Relative navigation method and device based on multi-source information fusion

Abstract

The application provides a relative navigation method and device based on multi-source information fusion, comprising: determining a signal state of a target node; in the case that the signal state is a signal, obtaining a target signal of the target node and node motion characteristics; estimating parameter information related to a relative position of the target node based on the target signal to obtain a signal matrix; determining a motion model according to the signal type and the node motion characteristics, and determining a measurement model according to the signal type; obtaining a fusion observation matrix based on the measurement model; obtaining a position estimation prediction value and an error covariance matrix prediction value of the target node at a target time based on the motion model; and obtaining a position estimation result and an error covariance matrix estimation result of the target node at the target time based on the fusion observation matrix, the position estimation prediction value and the error covariance matrix prediction value. According to the application, the model is adaptively selected according to the type and quantity of real-time observation information to fuse multi-source information, and joint estimation is performed, so that the utilization rate is higher and the positioning accuracy is higher.

Description

Technical Field

[0001] This invention relates to the field of positioning and navigation technology, and in particular to a relative navigation method and apparatus based on multi-source information fusion. Background Technology

[0002] Current mainstream multi-aircraft relative navigation technologies mainly consist of modules such as signal source selection, Kalman filtering, and navigation data extrapolation. The signal source selection module retains only high-quality signal source data based on parameters such as the signal source's location, time, location quality, and geometric positional relationship with the user. The Kalman filtering estimation module filters the selected signal source data, position parameter estimation results, and node motion equations, estimating node positions using the minimum root mean square error as the optimal estimation criterion. The navigation data extrapolation module uses its own inertial navigation system to infer its own trajectory during gaps in signal absence, compensating for the increased positioning error caused by low signal transmission frequencies.

[0003] Existing relative navigation technologies have low utilization rates of wireless signal measurement and inertial navigation sensor information, resulting in poor positioning accuracy. Summary of the Invention

[0004] This invention provides a relative navigation method and apparatus based on multi-source information fusion to address the shortcomings of existing technologies, such as low utilization of wireless signal measurement and inertial navigation sensor information and poor positioning accuracy, thereby achieving relative navigation with higher utilization of wireless signal measurement and inertial navigation sensor information and higher positioning accuracy.

[0005] This invention provides a relative navigation method based on multi-source information fusion, comprising:

[0006] Determine the signal status of the target node;

[0007] When the signal state is signaled, the target signal and node motion characteristics of the target node are acquired; wherein, the node motion characteristics are the motion characteristics of the target node, and the node motion characteristics include at least the type and motion mode of the target node;

[0008] Based on the target signal, parameter information related to the relative position of the target node is estimated to obtain the signal matrix;

[0009] The signal type of the target signal is determined, and a motion model is determined based on the signal type and the node motion characteristics. A measurement model is also determined based on the signal type.

[0010] Based on the signal matrix and the measurement model, a fused observation matrix is ​​obtained;

[0011] Based on the state vector of the target node at the previous moment, the estimation result of the error covariance matrix at the previous moment, and the clock bias, the predicted position and the predicted error covariance matrix of the target node at the target moment are obtained based on the motion model.

[0012] Based on the fused observation matrix, the predicted position value, and the predicted error covariance matrix, the target node's target time position estimation result and error covariance matrix estimation result are obtained.

[0013] According to the present invention, a relative navigation method based on multi-source information fusion estimates parameter information related to the relative position of the target node based on the target signal to obtain a signal matrix, specifically including:

[0014] For a target signal whose signal type is a signal source signal, the three-dimensional coordinates of the signal source, the relative distance measurement value, and the noise variance of the relative distance measurement of the target signal are obtained based on the signal source signal.

[0015] The signal source terminal identifier, three-dimensional coordinates of the signal source, relative distance measurement value, and noise variance of the relative distance measurement for each signal source are stored in a pre-constructed blank matrix to obtain the signal matrix.

[0016] According to the present invention, a relative navigation method based on multi-source information fusion estimates parameter information related to the relative position of the target node based on the target signal to obtain a signal matrix, specifically including:

[0017] For a target signal whose signal type is an inertial navigation measurement signal, the components of the target node's acceleration, velocity, and displacement in the x, y, and z axes, as well as the covariance matrix of the acceleration, velocity, and displacement measurement noise, are obtained based on the inertial navigation measurement signal.

[0018] The components of acceleration, velocity, and displacement in the x, y, and z axes of each inertial navigation measurement signal, as well as the covariance matrix of the acceleration, velocity, and displacement measurement noise, are stored in a pre-constructed blank matrix to obtain the signal matrix.

[0019] According to the relative navigation method based on multi-source information fusion provided by the present invention, a motion model is determined according to the signal type and the node motion characteristics, specifically including:

[0020] If the signal type does not include inertial navigation measurement signals, the motion model is determined to be a uniform motion model or a random walk model based on the node motion characteristics.

[0021] If the signal type includes inertial navigation measurement signals, the motion model is determined to be a uniformly accelerated motion model or the current statistical model based on the node motion characteristics.

[0022] According to the relative navigation method based on multi-source information fusion provided by the present invention, a measurement model is determined according to the type of signal, specifically including:

[0023] If the signal types include inertial navigation measurement signals and signal source signals, the measurement model is determined to be a ranging and inertial navigation measurement model;

[0024] If the signal type is a signal source signal, the measurement model is determined to be a ranging measurement model;

[0025] If the signal type is an inertial navigation measurement signal, the measurement model is determined to be an inertial navigation measurement model.

[0026] According to the relative navigation method based on multi-source information fusion provided by the present invention, a fused observation matrix is ​​obtained based on the signal matrix and the measurement model, specifically including:

[0027] Based on the signal matrix and the measurement model, the signal source measurement observation matrix and the inertial navigation measurement observation matrix are calculated.

[0028] The signal source measurement observation matrix and the inertial navigation measurement observation matrix are fused based on the first preset formula to obtain a fused observation matrix;

[0029] The first preset formula includes:

[0030]

[0031] Among them, H d，k H represents the signal source measurement and observation matrix; m，k H represents the inertial navigation measurement observation matrix; k This represents the fused observation matrix.

[0032] According to the relative navigation method based on multi-source information fusion provided by the present invention, based on the state vector of the target node at the previous moment, the estimation result of the error covariance matrix at the previous moment, and the clock bias, the method obtains the predicted position and the predicted error covariance matrix of the target node at the target moment based on the motion model, specifically including:

[0033] The estimated position and predicted error covariance matrix of the target time are calculated based on the second preset formula.

[0034] The second preset formula includes:

[0035]

[0036]

[0037] Where k represents the target time; Let represent the state vector of the target node at the previous moment; if the clock offset of the target moment is unknown, let b represent the estimated clock offset of the target moment at the previous moment. The elements in the matrix are set to the estimation results of the error covariance matrix at the previous time step; if the target time clock offset is known, b represents the target time clock offset value. Set the last row and last column elements to 0; F represents the estimated position prediction at the target time; k Represents a motion model; F represents k Transpose of; This represents the predicted value of the error covariance matrix at the target time.

