value is 5 to 10.

[0109] In a preferred embodiment, sequence consistency comparison is performed using the average topological change difference within a sliding window: ,in This represents the difference in topological changes between adjacent periods within the window.

[0110] Specifically, time consistency weight Difference in topological changes from the current periodic multipath structure Inversely proportional: ,in The preset scale parameter (typically 0.1) means that the smaller the period of topological change (the more continuous the structure), the higher the weight of the multipath structure observations in the joint solution; conversely, the weight is reduced, thereby suppressing the influence of anomalous observations on the positioning results.

[0111] For example, when hour, ;when hour, This effectively enables automatic downweighting of unreliable observations.

[0112] It should be noted that, through the above step S2, the present invention extracts highly reliable multipath structure observations (including time delay structure vectors and energy distribution vectors) from the original CIR sequence after double threshold screening, time domain continuity verification and weighting processing, providing input data with clear structure and controllable quality for subsequent spatiotemporal joint constraint equations.

[0113] It should also be noted that the multipath propagation delay structure vector generated above... It not only records the absolute arrival time of each path, but more importantly, it describes the relative time delay relationship and topological order between paths (i.e., the order and interval of each path according to arrival time). This structured time delay information contains the geometric constraint relationship between the underwater vehicle and the communication reference node, as well as between the carrier and the reflection interface.

[0114] Since multipath propagation delay has an inherent coupling relationship with propagation path length and time synchronization error (specifically manifested as: measurement delay = geometric propagation delay + clock deviation), the time deviation state variable can be directly embedded into the time delay expression of the multipath structure observation, laying the foundation for the coupling modeling in the subsequent step S3.

[0115] S3. Construct time deviation state variables based on bidirectional communication signals, and directly embed the time deviation state variables into the time delay expression of multipath structure observations, so that the multipath propagation time delay structure and time deviation state variables form a coupled expression relationship.

[0116] Specifically, the construction and embedding process of time deviation state variables includes:

[0117] S31: Based on the bidirectional communication signal between the underwater body and the communication reference node, establish the timestamp mapping relationship between uplink and downlink communication respectively, and express the round-trip propagation time as the superposition of the propagation time component and the time deviation state component.

[0118] Specifically, let the local clock of the water-based system be... The local clock of the communication reference node is The relative time deviation between the two is defined as (That is, the node clock minus the carrier clock) (Possible positive or negative). Combined with the four timestamps recorded in S1 , , , Taking downstream propagation (node ​​→ carrier) as an example: the node at local time Sending a response signal, the carrier at local time The signal was received. Due to clock asynchrony, the apparent propagation delay measured by the carrier is... The actual geometric propagation delay is The following conditions must be met between the two:

[0119]

[0120] in, The geometric distance between the carrier and the node. Effective speed of sound; The clock offset of the carrier relative to the node at the time of reception (here) - Therefore, when the carrier clock is slow (The measurement delay is too large). Accordingly, the uplink propagation delay (carrier → node) can be expressed as: This scheme prioritizes downlink propagation delay to avoid introducing additional symbol processing. It should be noted that, due to the short underwater communication cycle (typically 1-5 seconds), it can be considered that within one communication cycle... Approximately constant.

[0121] S32: Define the time deviation state component as a state variable that evolves over time, and describe it using a state transition model that includes linear drift terms and random disturbance terms.

[0122] In one specific implementation, the time deviation state variable is defined as: ,in For the first The relative clock offset at the receiving time of each communication cycle Let be the clock frequency drift rate (i.e., the rate of change of deviation per unit time). Its evolution is described using a first-order linear state transition model:

[0123]

[0124] Wherein, the state transition matrix , Communication cycle duration (in seconds); The process noise follows a zero-mean Gaussian distribution, and its covariance matrix is... , and Power spectral densities for clock skew and drift rate, respectively, are typically obtained from crystal oscillator datasheets (e.g., temperature-compensated crystal oscillators). , ).

[0125] S33: Embed the time deviation state variable into the time delay expression of the multipath structure observation, so that the multipath structure observation is transformed into a coupled observation that depends on the time deviation state variable. Moreover, the time deviation state variable does not participate in the positioning solution as a separate correction parameter, but participates in the joint solution as an endogenous variable of the spatiotemporal joint constraint equation system.