[0038] According to a relative navigation method based on multi-source information fusion provided by the present invention, the position estimation result and error covariance matrix estimation result of the target node at the target time are obtained based on the fused observation matrix, the predicted position estimate, and the predicted error covariance matrix, specifically including:

[0039] The position estimation result and error covariance matrix estimation result of the target node at the target time are calculated based on the third preset formula.

[0040] The third preset formula includes:

[0041]

[0042]

[0043]

[0044] Where k represents the target time; K k Indicates Kalman gain; H represents the predicted value of the target time error covariance matrix; k Represents the fused observation matrix; This represents the position estimation result of the target at that time. r represents the state vector of the target node at the previous time step. k Represents the target signal information vector, including the measurement information contained in all target signals received at the target time; I represents the identity matrix; H represents k transpose of; R k The covariance matrix representing the noise; This represents the estimation result of the target time error covariance matrix.

[0045] According to a relative navigation method based on multi-source information fusion provided by the present invention, the signal state of the target node is determined, and then the method further includes:

[0046] When the signal state is no signal, the node motion characteristics are acquired; wherein, the node motion characteristics are the motion characteristics of the target node, and the node motion characteristics include at least the type and motion mode of the target node;

[0047] The measurement model is determined to be a non-measurement model, and the motion model is determined to be a uniform motion model or a random walk model based on the node motion characteristics.

[0048] Based on the state vector of the target node at the previous moment, the estimation result of the error covariance matrix at the previous moment, and the motion model, the predicted position and the predicted error covariance matrix of the target node at the target moment are obtained.

[0049] The position estimation result of the target node at the target time is the position estimation prediction value, and the error covariance matrix estimation result of the target node at the target time is the error covariance matrix prediction value.

[0050] The present invention also provides a relative navigation device based on multi-source information fusion, comprising:

[0051] The determination unit is used to determine the signal state of the target node;

[0052] The acquisition unit is used to acquire the target signal and node motion characteristics of the target node when the signal state is signaled; wherein, the node motion characteristics are the motion characteristics of the target node, and the node motion characteristics include at least the type and motion mode of the target node;

[0053] The parameter unit is used to estimate parameter information related to the relative position of the target node based on the target signal, so as to obtain the signal matrix;

[0054] A model unit is used to determine the signal type of the target signal, determine a motion model based on the signal type and the node motion characteristics, and determine a measurement model based on the signal type.

[0055] A fusion unit is used to obtain a fused observation matrix based on the signal matrix and the measurement model.

[0056] The prediction unit is used to obtain the predicted position value and the predicted error covariance matrix value of the target node at the target time based on the motion model, according to the state vector of the target node at the previous time, the estimation result of the error covariance matrix at the previous time, and the clock bias.

[0057] The result unit is used to obtain the target time position estimation result and the error covariance matrix estimation result of the target node based on the fused observation matrix, the position estimation prediction value and the error covariance matrix prediction value.

[0058] The present invention also provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement the relative navigation method based on multi-source information fusion as described above.

[0059] The present invention also provides a non-transitory computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the relative navigation method based on multi-source information fusion as described above.

[0060] The present invention also provides a computer program product, including a computer program that, when executed by a processor, implements the relative navigation method based on multi-source information fusion as described above.

[0061] This invention provides a relative navigation method and apparatus based on multi-source information fusion. The method involves: determining the signal state of a target node; acquiring the target signal and node motion characteristics of the target node when the signal state is positive; wherein the node motion characteristics are the motion characteristics of the target node, including at least the type and motion mode of the target node; estimating parameter information related to the relative position of the target node based on the target signal to obtain a signal matrix; determining the signal type of the target signal and determining a motion model based on the signal type and the node motion characteristics; determining a measurement model based on the signal type; obtaining a fused observation matrix based on the signal matrix and the measurement model; obtaining the predicted position value and predicted error covariance matrix value of the target node at the target time based on the motion model, using the target node's state vector at the previous moment, the error covariance matrix estimation result at the previous moment, and clock bias; and obtaining the target node's position estimation result and error covariance matrix estimation result at the target time based on the fused observation matrix, the predicted position value, and the predicted error covariance matrix value. This invention can adaptively select a model to fuse multi-source information based on the type and quantity of real-time observation information of each node, and jointly estimate the node position and clock offset. It solves the problem of low performance utilization and poor positioning accuracy of positioning and navigation systems caused by the small number or low frequency of observation information in complex environments, and has good practicality and adaptability. Attached Figure Description

[0062] To more clearly illustrate the technical solutions in this invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of this invention. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0063] Figure 1This is one of the flowcharts illustrating the relative navigation method based on multi-source information fusion provided by the present invention;

[0064] Figure 2 This is the second flowchart of the relative navigation method based on multi-source information fusion provided by the present invention;

[0065] Figure 3 This is a schematic diagram of the structure of the relative navigation device based on multi-source information fusion provided by the present invention;

[0066] Figure 4 This is a schematic diagram of the structure of the electronic device provided by the present invention.

[0067] Figure label:

[0068] 310: Determination unit; 320: Acquisition unit; 330: Parameter unit; 340: Model unit; 350: Fusion unit; 360: Prediction unit; 370: Result unit;

[0069] 410: Processor; 420: Communication interface; 430: Memory; 440: Communication bus. Detailed Implementation

[0070] To make the objectives, technical solutions, and advantages of this invention clearer, the technical solutions of this invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of this invention. All other embodiments obtained by those skilled in the art based on the embodiments of this invention without creative effort are within the scope of protection of this invention.

[0071] Existing relative navigation technologies have low utilization rates of wireless signal measurement and inertial navigation sensor information, resulting in poor positioning accuracy. The main reasons for these problems are: the signal source selection module does not fully utilize the measurement information from poor-quality signal sources; and the Kalman filter estimation module and navigation data extrapolation module do not effectively fuse signal measurement and inertial navigation data, leading to high positioning errors and slow convergence speeds.

[0072] Based on this, the present invention provides a relative navigation method and apparatus based on multi-source information fusion.

[0073] The following is combined Figures 1-2 The present invention describes a relative navigation method based on multi-source information fusion. Figure 1 This is one of the flowcharts illustrating the relative navigation method based on multi-source information fusion provided by this invention, such as... Figure 1 As shown, it includes the following steps:

[0074] Step 110: Determine the signal status of the target node.

[0075] The target node, also known as the receiving node, is the node to be located and navigates and positions itself based on its relative position to the transmitting node. Signal states are divided into two types: with signal and without signal. With signal, the target node can receive target signals, which include the observation signals emitted by the transmitting node and the target node's own inertial navigation measurement signals. Observation signals include the signal source signals. It's important to note that the observation signals and the target node's own inertial navigation measurement signals are not necessarily acquired simultaneously; only the observation signals or only the inertial navigation measurement signals can be acquired. In the without signal state, the target node does not collect any target signals.

[0076] Step 120: When the signal state is signaled, acquire the target signal and node motion characteristics of the target node; wherein, the node motion characteristics are the motion characteristics of the target node, and the node motion characteristics include at least the type and motion mode of the target node.

[0077] When the target node is able to collect the target signal, such as Figure 2 As shown, this invention mainly consists of three parts: 1) target signal collection and parameter estimation, which can collect signals from an indefinite number of signal sources and estimate signal parameters, while simultaneously acquiring inertial navigation measurement data from the node itself; 2) a multi-source information fusion method, which can fuse ranging information and inertial navigation measurement information from an indefinite number of sources to fully extract position information from the measurement signals; and 3) a Kalman filter-based joint position and clock offset estimator, which can reduce clock errors and improve positioning accuracy. This invention is applicable to the fields of positioning and navigation and multi-aircraft formation, and features distributed autonomous decision-making positioning and navigation as well as multi-source information fusion networks.

[0078] First, the target signal and node motion characteristics are acquired. The target signal includes the observation signal emitted by the transmitting node and the inertial navigation measurement signal of the target node itself. Specifically, the number of observation signals is not fixed, and their sources can be diverse. At the target time, a variable number of multi-source observation signals are acquired. Furthermore, the observation signals can be signal source signals. The signal source is a node whose position is known in the absolute or relative coordinate system, or a node whose position has been estimated with high accuracy. The signal source broadcasts its signal at a fixed frequency or randomly, and the target node acquires it during the broadcast process.

[0079] In some embodiments, the collected signal source signal includes at least the identification of the signal source terminal, the estimated result of the signal source position, and the estimation error. Further, the estimated result of the signal source position and the estimation error in the signal source signal can originate from prior information or from the solution results of the absolute or relative positioning method of the signal source terminal. In some embodiments, the estimated result of the signal source position and the estimation error in the signal source signal can be replaced by positioning measurement information; the inertial navigation measurement signal can be directly collected, and the estimated result of the signal source position and the estimation error can be calculated based on the positioning measurement information in the inertial navigation measurement signal. That is, the target signal may also include the inertial navigation measurement signal.

[0080] Inertial navigation measurement signals are signals output by an inertial navigation device after measurement. The inertial navigation device is installed on the target node and is the inertial navigation device itself. In some embodiments, the collected inertial navigation measurement signals include one or more of acceleration measurements, velocity measurements, and displacement measurements.

[0081] To distinguish it from the target node, the node emitting the observation signal is called the transmitting node. For observation signals (signal source type), the transmitting node is the signal source. In some embodiments, the transmitting node can be a node whose position is known in an absolute or relative coordinate system, or a node whose position has been estimated with high accuracy. For example, an aircraft. For target signals (inertial navigation measurement type), the transmitting node is the inertial navigation device.

[0082] Node motion characteristics refer to the motion characteristics of the target node (i.e., the receiving node). Node motion characteristics include the type of the target node and its motion mode. For example, when the target node is an aircraft, the node's motion characteristics include the type of aircraft and the motion mode it is performing. Aircraft types include, for example, fixed-wing aircraft, multi-rotor aircraft, and helicopters. The motion modes the aircraft is performing include, for example, takeoff, cruise, hovering, and landing.

[0083] For ease of explanation, taking an aerial platform as an example, given a three-dimensional Cartesian coordinate system, the state parameters of the target node at time k are modeled as the following 10-dimensional vector:

[0084] [x, y, z, v] x v y v z a x a y a z b] T ,

[0085] In the formula, x, y, and z represent the coordinates of the node to be located (i.e., the target node) on the x, y, and z axes, respectively, and v x v y v zThese represent the components of velocity along the x, y, and z axes, respectively. x a y a z ...

[0086] It should be noted that the relative navigation method based on multi-source information fusion provided in this invention is described using an aerial platform application as an example. However, the application scenarios of the relative navigation method based on multi-source information fusion provided in this invention are not limited to aerial platforms, but also include relative navigation application scenarios such as ground platforms, surface platforms, and underwater platforms. It can also be used in application scenarios requiring relative positioning and navigation, such as paratrooper mobilization and emergency search and rescue, where satellite denial or weak signals are present. This description should not be construed as limiting the invention.

[0087] Step 130: Estimate the parameter information related to the relative position of the target node based on the target signal to obtain the signal matrix.

[0088] The parameter estimation method estimates the parameters related to the relative position of the target node from the target signal. Specifically, the parameter estimation method may include distance estimation based on the received signal strength, distance estimation based on the time of arrival, etc. It should be noted that the parameter estimation method is not limited to the above two methods. In the specific implementation process, the parameter estimation method can be selected as needed, and this invention does not impose any restrictions on it.

[0089] The estimation results of parameter information are uniformly represented in matrix form as a signal matrix.

[0090] Preferably, for the signal source signal, the signal matrix is ​​constructed as follows: An M-row, N-column matrix variable is defined, where M ≥ 1 and N ≥ 6. Each row of this matrix variable stores one signal source signal, with a maximum of M signals. Each row of this matrix variable includes at least N ≥ 6 elements, namely, the signal source terminal's identifier, the signal source's three-dimensional coordinates, the relative distance measurement value, and the noise variance of the relative distance measurement. If the number of received signal source signals at a certain moment is less than M, then all elements in the rows of this matrix variable without messages are set to -1. It should be noted that the signal source terminal's identifier is included in the signal source signal, and the signal source's three-dimensional coordinates, relative distance measurement value, and relative distance measurement noise variance are all estimated from the signal source signal using parameter estimation methods.

[0091] Preferably, for the inertial navigation measurement signal, the signal matrix is ​​constructed as follows: a 9-row, 4-column matrix variable is defined, where the first column stores the components of acceleration, velocity, and displacement along the x, y, and z axes, respectively, and the last three columns store the covariance matrices of the acceleration, velocity, and displacement measurement noise, respectively. If no acceleration measurement is performed at that moment, all elements in rows 1 to 3 of the matrix are set to -1; if no velocity measurement is performed at that moment, all elements in rows 4 to 6 of the matrix are set to -1; if no displacement measurement is performed at that moment, all elements in rows 7 to 9 of the matrix are set to -1; and if no inertial navigation measurement is performed at that moment, all elements in the matrix are set to -1. It should be noted that the components of acceleration, velocity, and displacement along the x, y, and z axes, and the covariance matrices of the acceleration, velocity, and displacement measurement noise in the last three columns, are all calculated based on the inertial navigation measurement signal using the Kalman filtering method.

[0092] Step 140: Determine the signal type of the target signal, and determine the motion model based on the signal type and the node motion characteristics, and determine the measurement model based on the signal type.

[0093] After acquiring the target signal, it is necessary to fuse the information contained in the multi-source observation signals, such as... Figure 2 A multi-dimensional information fusion model is selected for information fusion. This model includes a motion model and a measurement model. Specifically, the motion model is selected based on the node motion characteristics and the available measurement types of the target signal, and the measurement model is selected based on the available measurement types. The fused observation matrix is ​​then calculated.