[0126] Specifically, the absolute propagation delay (denoted as ) in multipath structure observations ,Should This refers to the actual propagation time of the signal from the communication reference node to the underwater receiver, which is different from the normalized relative arrival time in S2. Originally determined solely by geometric distance: After embedding the time offset state variable, the expression is modified according to the downlink propagation delay relationship of S31 as follows:

[0127]

[0128] To clearly express the mathematical coupling relationship between multipath structural observations and time-deviation state variables, this invention establishes the following structured observation model (based on nodes). The (Taking a multi-path as an example)

[0129]

[0130] It should be noted that the virtual reflection points of each multipath in this expression... It is not an independent variable, but rather determined by the location of the carrier. Since the geometric parameters of the reflection interface are uniquely determined, there is a deterministic geometric constraint relationship between the time delay observations of each path, and they are not independent distance measurements.

[0131] in, The first measured The absolute propagation delay of multiple paths; For the first The virtual reflection point position corresponding to the multipath ( hour That is, the first diameter corresponds to the actual position of the carrier. For a unified time deviation state variable; For observation noise. This expression clearly reveals that the time delay observation consists of three parts: geometric propagation distance (spatial term), clock deviation (time term), and noise, and is a computable mathematical model.

[0132] It is particularly important to emphasize that the observation model constructed in this invention does not treat each multipath as an independent ranging source for single-path TOA estimation, but rather utilizes the multipath propagation time delay structure vector. The relative time delay relationships and topological order (i.e., the order and intervals of the paths arranged by arrival time) described are used in constraint construction in the form of a path set. Specifically, the relative time delay difference between each path... The carrier location and the geometric parameters of the reflective interface form a deterministic functional relationship, which together form a joint constraint on the spatial location, rather than a single path participating independently in distance estimation.

[0133] For example, the time delay difference between the surface reflection path and the direct path is uniquely determined by the carrier depth and the node depth; the time delay difference between the seabed reflection path and the direct path is jointly determined by the carrier depth, water depth, and node depth. The relative time delay structure of multiple nodes and multiple paths constitutes an overdetermined constraint system, significantly enhancing geometric observability. This invention utilizes this multipath structure information, rather than relying solely on absolute time delay ranging, to achieve spatiotemporal integrated joint solution.

[0134] More generally, for the first Multiple paths, their time delay It can be represented as:

[0135]

[0136] in This represents the virtual propagation distance corresponding to this path (which may be greater than the direct distance). This value, relative to the arrival time difference of the first path, is directly taken from the relative time delay in the multipath structure observations generated by S2 (i.e., ,and Through the above embedding, the originally independent multipath delay observations are transformed. Become dependent Coupled observations.

[0137] Crucially, time-biased state variables Instead of being used as a pre-corrected parameter, it is treated as a state variable to be estimated, and its predicted value is influenced by the observation delay. Enter the optimization objective function (see the cost function in S5 for details). and This allows for its observability in the joint solution process. The variable is related to the carrier position. Together they constitute the optimization variables This is estimated synchronously in the subsequent joint solution of S5. This allows the multipath propagation delay structure to be solved simultaneously. It forms a strong coupling constraint with spatial location.

[0138] In one specific implementation, to further improve the accuracy of embedding, the influence of sound velocity variations with depth, temperature, and salinity can be considered, and the sound velocity... Modeled as a function of spatial location At this point, the above embedded expression becomes:

[0139]

[0140] in, This represents the actual propagation trajectory (spatial curve) of the signal from the communication reference node to the underwater body. Let be the arc length parameter along this trajectory; The corresponding arc length on the trajectory Spatial coordinates of the location; It is an arc-length infinitesimal element. However, to reduce computational complexity, a constant sound velocity approximation is usually used in shallow water and short baseline scenarios, and the sound velocity variation is included in the process noise.

[0141] It should be noted that, through the above step S3, this invention utilizes downlink propagation delay to construct an explicit coupling relationship between the time deviation state variable and the multipath delay, and incorporates clock deviation as an endogenous variable into the observation model, thereby realizing a unified representation of spatiotemporal information and providing a precise mathematical expression for constructing an indivisible spatiotemporal joint constraint equation set in step S4.