[0094] The range of motion models to choose from includes, but is not limited to, uniform motion models, uniformly accelerated motion models, random walk models, and current statistical models. Specifically, when inertial navigation is unavailable (i.e., no inertial navigation measurement signal is available), the motion model can be a uniform motion model, a random walk model, etc., with the specific model selected based on the nodal motion characteristics. For example, a uniform motion model can be used when an aircraft is cruising, while a random walk model can be used when flying to avoid obstacles. When inertial navigation is available (i.e., there is an inertial navigation measurement signal is available), the motion model can be a uniformly accelerated motion model, a current statistical model, etc., with the specific model selected based on the nodal motion characteristics. For example, a uniformly accelerated motion model can be used when an aircraft is circling, while a current statistical model can be used when flying to avoid obstacles. In some embodiments, the uniform motion model models the motion of the node as uniform motion; the uniform acceleration motion model models the motion of the node as uniform acceleration; the random walk model models the motion of the node as a random walk model according to a certain probability distribution, wherein the probability distribution can be a Gaussian distribution; the current statistical model models the current probability density of the node acceleration as a modified Rayleigh distribution, the mean of which is the predicted value of the current acceleration.

[0095] It is important to note that, when the signal status is "signal present," the case of "no inertial navigation measurement signal" includes the case where only the signal source signal is present. The case of "with inertial navigation measurement signal" includes the case where only the inertial navigation measurement signal is present, and the case where both the inertial navigation measurement signal and the signal source signal are present. In particular, when the signal status is "no signal," the case of "no inertial navigation measurement signal" also includes the case of "no signal at all."

[0096] The range of measurement models includes ranging and inertial navigation measurement models, ranging measurement models, and inertial navigation measurement models. In some embodiments, the ranging measurement model is a wireless signal arrival timestamp measurement model, the inertial navigation measurement model is a model for directly or indirectly measuring the acceleration, velocity, and displacement of a node, and the ranging and inertial navigation measurement model is the union of the ranging measurement model and the inertial navigation measurement model.

[0097] Preferably, when the signal status is "signal present," if both the signal source signal and the inertial navigation measurement signal are collected simultaneously in step 120 (i.e., the target signal includes both inertial navigation measurement signals and signal source signals), then the ranging and inertial navigation measurement models are selected. If only the signal source signal is collected in step 120 (i.e., the target signal is a signal source signal), then the ranging measurement model is selected. If only inertial navigation measurement information is collected in step 120 (i.e., the signal type is an inertial navigation measurement signal), then the inertial navigation measurement model is selected.

[0098] Specifically, when there is no signal, neither the signal source nor the inertial navigation measurement information is collected. The measurement model is selected as the no-measurement model, and the range of motion models includes uniform motion models, random walk models, etc. The specific model is selected based on the motion characteristics of the node.

[0099] Step 150: Obtain the fused observation matrix based on the measurement model according to the signal matrix.

[0100] For a signal source, the measurement model for the arrival time of the i-th signal at time k is as follows:

[0101]

[0102] Among them, (x k y k , z k (p) represents the three-dimensional coordinates of the receiving node. ix，k p iy，k p iz，k Let ) represent the three-dimensional coordinates of the i-th signal source, and v di，k It measures noise; the noise variance of the relative distance measurement is v. di，kThe equation, assuming that the measurement noises are independent and follow a Gaussian distribution, gives the corresponding observation matrix as:

[0103]

[0104] in, This indicates the relative distance measurement.

[0105] The observation matrix of all signal sources is denoted as H. d，k Its expression is shown below, and it is called the signal source measurement observation matrix:

[0106]

[0107] For inertial navigation measurement signals, the equations for acceleration, velocity, and displacement measurement are as follows:

[0108] [m ax，k m ay，k m az，k ] = [a x，k a y，k a z，k ]+[w ax，k w ay，k w az，k ],

[0109] [m vx，k m vy，k m vz，k ] = [v x，k v y，k v z，k ]+[w vx，k w vy，k w vz，k ],

[0110] [m x，k m y，k m z，k ] = [x k y k z k ]+[w x，k w y，k w z，k ],

[0111] Among them, a x，k a y，k a z，k The acceleration a is represented in sequence. k The components of v along the x, y, and z axes x，k v y，k v z，k The velocity v is represented in sequence. k Components along the x, y, and z axes, x k y k , zk The components of the displacement along the x, y, and z axes are represented in that order, respectively. ax，k w ay，k w az，k [w] represents the covariance matrix of the acceleration measurement noise. vx，k w vy，k w vz，k [w] represents the covariance matrix of the velocity measurement noise. x，k w y，k w z，k Let ] represent the covariance matrix of the displacement measurement noise, and w be the measurement noise. It is assumed that the measurement noises are independent and follow a Gaussian distribution. The observation matrix corresponding to the inertial navigation measurement is as follows, called the inertial navigation measurement observation matrix:

[0112] H m，k =diag([l p l p l p l v l v l v l a l a l a 0])

[0113] In the formula,

[0114] Furthermore, the signal source measurement and observation matrix H d，k and inertial navigation measurement observation matrix H m，k The data are fused into a fused observation matrix according to the first preset formula, which is as follows:

[0115]

[0116] Among them, H d，k H represents the signal source measurement and observation matrix; m，k H represents the inertial navigation measurement observation matrix; k This represents the fused observation matrix. It's important to note that when the signal state is "signal present," the target signal may not simultaneously include both the source signal and the inertial navigation measurement signal. When the target signal at the target time does not include the inertial navigation measurement signal, the inertial navigation measurement observation matrix H... m，k Similarly, when the target signal at the target time does not include the signal source signal, the signal source measurement and observation matrix H is an empty set. d，k It is an empty set. Specifically, in the case of no signal, the signal source measurement and observation matrix H... d，k and inertial navigation measurement observation matrix H m，k All of them are empty sets, meaning that the fused observation matrix is ​​an empty set at this time.

[0117] Step 160: Based on the state vector of the target node at the previous moment, the estimation result of the error covariance matrix at the previous moment, and the clock bias, obtain the predicted position value and the predicted value of the error covariance matrix of the target node at the target moment based on the motion model.

[0118] After fusing information from multi-source observation signals to obtain a fused observation matrix, a joint position and clock offset estimator is used to simultaneously estimate the node position and clock offset. Further, the motion model is first used based on the target node's state vector from the previous time step. Estimation result of the error covariance matrix at the previous time step And clock bias, the estimated predicted value of the target node's position at the target time. Sum of error covariance matrix predictions Make an estimate.

[0119] Step 170: Based on the fused observation matrix, the predicted position value, and the predicted error covariance matrix, obtain the position estimation result and the error covariance matrix estimation result of the target node at the target time.

[0120] Estimating the predicted value using the obtained target node's target time position. Sum of error covariance matrix predictions and the fused observation matrix H k Calculate the Kalman gain K k Then, the position estimation result of the target at that time is calculated. Estimation results of the error covariance matrix

[0121] The position and clock offset joint estimator of this invention can reduce clock errors, improve ranging accuracy, and thus improve the estimation accuracy of node positions.

[0122] Based on the above embodiments, in this method, according to the state vector of the target node at the previous moment, the estimation result of the error covariance matrix at the previous moment, and the clock bias, the predicted position and error covariance matrix of the target node at the target moment are obtained based on the motion model, specifically including:

[0123] The estimated position and predicted error covariance matrix of the target time are calculated based on the second preset formula.

[0124] The second preset formula includes:

[0125]

[0126]

[0127] Where k represents the target time; Let represent the state vector of the target node at the previous moment; if the clock offset of the target moment is unknown, let b represent the estimated clock offset of the target moment at the previous moment. The elements in the matrix are set to the estimation results of the error covariance matrix at the previous time step; if the target time clock offset is known, b represents the target time clock offset value. Set the last row and last column elements to 0; F represents the estimated position prediction at the target time; k Represents a motion model; F represents k Transpose of; This represents the predicted value of the error covariance matrix at the target time.

[0128] Specifically, a joint position and clock bias estimator is used to simultaneously estimate the node position and clock bias. The node state vector at the previous time step is denoted as... The result of the error covariance matrix estimation is denoted as In the absence of a known clock bias, The last element b is set to the clock bias estimate of the previous time step. The elements in the array are set to the error covariance matrix estimation results from the previous time step. Given the clock bias, the following is used: The last element b is set to the known clock offset, and... Set the last row and last column elements to 0.

[0129] Based on the determined motion model F k It can calculate the predicted value of the target's position at a given time. The predicted values ​​of the sum of error covariance matrix

[0130] Based on the above embodiments, the method obtains the target node's target time position estimation result and error covariance matrix estimation result according to the fused observation matrix, the predicted position estimate, and the predicted error covariance matrix, specifically including:

[0131] The position estimation result and error covariance matrix estimation result of the target node at the target time are calculated based on the third preset formula.

[0132] The third preset formula includes:

[0133]

[0134]

[0135]

[0136] Where k represents the target time; Kk Indicates Kalman gain; H represents the predicted value of the target time error covariance matrix; k Represents the fused observation matrix; This represents the position estimation result of the target at that time. r represents the state vector of the target node at the previous time step. k The target signal information vector represents the measurement information contained in all target signals received at the target time, including ranging information of the signal source signal, inertial navigation system (INS) measurement of acceleration, INS measurement of velocity, and INS measurement of displacement; I represents the identity matrix. H represents k transpose of; R k The covariance matrix representing the noise; This represents the estimation result of the target time error covariance matrix.

[0137] Specifically, based on the calculated fusion observation matrix H k Kalman gain can be calculated. Then calculate the target's position estimation result at that time. Estimation results of the error covariance matrix at the target time The clock bias estimation result for the node is The last element of the equation gives the estimated node position as follows: The first three elements correspond to the three-dimensional coordinates of the node.

[0138] The relative navigation method based on multi-source information fusion provided by this invention can filter and estimate signals from an indefinite number of signal sources, and can simultaneously estimate position and clock offset. Specifically, at each moment, if there are no new signal sources, the current position is estimated based on inertial navigation measurements and motion equations; if there are new signal sources, regardless of the number of messages, they can be fused and filtered with inertial navigation measurements and motion equations. If there is no inertial navigation information in the above process, the position is calculated solely based on the motion equations and signal source signals.

[0139] Based on the above embodiments, the method, after determining the signal state of the target node, further includes:

[0140] When the signal state is no signal, the node motion characteristics are acquired; wherein, the node motion characteristics are the motion characteristics of the target node, and the node motion characteristics include at least the type and motion mode of the target node;

[0141] The measurement model is determined to be a non-measurement model, and the motion model is determined to be a uniform motion model or a random walk model based on the node motion characteristics.

[0142] Based on the state vector of the target node at the previous moment, the estimation result of the error covariance matrix at the previous moment, and the motion model, the predicted position and error covariance matrix of the target node at the target moment are obtained.

[0143] The position estimation result of the target node at the target time is the position estimation prediction value, and the error covariance matrix estimation result of the target node at the target time is the error covariance matrix prediction value.

[0144] Specifically, when the signal state is no signal, the target node does not collect any target signal, and the node motion characteristics of the target node are acquired. Node motion characteristics include the type of target node and its motion mode. For example, when the target node is an aircraft, the node's motion characteristics include the type of aircraft and the aircraft's current motion mode. Aircraft types include, for example, fixed-wing aircraft, multi-rotor aircraft, and helicopters. The aircraft's current motion mode includes, for example, takeoff, cruise, hovering, and landing.

[0145] When the signal state is no signal, the signal matrix is ​​an empty set. The motion model is determined based on the node motion characteristics. In actual operation, no signal corresponds to no inertial navigation measurement signal. The motion model can be a uniform motion model, a random walk model, etc. The specific model is selected based on the node motion characteristics. For example, a uniform motion model can be used when the aircraft is cruising, while a random walk model can be used when flying to avoid obstacles. The measurement model is directly determined as a no-measurement model, which means there is neither a signal source measurement nor inertial navigation measurement; the corresponding observation matrix is ​​an empty set.

[0146] When the signal state is no signal, the signal source measurement and observation matrix H d，k and inertial navigation measurement observation matrix H m，k All are empty sets, and the fused observation matrix is ​​also an empty set.

[0147] At this point, based on the target node's state vector from the previous moment, the error covariance matrix estimation result from the previous moment, and the motion model, the predicted position and error covariance matrix of the target node at the target moment are obtained. Specifically, the node's state vector from the previous moment is denoted as... The result of the error covariance matrix estimation is denoted as In the absence of a known clock bias, The last element b is set to the clock bias estimate of the previous time step. The elements in the array are set to the error covariance matrix estimation results from the previous time step. Given the clock bias, the following is used: The last element b is set to the known clock offset, and... Set the last row and last column elements to 0.

[0148] Based on the determined motion model F k It can calculate the predicted value of the target's position at a given time. The predicted values ​​of the sum of error covariance matrix

[0149] The target's position estimation result at that time is: The estimation result of the error covariance matrix at the target time is as follows

[0150] The relative navigation method based on multi-source information fusion provided by this invention can solve the problem of poor performance of positioning and navigation systems when the number of observations is small or the frequency of observations is low, based on real-time observation information from each node. It has good practicality and adaptability. Furthermore, this invention can fuse a variable number of signal source messages and inertial navigation measurement information, and can update the estimated values ​​of target position and clock offset in real time, possessing efficient and high-precision relative navigation capabilities. It can effectively solve the problem of poor performance of positioning and navigation systems in complex environments due to the small number of signal source messages, low transmission frequency, or instability.

[0151] It should be noted that the relative navigation method based on multi-source information fusion provided in this invention is described using an aerial platform application as an example. However, the application scenarios of the relative navigation method based on multi-source information fusion provided in this invention are not limited to aerial platforms, but also include relative navigation application scenarios such as ground platforms, surface platforms, and underwater platforms. It can also be used in application scenarios requiring relative positioning and navigation, such as paratrooper mobilization and emergency search and rescue, where satellite denial or weak signals are present. This description should not be construed as limiting the invention.