[0142] S4. Based on the coupling expression relationship corresponding to different communication reference nodes, construct a spatiotemporal joint constraint equation set, which includes both spatial propagation constraint terms and temporal state constraint terms, and the two are bound together through the same multipath structure observation.

[0143] Specifically, the construction process of the spatiotemporal joint constraint equation system includes:

[0144] For different communication reference nodes, local constraint equations are established based on their respective coupling expression relationships, and a unified time deviation state variable is introduced into each local constraint equation so that the local constraint equations of different nodes share the same time state space. Among them, the spatial propagation constraint term and the time state constraint term in each local constraint equation are bound through the same multipath structure observation, rather than being constructed independently and then spliced ​​together.

[0145] The local constraint equations of different communication reference nodes are subjected to consistency coupling processing, so that all local constraint equations satisfy the global consistency condition in the same state variable space, thereby forming an indivisible spatiotemporal joint constraint equation set.

[0146] In one specific implementation, the steps for establishing the local constraint equations are as follows:

[0147] Assume there is a total One communication reference node ( ),node The position coordinates are The location of the water-based body is For nodes Its multipath structure observations include: the absolute propagation delay of the first path. (Obtained from S33), and the arrival time difference vector relative to the first path. (Obtained from S2, where) ,and According to the coupling expression in S33, for node... Given the initial diameter, the following local constraint equations are established:

[0148]

[0149] in, As carrier and node The Euclidean distance between them (spatial propagation constraint). For a unified time deviation state variable (time state constraint term). To account for observational noise. In this equation, the spatial and temporal terms pass through the same observation. Bound together, inseparable.

[0150] For the Multi-path ( Its delay can be expressed as:

[0151]

[0152] Substituting the coupling expression of S33 into the equation, we obtain the auxiliary constraint equation:

[0153]

[0154] in, For the first The virtual reflection point locations corresponding to the multipath. It should be noted that... Location of the water-based body The geometric parameters of the reflective interface (such as the sea surface or seabed) are uniquely determined and do not depend on the node index. In one specific implementation, for the water surface reflection path, the mirror image point of the carrier with respect to the water surface is taken: (Assume the water surface is) (Plane); for the seabed reflection path, take the mirror point about the seabed plane: ,in The water depth. This auxiliary equation is independent of... It is used to enhance the geometric strength of spatial constraints.

[0155] In one specific implementation, the consistency coupling process refers to simultaneously solving the local constraint equations of the first diameter of all nodes to form a system based on the carrier position. and time deviation state variables For an overdetermined system of equations with common optimization variables. For example, for With nodes, construct the following system of nonlinear equations:

[0156]

[0157] It should be noted that in the equations on the right... These correspond to the position coordinates of each node, consistent with the superscript on the left.

[0158] The characteristic of this system of equations is that all equations share the same... ,and and Appearing simultaneously in each equation, any pair All estimates will be passed It affects all equations, and vice versa. Therefore, and The equations must be jointly estimated within a unified solution framework and cannot be decomposed into independent location subproblems and time synchronization subproblems. This is the technical meaning of "indivisible spatiotemporal joint constraint equations" in this invention.

[0159] Furthermore, at least two communication reference nodes achieve state sharing through a unified time deviation state variable, including:

[0160] S41: The multipath structure observations generated by each communication reference node are mapped to a unified time-bias state variable space, and cross-node observation correlation is achieved through this shared time-bias state variable.

[0161] Specifically, since the local constraint equations of all nodes contain the same When using the observations of node 1 to... When making the estimation, the estimation result will automatically affect nodes 2, 3, ... The equation residuals. In other words, It has become a "bridge" connecting observation information from different nodes, enabling cross-node information sharing. For example, if node 1 has better channel quality, its... The correction will immediately improve the positioning accuracy of other nodes.

[0162] S42: When constructing the spatiotemporal joint constraint equation set, state variable consistency constraints are applied to the local constraint equations of each node, so that the multipath structure observation errors of different nodes are propagated and corrected through the same time deviation state variable. This enables the multipath structure observations of each communication reference node to be jointly solved in a unified time deviation state variable space, rather than being processed independently.