[0152] This invention provides a relative navigation method based on multi-source information fusion. The method involves: determining the signal state of a target node; acquiring the target signal and node motion characteristics of the target node when the signal state is positive; wherein the node motion characteristics are the motion characteristics of the target node, including at least the type and motion mode of the target node; estimating parameter information related to the relative position of the target node based on the target signal to obtain a signal matrix; determining the signal type of the target signal and determining a motion model based on the signal type and the node motion characteristics; determining a measurement model based on the signal type; obtaining a fused observation matrix based on the signal matrix and the measurement model; obtaining the predicted position value and predicted error covariance matrix value of the target node at the target time based on the motion model, using the target node's state vector at the previous time step, the error covariance matrix estimation result at the previous time step, and clock bias; and obtaining the target node's position estimation result and error covariance matrix estimation result at the target time based on the fused observation matrix, the predicted position value, and the predicted error covariance matrix value. This invention can adaptively select a model to fuse multi-source information based on the type and quantity of real-time observation information of each node, and jointly estimate the node position and clock offset. It solves the problem of low performance utilization and poor positioning accuracy of positioning and navigation systems caused by the small number or low frequency of observation information in complex environments, and has good practicality and adaptability.

[0153] The relative navigation device based on multi-source information fusion provided by the present invention will be described below. The relative navigation device based on multi-source information fusion described below and the relative navigation method based on multi-source information fusion described above can be referred to in correspondence.

[0154] Figure 3 This is a schematic diagram of the relative navigation device based on multi-source information fusion provided by the present invention, as shown below. Figure 3 As shown, it includes a determination unit 310, an acquisition unit 320, a parameter unit 330, a model unit 340, a fusion unit 350, a prediction unit 360, and a result unit 370, wherein,

[0155] Determining unit 310 is used to determine the signal state of the target node;

[0156] The acquisition unit 320 is used to acquire the target signal and node motion characteristics of the target node when the signal state is signaled; wherein, the node motion characteristics are the motion characteristics of the target node, and the node motion characteristics include at least the type and motion mode of the target node;

[0157] The parameter unit 330 is used to estimate parameter information related to the relative position of the target node based on the target signal, so as to obtain a signal matrix;

[0158] Model unit 340 is used to determine the signal type of the target signal, determine a motion model based on the signal type and the node motion characteristics, and determine a measurement model based on the signal type.

[0159] Fusion unit 350 is used to obtain a fused observation matrix based on the signal matrix and the measurement model;

[0160] The prediction unit 360 is used to obtain the predicted position value and the predicted error covariance matrix value of the target node at the target time based on the motion model, according to the state vector of the target node at the previous time, the estimation result of the error covariance matrix at the previous time, and the clock bias.

[0161] The result unit 370 is used to obtain the target time position estimation result and the error covariance matrix estimation result of the target node based on the fused observation matrix, the position estimation prediction value and the error covariance matrix prediction value.

[0162] Based on the above embodiments, in this device, parameter information related to the relative position of the target node is estimated based on the target signal to obtain a signal matrix, specifically including:

[0163] For a target signal whose signal type is a signal source signal, the three-dimensional coordinates of the signal source, the relative distance measurement value, and the noise variance of the relative distance measurement of the target signal are obtained based on the signal source signal.

[0164] The signal source terminal identifier, three-dimensional coordinates of the signal source, relative distance measurement value, and noise variance of the relative distance measurement for each signal source are stored in a pre-constructed blank matrix to obtain the signal matrix.

[0165] Based on the above embodiments, in this device, parameter information related to the relative position of the target node is estimated based on the target signal to obtain a signal matrix, specifically including:

[0166] For a target signal whose signal type is an inertial navigation measurement signal, the components of the target node's acceleration, velocity, and displacement in the x, y, and z axes, as well as the covariance matrix of the acceleration, velocity, and displacement measurement noise, are obtained based on the inertial navigation measurement signal.

[0167] The components of acceleration, velocity, and displacement in the x, y, and z axes of each inertial navigation measurement signal, as well as the covariance matrix of the acceleration, velocity, and displacement measurement noise, are stored in a pre-constructed blank matrix to obtain the signal matrix.

[0168] Based on the above embodiments, in this device, determining the motion model according to the signal type and the node motion characteristics specifically includes:

[0169] If the signal type does not include inertial navigation measurement signals, the motion model is determined to be a uniform motion model or a random walk model based on the node motion characteristics.

[0170] If the signal type includes inertial navigation measurement signals, the motion model is determined to be a uniformly accelerated motion model or the current statistical model based on the node motion characteristics.

[0171] Based on the above embodiments, in this device, determining the measurement model according to the signal type specifically includes:

[0172] If the signal types include inertial navigation measurement signals and signal source signals, the measurement model is determined to be a ranging and inertial navigation measurement model;

[0173] If the signal type is a signal source signal, the measurement model is determined to be a ranging measurement model;

[0174] If the signal type is an inertial navigation measurement signal, the measurement model is determined to be an inertial navigation measurement model.

[0175] Based on the above embodiments, in this device, obtaining the fused observation matrix based on the measurement model according to the signal matrix specifically includes:

[0176] Based on the signal matrix and the measurement model, the signal source measurement observation matrix and the inertial navigation measurement observation matrix are calculated.

[0177] The signal source measurement observation matrix and the inertial navigation measurement observation matrix are fused based on the first preset formula to obtain a fused observation matrix;

[0178] The first preset formula includes:

[0179]

[0180] Among them, H d，k H represents the signal source measurement and observation matrix; m，k H represents the inertial navigation measurement observation matrix; k This represents the fused observation matrix.

[0181] Based on the above embodiments, in this device, according to the state vector of the target node at the previous moment, the estimation result of the error covariance matrix at the previous moment, and the clock bias, the predicted position value and the predicted value of the error covariance matrix of the target node at the target moment are obtained based on the motion model, specifically including:

[0182] The estimated position and predicted error covariance matrix of the target time are calculated based on the second preset formula.

[0183] The second preset formula includes:

[0184]

[0185]

[0186] Where k represents the target time; Let represent the state vector of the target node at the previous moment; if the clock offset of the target moment is unknown, let b represent the estimated clock offset of the target moment at the previous moment. The elements in the matrix are set to the estimation results of the error covariance matrix at the previous time step; if the target time clock offset is known, b represents the target time clock offset value. Set the last row and last column elements to 0; F represents the estimated position prediction at the target time; k Represents a motion model; F represents k Transpose of; This represents the predicted value of the error covariance matrix at the target time.

[0187] Based on the above embodiments, in this device, the position estimation result and error covariance matrix estimation result of the target node at the target time are obtained according to the fused observation matrix, the predicted position estimate, and the predicted error covariance matrix, specifically including:

[0188] The position estimation result and error covariance matrix estimation result of the target node at the target time are calculated based on the third preset formula.

[0189] The third preset formula includes:

[0190]

[0191]

[0192]

[0193] Where k represents the target time; K k Indicates Kalman gain; H represents the predicted value of the target time error covariance matrix; k Represents the fused observation matrix; This represents the position estimation result of the target at that time. r represents the state vector of the target node at the previous time step. k Represents the target signal information vector, including the measurement information contained in all target signals received at the target time; I represents the identity matrix; Hk represents the transpose of Hk; Rk represents the covariance matrix of the noise. This represents the estimation result of the target time error covariance matrix.

[0194] Based on the above embodiments, the device, after determining the signal state of the target node, further includes:

[0195] When the signal state is no signal, the node motion characteristics are acquired; wherein, the node motion characteristics are the motion characteristics of the target node, and the node motion characteristics include at least the type and motion mode of the target node;

[0196] The measurement model is determined to be a non-measurement model, and the motion model is determined to be a uniform motion model or a random walk model based on the node motion characteristics.

[0197] Based on the state vector of the target node at the previous moment, the estimation result of the error covariance matrix at the previous moment, and the motion model, the predicted position and the predicted error covariance matrix of the target node at the target moment are obtained.

[0198] The position estimation result of the target node at the target time is the position estimation prediction value, and the error covariance matrix estimation result of the target node at the target time is the error covariance matrix prediction value.

[0199] This invention provides a relative navigation method and apparatus based on multi-source information fusion. The method involves: determining the signal state of a target node; acquiring the target signal and node motion characteristics of the target node when the signal state is positive; wherein the node motion characteristics are the motion characteristics of the target node, including at least the type and motion mode of the target node; estimating parameter information related to the relative position of the target node based on the target signal to obtain a signal matrix; determining the signal type of the target signal and determining a motion model based on the signal type and the node motion characteristics; determining a measurement model based on the signal type; obtaining a fused observation matrix based on the signal matrix and the measurement model; obtaining the predicted position value and predicted error covariance matrix value of the target node at the target time based on the motion model, using the target node's state vector at the previous moment, the error covariance matrix estimation result at the previous moment, and clock bias; and obtaining the target node's position estimation result and error covariance matrix estimation result at the target time based on the fused observation matrix, the predicted position value, and the predicted error covariance matrix value. This invention can adaptively select a model to fuse multi-source information based on the type and quantity of real-time observation information of each node, and jointly estimate the node position and clock offset. It solves the problem of low performance utilization and poor positioning accuracy of positioning and navigation systems caused by the small number or low frequency of observation information in complex environments, and has good practicality and adaptability.

[0200] Figure 4 An example is a schematic diagram of the physical structure of an electronic device, such as... Figure 4As shown, the electronic device may include: a processor 410, a communication interface 420, a memory 430, and a communication bus 440, wherein the processor 410, the communication interface 420, and the memory 430 communicate with each other through the communication bus 440. The processor 410 can call logic instructions in the memory 430 to execute a relative navigation method based on multi-source information fusion. This method includes: determining the signal state of a target node; if the signal state is signal-enabled, acquiring the target signal and node motion characteristics of the target node; wherein the node motion characteristics are the motion characteristics of the target node, and the node motion characteristics include at least the type and motion mode of the target node; estimating parameter information related to the relative position of the target node based on the target signal to obtain a signal matrix; determining the signal type of the target signal, and determining a motion model based on the signal type and the node motion characteristics, and determining a measurement model based on the signal type; obtaining a fused observation matrix based on the signal matrix and the measurement model; obtaining a predicted position value and a predicted error covariance matrix value of the target node at the target time based on the motion model, according to the state vector of the target node at the previous time, the error covariance matrix estimation result at the previous time, and the clock bias; and obtaining a predicted position value and an estimated error covariance matrix value of the target node at the target time based on the fused observation matrix, the predicted position value, and the predicted error covariance matrix value.

[0201] Furthermore, the logical instructions in the aforementioned memory 430 can be implemented as software functional units and, when sold or used as independent products, can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, essentially, or the part that contributes to the prior art, or a part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0202] On the other hand, the present invention also provides a computer program product, which includes a computer program that can be stored on a non-transitory computer-readable storage medium. When the computer program is executed by a processor, the computer is able to execute the relative navigation method based on multi-source information fusion provided by the above methods. The method includes: determining the signal state of a target node; when the signal state is that there is a signal, acquiring the target signal and node motion characteristics of the target node; wherein the node motion characteristics are the motion characteristics of the target node, and the node motion characteristics include at least the type and motion mode of the target node; estimating the relative position with the target node based on the target signal. The relevant parameter information is set to obtain a signal matrix; the signal type of the target signal is determined, and a motion model is determined based on the signal type and the node motion characteristics; a measurement model is determined based on the signal type; a fused observation matrix is ​​obtained based on the signal matrix and the measurement model; based on the state vector of the target node at the previous time step, the error covariance matrix estimation result at the previous time step, and the clock bias, the predicted position value and the predicted error covariance matrix value of the target node at the target time are obtained based on the motion model; the predicted position value and the predicted error covariance matrix value of the target node at the target time are obtained based on the fused observation matrix, the predicted position value, and the predicted error covariance matrix value.

[0203] In another aspect, the present invention also provides a non-transitory computer-readable storage medium storing a computer program thereon, which, when executed by a processor, implements the relative navigation method based on multi-source information fusion provided by the methods described above. This method includes: determining the signal state of a target node; when the signal state is signal-present, acquiring the target signal and node motion characteristics of the target node; wherein the node motion characteristics are the motion characteristics of the target node, and the node motion characteristics include at least the type and motion mode of the target node; estimating parameter information related to the relative position of the target node based on the target signal to obtain a signal matrix; determining the signal type of the target signal, and determining a motion model based on the signal type and the node motion characteristics, and determining a measurement model based on the signal type; obtaining a fused observation matrix based on the signal matrix and the measurement model; obtaining a predicted position value and a predicted error covariance matrix value of the target node at the target time based on the motion model, according to the state vector of the target node at the previous time, the error covariance matrix estimation result at the previous time, and the clock bias; and obtaining a predicted position value and an estimated error covariance matrix value of the target node at the target time based on the fused observation matrix, the predicted position value, and the predicted error covariance matrix value.

[0204] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs. Those skilled in the art can understand and implement this without any creative effort.

[0205] Through the above description of the embodiments, those skilled in the art can clearly understand that each embodiment can be implemented by means of software plus necessary general-purpose hardware platforms, and of course, it can also be implemented by hardware. Based on this understanding, the above technical solutions, in essence or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute the methods described in the various embodiments or some parts of the embodiments.

[0206] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A relative navigation method based on multi-source information fusion, characterized in that, include: Determine the signal status of the target node; When the signal state is signaled, the target signal and node motion characteristics of the target node are acquired; wherein, the node motion characteristics are the motion characteristics of the target node, and the node motion characteristics include at least the type and motion mode of the target node; Based on the target signal, parameter information related to the relative position of the target node is estimated to obtain the signal matrix; The signal type of the target signal is determined, and a motion model is determined based on the signal type and the node motion characteristics. A measurement model is also determined based on the signal type. Based on the signal matrix and the measurement model, a fused observation matrix is ​​obtained; Based on the state vector of the target node at the previous moment, the estimation result of the error covariance matrix at the previous moment, and the clock bias, the predicted position and the predicted error covariance matrix of the target node at the target moment are obtained based on the motion model. Based on the fused observation matrix, the predicted position value, and the predicted error covariance matrix, the position estimation result and the error covariance matrix estimation result of the target node at the target time are obtained; Wherein, the clock offset is the estimated value of the clock offset at the previous time when the clock offset at the target time is unknown, and the clock offset at the target time is the clock offset value at the target time when the clock offset at the target time is known; when the clock offset at the target time is known, the last row and last column elements of the estimation result of the error covariance matrix at the previous time are set to 0; After determining the signal state of the target node, the method further includes: acquiring the node motion characteristics when the signal state is no signal; wherein, the node motion characteristics are the motion characteristics of the target node, and the node motion characteristics include at least the type and motion mode of the target node; determining that the measurement model is a measurement-free model, and determining the motion model as a uniform motion model or a random walk model based on the node motion characteristics; obtaining the target node's position estimate and error covariance matrix prediction value at the target time based on the target node's state vector at the previous time, the error covariance matrix estimation result at the previous time, and the motion model; the target node's position estimate result at the target time is the position estimate prediction value, and the target node's error covariance matrix estimation result at the target time is the error covariance matrix prediction value.