[0163] In a preferred embodiment, state variable consistency constraints are introduced by a unified state variable vector. ( For the clock drift rate (see S32), and enforce the same rate in all local constraint equations. To achieve this, specifically, construct the following global cost function, which consists of two items:

[0164] The first term is the first-path cost term, which reflects the observation error of the first-path propagation delay at each communication reference node:

[0165] ;

[0166] The second term is the multipath auxiliary cost term, reflecting the observation error of the relative time delay structure of multipath at each node:

[0167] ;

[0168] The complete global cost function is then:

[0169]

[0170] in, For nodes Link-state consistency weight (derived from the link-state consistency metric in S11) Export). The time consistency weights for the multipath components (generated by S2) are used. By minimizing this global cost function, the observation errors of all nodes are processed through the same... and Through propagation and joint correction, a truly joint solution was achieved.

[0171] It is important to emphasize that in the above cost function and Simultaneously appearing and In this context, the two cannot be mathematically separated into independent location subproblems and time synchronization subproblems; that is, the optimization update of either variable depends on the current estimate of the other variable.

[0172] It should be noted that, through the above step S4, the present invention combines the coupled observations of multiple communication reference nodes into an indivisible spatiotemporal joint constraint equation set, realizing the physical binding of spatial constraints and time constraints and cross-node state sharing, and providing a complete mathematical model for subsequent synchronous solution of carrier position and time deviation.

[0173] S5. By jointly solving the spatiotemporal constraint equations, the spatial location information and time deviation state estimation results of the underwater vehicle are obtained simultaneously.

[0174] Specifically, the joint solution process includes:

[0175] The spatiotemporal joint constraint equations are transformed into a nonlinear optimization problem with variable constraints, where the optimization variables include both the spatial location variables of the underwater vehicle and the time deviation state variables.

[0176] A unified cost function based on the observation error of multipath structure is constructed, and spatial error and temporal deviation state error are incorporated into the same optimization objective function for joint minimization.

[0177] During the iterative solution process, the spatial location variable and the time deviation state variable are updated simultaneously, so that the update result of any variable depends on the current estimate of the other variable, thus forming a strongly coupled bivariate solution process.

[0178] In one specific implementation, the spatiotemporal joint constraint equations are transformed into a nonlinear optimization problem with variable constraints, specifically including the following steps: The global cost function constructed in S4... As the optimization objective, the optimization variables are... The optimization problem can be expressed as:

[0179]

[0180] in, For the first path cost item, This is the multipath auxiliary cost term (see S42). Since the observation equations are all nonlinear functions, this problem is a nonlinear least squares optimization problem.

[0181] In one specific implementation, a unified cost function based on multipath structure observation errors is constructed, incorporating spatial errors and temporal deviation state errors into the same optimization objective function for joint minimization. Specifically: the above Spatial error has been reflected in Deviation from theoretical value) and time deviation state error (reflected in The deviation from the observed values ​​is incorporated into the same objective function. For example, the Levenberg-Marquardt algorithm is used for iterative solution, with the iterative update formula as follows:

[0182]

[0183] in, The residual vector is formed by stacking the first path residuals and multipath auxiliary residuals of all nodes; Let be the Jacobian matrix of the residual vector with respect to the optimization variables; The damping factor, It is an identity matrix.

[0184] In one specific implementation, the Jacobian matrix The structure reflects the strong coupling relationship between the two variables. For nodes... First diameter residual Its partial derivative is:

[0185] , ,

[0186] For multipath auxiliary residuals Its partial derivatives involve and virtual reflection points (The latter is also) (functions). Specifically:

[0187]

[0188] in, It can be obtained from the mirror relationship using the chain rule (e.g., when reflecting off a water surface). Then this item is ,and ).therefore It includes two contributions that further strengthen the coupling between spatial location and temporal deviation.

[0189] In each iteration, and Update volume and From the same system of linear equations Simultaneously, it was found that the update of any variable depends on the current estimate of the other variable, thus achieving a strongly coupled bivariate solution.

[0190] It is important to emphasize that the time deviation state variable Through residual function and Directly affects the cost function The value of is thus continuously corrected during optimization iterations. Specifically, The estimation error will cause the first path prediction delay for all nodes. This produces a systematic bias, which is reflected in the residuals and expressed through the Jacobian matrix. Dissemination to Updates .therefore, It is a fully observable state variable.