2. The relative navigation method based on multi-source information fusion according to claim 1, characterized in that, Based on the target signal, parameter information related to the relative position of the target node is estimated to obtain a signal matrix, specifically including: For a target signal whose signal type is a signal source signal, the three-dimensional coordinates of the signal source, the relative distance measurement value, and the noise variance of the relative distance measurement of the target signal are obtained based on the signal source signal. The signal source terminal identifier, three-dimensional coordinates of the signal source, relative distance measurement value, and noise variance of the relative distance measurement for each signal source are stored in a pre-constructed blank matrix to obtain the signal matrix.

3. The relative navigation method based on multi-source information fusion according to claim 1, characterized in that, Based on the target signal, parameter information related to the relative position of the target node is estimated to obtain a signal matrix, specifically including: For a target signal whose signal type is an inertial navigation measurement signal, the components of the target node's acceleration, velocity, and displacement in the x, y, and z axes, as well as the covariance matrix of the acceleration, velocity, and displacement measurement noise, are obtained based on the inertial navigation measurement signal. The components of acceleration, velocity, and displacement in the x, y, and z axes of each inertial navigation measurement signal, as well as the covariance matrix of the acceleration, velocity, and displacement measurement noise, are stored in a pre-constructed blank matrix to obtain the signal matrix.

4. The relative navigation method based on multi-source information fusion according to any one of claims 1-3, characterized in that, The motion model is determined based on the signal type and the node motion characteristics, specifically including: If the signal type does not include inertial navigation measurement signals, the motion model is determined to be a uniform motion model or a random walk model based on the node motion characteristics. If the signal type includes inertial navigation measurement signals, the motion model is determined to be a uniformly accelerated motion model or the current statistical model based on the node motion characteristics.

5. The relative navigation method based on multi-source information fusion according to any one of claims 1-3, characterized in that, The measurement model is determined based on the type of signal, specifically including: If the signal types include inertial navigation measurement signals and signal source signals, the measurement model is determined to be a ranging and inertial navigation measurement model; If the signal type is a signal source signal, the measurement model is determined to be a ranging measurement model; If the signal type is an inertial navigation measurement signal, the measurement model is determined to be an inertial navigation measurement model.

6. The relative navigation method based on multi-source information fusion according to claim 5, characterized in that, The fused observation matrix is ​​obtained based on the signal matrix and the measurement model, specifically including: Based on the signal matrix and the measurement model, the signal source measurement observation matrix and the inertial navigation measurement observation matrix are calculated. The signal source measurement observation matrix and the inertial navigation measurement observation matrix are fused based on the first preset formula to obtain a fused observation matrix; The first preset formula includes: ; in, Represents the signal source measurement and observation matrix; Represents the inertial navigation measurement observation matrix; Represents the fused observation matrix. Indicates the target time.

7. The relative navigation method based on multi-source information fusion according to claim 1, characterized in that, Based on the target node's state vector from the previous moment, the error covariance matrix estimation result from the previous moment, and the clock bias, the predicted position and error covariance matrix of the target node at the target moment are obtained based on the motion model, specifically including: The estimated position and predicted error covariance matrix of the target time are calculated based on the second preset formula. The second preset formula includes: ， ， in, Indicates the target time; , represents the state vector of the target node at the previous moment; These represent the target nodes at... , and The coordinates of the axis; They represent speeds at... , and The components of the axis; They represent acceleration at , and The components of the axis; if the target clock offset is unknown, This represents the estimated clock offset value at the time preceding the target time. The elements in the matrix are set to the estimation results of the error covariance matrix at the previous time step; if the clock offset of the target time step is known, Indicates the target clock offset. Set the last row and last column elements to 0; This represents the estimated position prediction of the target at that time. Represents a motion model; express transpose; This represents the predicted value of the error covariance matrix at the target time.

8. The relative navigation method based on multi-source information fusion according to claim 1, characterized in that, Based on the fused observation matrix, the predicted position value, and the predicted error covariance matrix, the target node's target time position estimation result and error covariance matrix estimation result are obtained, specifically including: The position estimation result and error covariance matrix estimation result of the target node at the target time are calculated based on the third preset formula. The third preset formula includes: ， ， ， in, Indicates Kalman gain; This represents the predicted value of the error covariance matrix at the target time. Represents the fused observation matrix; This represents the position estimation result of the target at that time. This represents the state vector of the target node at the previous moment; It represents the target signal information vector, including the measurement information contained in all target signals received at the target time; Represents the identity matrix; express transpose; The covariance matrix representing the noise; This represents the estimation result of the target time error covariance matrix.

9. A relative navigation device based on multi-source information fusion, characterized in that, include: The determination unit is used to determine the signal state of the target node; The acquisition unit is used to acquire the target signal and node motion characteristics of the target node when the signal state is signaled; wherein, the node motion characteristics are the motion characteristics of the target node, and the node motion characteristics include at least the type and motion mode of the target node; The parameter unit is used to estimate parameter information related to the relative position of the target node based on the target signal, so as to obtain the signal matrix; A model unit is used to determine the signal type of the target signal, determine a motion model based on the signal type and the node motion characteristics, and determine a measurement model based on the signal type. A fusion unit is used to obtain a fused observation matrix based on the signal matrix and the measurement model. The prediction unit is used to obtain the predicted position value and the predicted error covariance matrix value of the target node at the target time based on the motion model, according to the state vector of the target node at the previous time, the estimation result of the error covariance matrix at the previous time, and the clock bias. The result unit is used to obtain the target time position estimation result and the error covariance matrix estimation result of the target node based on the fused observation matrix, the position estimation prediction value and the error covariance matrix prediction value; Wherein, the clock offset is the estimated value of the clock offset at the previous time when the clock offset at the target time is unknown, and the clock offset at the target time is the clock offset value at the target time when the clock offset at the target time is known; when the clock offset at the target time is known, the last row and last column elements of the estimation result of the error covariance matrix at the previous time are set to 0; After determining the signal state of the target node, the relative navigation device based on multi-source information fusion is used to: acquire node motion characteristics when the signal state is no signal; wherein, the node motion characteristics are the motion characteristics of the target node, and the node motion characteristics include at least the type and motion mode of the target node; determine that the measurement model is a measurement-free model, and determine the motion model as a uniform motion model or a random walk model based on the node motion characteristics; obtain the target node's position estimate and error covariance matrix prediction value at the target time based on the target node's state vector at the previous moment, the error covariance matrix estimation result at the previous moment, and the motion model; the target node's position estimate at the target time is the position estimate prediction value, and the target node's error covariance matrix estimation result at the target time is the error covariance matrix prediction value.

Images