[0191] As a supplement to the aforementioned gradient-based Levenberg-Marquardt solution method, this embodiment provides another preferred implementation: the joint solution of the spatiotemporal joint constraint equations employs a swarm intelligence algorithm for iterative optimization. For example, a particle swarm optimization algorithm is used to optimize the global cost function. Perform a minimization solution:

[0192] First, perform algorithm initialization: set the underwater body spatial location variable. , , and time deviation state variables , Encoded as the position vector of the particle. subscript Indicates particle index, This represents the particle swarm size (typically ranging from 20 to 50). The position of each particle is randomly initialized within a reasonable search space. and speed (Velocity components are initialized to zero or small random values). Each particle maintains its own historical best position. (i.e., the position corresponding to the minimum fitness reached by the particle) and the global optimal position of the population. (That is, the best among all the best positions in the history of all particles).

[0193] Secondly, iterative updates are performed: In each iteration, the velocity and position of each particle are updated iteratively according to the following equation:

[0194]

[0195]

[0196] in, For the number of iterations, For particles In the The velocity vector at the next iteration; This is the inertial weight, typically ranging from 0.5 to 0.9, used to balance global and local searches; and The learning factor (or acceleration constant) typically has a value of 2.0, representing individual cognition and social impact, respectively. , for A random number that is uniformly distributed within an interval. For particles The historical best position; This is the historical best position for the entire particle swarm.

[0197] After the update, calculate the fitness value for each particle's new position. (i.e., the global cost function value), and update and .

[0198] The iteration termination condition is: reaching the preset maximum number of iterations (typically 100-500) or achieving the globally optimal fitness function multiple times consecutively (e.g., 10 times). The relative rate of change is less than a preset threshold (e.g.) ).

[0199] Final solution: After the algorithm converges, the global optimum is:

[0200] That is, the desired optimization variable. The optimal estimate, where The spatial location estimation results of the underwater vehicle. , The results show the time deviation state and drift rate estimation. This swarm intelligence solution method does not require the calculation of the Jacobian matrix, is applicable to discontinuous and non-convex optimization problems, and enhances the reliability of the algorithm in complex underwater environments.

[0201] Furthermore, during the iterative solution process, a reverse constraint mechanism based on multipath structure observations is introduced, including:

[0202] By utilizing the deviation relationship between the multipath propagation delay structure vector and the theoretical propagation time, a reverse correction term for the time deviation state variable is constructed;

[0203] By incorporating the reverse correction term as part of the state update equation, the update of the time-biased state variable depends simultaneously on the historical time state and the current multipath structure observation.

[0204] The temporal state evolution process is thus simultaneously influenced by both time series constraints and spatial propagation constraints, forming a closed-loop correction for time-biased state variables.

[0205] In a preferred embodiment, the specific implementation process of the reverse constraint mechanism is as follows:

[0206] First, define the consistency residual of the multipath structure. The residual reflects the degree of matching between the multipath delay structure and the currently estimated carrier position.

[0207] Then, construct the reverse correction term for the time deviation state variable. The calculation formula is as follows:

[0208]

[0209] in, This is a preset inverse correction gain coefficient. In a preferred embodiment, Adaptive adjustment based on the statistical characteristics of the current observed residuals: ,in For the consistency residual of multipath structure standard deviation The standard deviation of the time-biased state variable. The preset maximum gain coefficient (typically 0.1). In another preferred embodiment, The optimal value is determined based on the minimum mean square error criterion and obtained through offline simulation or online learning. This adaptive mechanism avoids the mismatch problem of fixed gain coefficients when channel conditions change, and improves the stability and engineering applicability of the inverse constraint.

[0210] because It does not have an explicit dependency. ,but The estimate is affected by Therefore, the impact is:

[0211] ;

[0212] in, The sensitivity matrix to time deviation of the position estimate from the previous iteration can be approximated, or it can be given by the cross-covariance matrix in the extended Kalman filter framework. In a simplified implementation, the Jacobian matrix from the previous iteration is used directly for calculation: ,in It indicates a false reversal.

[0213] The reverse correction term is incorporated into the state transition model of the time deviation state variable. Combined with the state transition model of S32, the time deviation prediction after closed-loop correction is obtained:

[0214]

[0215] in, Let $\mathbf{a}$ be the clock drift rate. Correspondingly, the update of the covariance matrix also needs to consider the uncertainty of the inverse correction term. Specifically, the covariance update formula in the extended Kalman filter framework can be used: ,in, Here is the state transition matrix. The process noise covariance matrix is... This is the covariance matrix of the inverse correction term. For example, Can be set to Or, through experience, it can be set as a diagonal matrix. The typical value is Second.

[0216] It should be noted that, through the aforementioned adaptive reverse constraint mechanism, the correction amount of the time deviation state variable no longer depends on fixed empirical parameters, but is dynamically adjusted according to the statistical characteristics of the current observation data. This enables the system to obtain strong closed-loop correction when the channel quality is good, and automatically reduce the correction strength when the channel quality deteriorates, thus avoiding the risk of divergence in erroneous correction.

[0217] Through the above mechanism, the time deviation state variable is not only constrained by the physical characteristics of the clock itself (linear drift + random perturbation), but also by the inverse correction from multipath structure observations, thus forming a... The closed-loop control significantly improves the time synchronization accuracy during long-endurance operation.

[0218] It should be noted that, through the above step S5, the present invention uses a nonlinear optimization method to solve the spatiotemporal joint constraint equations in a bivariate strongly coupled manner, and introduces a reverse constraint mechanism based on a multipath structure. Under a unified optimization framework, the spatial position and time deviation of the underwater vehicle are solved simultaneously, thereby achieving high-precision underwater navigation and positioning.

[0219] In summary, this invention achieves synchronous and strongly coupled solution of underwater body position and clock deviation by directly embedding the time deviation state variable into the time delay expression of multipath structure observations and constructing an indivisible spatiotemporal joint constraint equation set that binds spatial propagation constraints and temporal state constraints through the same multipath observation. This fundamentally overcomes the defect of error accumulation caused by the step-by-step processing of positioning and time synchronization in traditional methods, significantly improving the navigation and positioning accuracy and reliability in complex underwater environments. Furthermore, this invention effectively suppresses the "structural break" anomaly component caused by sudden interference and non-line-of-sight propagation in time-varying underwater channels by performing dual-threshold screening of channel impulse response sequences, constructing multipath structure vectors, and verifying the consistency of continuous time windows. It also introduces time consistency weights to adaptively weight high-quality multipath observations, thus providing robust and structured input data for spatiotemporal joint solution, enabling the navigation system to maintain stable and accurate positioning output even under harsh hydrological conditions. This invention models the time deviation state variable as an evolutionary state containing linear drift and random perturbation, and embeds it as an endogenous variable into the multipath time delay expression. At the same time, it introduces a reverse constraint mechanism based on multipath structure consistency during the iterative solution process, so that the clock deviation update is subject to dual closed-loop correction by historical time series constraints and spatial propagation constraints. This achieves continuous self-calibration of time synchronization accuracy under long flight time conditions and effectively reduces the dependence on high-precision crystal oscillators.

[0220] Example 2, a second embodiment of the present invention, provides a multi-source fusion communication-assisted navigation and positioning system for underwater vehicles, applied to the method of Example 1. This system is deployed in the navigation computer of an underwater vehicle (such as an AUV or ROV) as a hardware module or a combination of hardware and software, and can output the vehicle's spatial position information and time deviation state estimation results in real time and with high accuracy. Specifically, it includes the following modules:

[0221] The communication signal acquisition module is used to acquire the bidirectional communication signals between the underwater body and the communication reference node.

[0222] The multipath structure observation generation module is used to generate multipath structure observations that include multipath propagation delay structure and path energy distribution, and to serve as input for propagation structure constraints.

[0223] The time deviation state modeling module is used to construct time deviation state variables and directly embed them into the time delay expression of multipath structure observations to form a coupled expression relationship.

[0224] The spatiotemporal joint constraint construction module is used to construct a set of spatiotemporal joint constraint equations based on the coupling expression relationships corresponding to different communication reference nodes.

[0225] The joint solution module is used to simultaneously solve the spatial location variables and time deviation state variables of the underwater body within the same optimization framework, and output the spatial location information and time deviation state estimation results.

[0226] It should be noted that the above modules can be deployed in the navigation computer or signal processing unit of the underwater vehicle in the form of software, hardware, or a combination of both. In one specific implementation, the communication signal acquisition module consists of an underwater acoustic transducer, an analog-to-digital converter, and a digital signal processor; the multipath structure observation generation module, the time deviation state modeling module, the spatiotemporal joint constraint construction module, and the joint solution module are integrated into an embedded processor (such as a DSP, FPGA, or ARM), and implement the corresponding functions by executing fixed algorithm code.

[0227] Through the above system, the present invention can output the spatial position information and time deviation state estimation results of the underwater vehicle in real time and with high accuracy, realizing the integration of communication and navigation.

[0228] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention, and all such modifications or substitutions should be covered within the scope of the claims of the present invention.

Claims

1. A multi-source fusion communication-assisted navigation and positioning method for underwater vehicles, characterized in that, Includes the following steps: Acquire bidirectional communication signals between the underwater vehicle and at least two communication reference nodes, wherein the bidirectional communication signals contain a transmission timestamp, a reception timestamp, and a channel impulse response sequence within the same communication cycle; Multipath components are extracted from the channel impulse response sequence to generate multipath structure observations that include multipath propagation delay structure and path energy distribution; Based on the bidirectional communication signal, a time deviation state variable is constructed, and the time deviation state variable is directly embedded into the time delay expression of the multipath structure observation, so that the multipath propagation time delay structure and the time deviation state variable form a coupled expression relationship. A spatiotemporal joint constraint equation set is constructed based on the coupling expression relationship corresponding to different communication reference nodes. The spatiotemporal joint constraint equation set contains both spatial propagation constraint terms and temporal state constraint terms, and the two are bound together through the same multipath structure observation. By jointly solving the spatiotemporal constraint equations, the spatial location information and time deviation state estimation results of the underwater vehicle can be obtained simultaneously.

2. The multi-source fusion communication-assisted navigation and positioning method for underwater vehicles as described in claim 1, characterized in that: Generating the multipath structure observations includes: The channel impulse response sequence is subjected to time-domain matched filtering to obtain a multipath response function, and effective response components that satisfy the dual threshold conditions of amplitude threshold and duration are identified in the multipath response function. A multipath propagation delay structure vector is constructed for the effective response components according to their occurrence time order, and the corresponding path energy distribution vector is generated simultaneously. The consistency of the multipath propagation delay structure vector is checked in a continuous time window. By comparing the multipath propagation delay structure vectors in adjacent time windows, the topological change difference value between the two is obtained. When the topological change difference value exceeds the preset structural continuity constraint threshold, the corresponding response component is marked as a structural fracture component and suppressed. Using the remaining response components after suppression, the multipath propagation delay structure vector and path energy distribution vector are reconstructed, and the reconstructed vectors are used as the final multipath structure observation.

3. The multi-source fusion communication-assisted navigation and positioning method for underwater vehicles as described in claim 1, characterized in that: The construction and embedding process of the time deviation state variable specifically includes: Based on the bidirectional communication signal between the underwater body and the communication reference node, the timestamp mapping relationship between uplink and downlink communication is established respectively, and the round-trip propagation time is expressed as the superposition of the propagation time component and the time deviation state component. The time deviation state component is defined as a state variable that evolves over time, and is described by a state transition model that includes linear drift terms and random disturbance terms. The time deviation state variable is embedded into the time delay expression of the multipath structure observation, so that the multipath structure observation is transformed into a coupled observation that depends on the time deviation state variable. The time deviation state variable does not participate in the localization solution as a separate correction parameter, but participates in the joint solution as an endogenous variable of the spatiotemporal joint constraint equation system.

4. The multi-source fusion communication-assisted navigation and positioning method for underwater vehicles as described in claim 1, characterized in that: The construction process of the spatiotemporal joint constraint equation set specifically includes: For different communication reference nodes, local constraint equations are established based on their respective coupling expression relationships, and a unified time deviation state variable is introduced into each local constraint equation so that the local constraint equations of different nodes share the same time state space. Among them, the spatial propagation constraint term and the time state constraint term in each local constraint equation are bound through the same multipath structure observation, rather than being constructed independently and then spliced ​​together. The local constraint equations of different communication reference nodes are subjected to consistency coupling processing, so that all local constraint equations satisfy the global consistency condition in the same state variable space, thereby forming an indivisible spatiotemporal joint constraint equation set.

5. The multi-source fusion communication-assisted navigation and positioning method for underwater vehicles as described in claim 1, characterized in that: The joint solution process specifically includes: The spatiotemporal joint constraint equations are transformed into a nonlinear optimization problem with variable constraints, wherein the optimization variables include both the spatial location variable of the underwater vehicle and the temporal deviation state variable. A unified cost function based on the observation error of multipath structure is constructed, and spatial error and temporal deviation state error are incorporated into the same optimization objective function for joint minimization. During the iterative solution process, the spatial location variable and the time deviation state variable are updated simultaneously, so that the update result of any variable depends on the current estimate of the other variable, thus forming a strongly coupled bivariate solution process.

6. The multi-source fusion communication-assisted navigation and positioning method for underwater vehicles as described in claim 4, characterized in that: At least two of the aforementioned communication reference nodes achieve state sharing through the unified time deviation state variable, including: The multipath structure observations generated by each communication reference node are mapped to a unified time-bias state variable space, and cross-node observation correlation is achieved through this shared time-bias state variable. When constructing the spatiotemporal joint constraint equation set, state variable consistency constraints are applied to the local constraint equations of each node. This allows the multipath structure observation errors of different nodes to be propagated and corrected through the same time deviation state variable, thereby enabling the multipath structure observations of each communication reference node to be jointly solved in a unified time deviation state variable space, rather than being processed independently.

7. The multi-source fusion communication-assisted navigation and positioning method for underwater vehicles as described in claim 2, characterized in that: The screening process for structural continuity constraints on the observed data of the multipath structure specifically includes: The sequence consistency of the multipath propagation delay structure vectors generated in multiple consecutive communication cycles is compared, and the differences in topological changes between the structure vectors are calculated. When the difference in topological change exceeds the structural continuity constraint threshold, the corresponding multipath response component is marked as a structural fracture component. A time consistency weight is introduced into the multipath structural observations retained after screening by structural continuity constraints, so that they participate in the calculation in a weighted form in the subsequent spatiotemporal joint constraint equations.

8. The multi-source fusion communication-assisted navigation and positioning method for underwater vehicles as described in claim 5, characterized in that: In the iterative solution process, a reverse constraint mechanism based on multipath structure observations is introduced, including: By utilizing the deviation relationship between the multipath propagation delay structure vector and the theoretical propagation time, a reverse correction term for the time deviation state variable is constructed; The reverse correction term is incorporated as part of the state update equation, so that the update of the time-biased state variable depends on both the historical time state and the current multipath structure observation. The temporal state evolution process is thus simultaneously influenced by both time series constraints and spatial propagation constraints, forming a closed-loop correction for time-biased state variables.

9. The multi-source fusion communication-assisted navigation and positioning method for underwater vehicles as described in claim 1, characterized in that: Before acquiring the bidirectional communication signal, the communication reference node undergoes a filtering process based on link state consistency. This filtering process includes: Based on the channel impulse response stability and round-trip propagation time consistency of the communication link, a link state consistency index is constructed. When the link state consistency index is lower than a preset threshold, the corresponding communication reference node is excluded from the spatiotemporal joint constraint equation set. The link selection result not only affects the input data set, but also directly affects the observability structure of the time deviation state variable, thereby forming a binding relationship between the selected set of communication reference nodes and the time deviation state variable space.

10. A multi-source fusion communication-assisted navigation and positioning system for underwater vehicles, applied to the multi-source fusion communication-assisted navigation and positioning method for underwater vehicles as described in any one of claims 1-9, characterized in that, include: The communication signal acquisition module is used to acquire the bidirectional communication signal between the underwater body and the communication reference node; The multipath structure observation generation module is used to generate multipath structure observations that include multipath propagation delay structure and path energy distribution, and to serve as propagation structure constraint inputs. The time deviation state modeling module is used to construct time deviation state variables and directly embed them into the time delay expression of the multipath structure observations to form a coupled expression relationship. The spatiotemporal joint constraint construction module is used to construct a set of spatiotemporal joint constraint equations based on the coupling expression relationships corresponding to different communication reference nodes. The joint solution module is used to simultaneously solve the spatial location variables and time deviation state variables of the underwater body within the same optimization framework, and output the spatial location information and time deviation state estimation results.

Images