Node optimization method for cluster collaborative navigation

Abstract

The invention discloses a cluster collaborative navigation node optimization method based on bionic neighborhood perception. The cluster collaborative navigation node optimization method mainly aims at cluster dynamic collaborative positioning composed of low-cost multiple carriers. A comprehensive node evaluation model matrix is established, a multi-index fused comprehensive node evaluation model is established by using indexes such as geometric distribution, information uncertainty, jump stability and collaborative signal quality of cluster collaborative nodes, fish swarm gathering and preferred foraging behaviors are enlightened, a bionic enhancement operator is established, and the optimal foraging behavior is evaluated. A distance attenuation factor and a signal fitness factor are introduced on the basis of entropy weight multi-index scoring, and the robustness and global adaptability of a node optimization algorithm are enhanced, so that a low-cost and high-efficiency cluster collaborative navigation system node optimization robustness framework in a dynamic environment is realized.

Description

Technical Field

[0001] This invention belongs to the field of cooperative navigation technology, specifically relating to a node selection method for cluster cooperative navigation based on a comprehensive node evaluation model and biomimetic enhancement. Background Technology

[0002] In recent years, unmanned surface vessels (USVs) have received widespread attention in both military and civilian fields, such as military reconnaissance, environmental monitoring, material transport, disaster relief, and fishing. USV swarm technology, due to its high efficiency and flexibility, has become a hot research topic. During collaborative operations by USV swarms, the reliability and real-time performance of navigation and positioning directly affect the safety and efficiency of mission execution.

[0003] Due to the complexity and diversity of the marine environment and the tasks being performed, the expansion of the swarm size and the frequent changes in topology have dramatically increased the communication and computational burden on the system. If all nodes maintain a fully connected communication mode, it will not only consume limited communication bandwidth but also cause redundant information transmission, reducing system response speed. Simultaneously, some nodes, due to environmental interference, sensor accuracy limitations, or signal attenuation, have significant errors in their positioning information. If this information is adopted indiscriminately, it will inevitably negatively impact the accuracy of cooperative navigation. Most existing cooperative navigation methods are based on global information sharing, assuming that the pose or measurement data provided by each node is reliable, but lack an effective mechanism for evaluating the information value and reliability of different nodes. This approach is acceptable when the number of nodes is limited, but it is difficult to meet the dual requirements of accuracy and real-time performance in dynamic topology and large-scale unmanned surface vessel (USV) swarm environments. To reduce system overhead while ensuring navigation accuracy, there is an urgent need for a node selection method that can dynamically select communication and cooperation partners based on node information contribution, reliability, and environmental conditions, thereby achieving efficient support for the cooperative navigation process of USV swarms. Summary of the Invention

[0004] The purpose of this invention is to overcome the above-mentioned defects and provide a node selection method for cluster cooperative navigation.

[0005] The objective of this invention is achieved through the following technical solution: a node selection method for cluster cooperative navigation, comprising the following steps:

[0006] All nodes within the cluster are equipped with the same low-cost navigation equipment, including body sensors, cooperative ranging sensors, and communication devices. The node selection method includes the following steps:

[0007] S1: Collect dynamic information from other nodes within the cluster system;

[0008] S2: Establish a comprehensive evaluation model matrix for collaborative navigation nodes;

[0009] S3: Calculate the multi-index weight vector using the entropy weight method ;

[0010] S4: Based on fish aggregation and selective feeding behavior, a biomimetic enhancement operator is established, and a distance attenuation factor and a signal fitness factor are introduced.

[0011] S5: Achieve robust node selection for a low-cost cluster collaborative navigation system in dynamic environments;

[0012] S6: Based on the selected nodes, information fusion is completed to obtain the navigation information of the current node.

[0013] Furthermore, in S1, dynamic information of other nodes within the cluster system is collected, including: the number of cooperating nodes within the communication range, the location of the cooperating nodes within the communication range, the location information matrix of the filtered estimation, the relative distance information, the reliability index of the ranging signal, and the reliability index of the main sensor.

[0014] Furthermore, in S2, a comprehensive evaluation model matrix for collaborative navigation nodes is established. Utilizing indicators such as the geometric distribution of collaborative nodes in the cluster, information uncertainty, jump stability, and collaborative signal quality, a multi-indicator fusion comprehensive node evaluation model is constructed, including the following steps:

[0015] S2-1: Define the comprehensive node evaluation model matrix ,

[0016] In the formula, i represents the current node number. , This represents the total number of nodes in the cluster and the corresponding number of each node. The set of candidate cooperating nodes within the communication range of node i. quantity, Dimensions representing comprehensive evaluation indicators This represents the quantified value of the candidate node j on the index m.

[0017] S2-2: Calculate the quantitative indicators for each candidate node, specifically including:

[0018] (1) Geometric distribution index of collaborative nodes ,in, The direction vector of node j relative to the current node i is represented by the direction vector of node j. To simplify the calculation and improve the efficiency of node selection, this method proposes to use the observation cosine of the cooperative nodes instead of the cooperative accuracy factor. The observation direction cosine between each pair of cooperative nodes is calculated, and the minimum cosine value is used as the geometric distribution index to avoid multiple cooperative nodes from collinearity and optimize the node selection target from the perspective of topology.

[0019] (2) Information uncertainty index of filtering estimation ,in The submatrix representing the position covariance of node j during the information fusion process. The largest eigenvalue corresponding to the covariance. The amount of information is small to avoid computational anomalies. This method uses the position information matrix of the cooperative nodes to calculate the estimation uncertainty of the main direction, which is used to characterize the estimation accuracy of the cooperative nodes and optimize node selection from the perspective of information gain.

[0020] (3) Jump stability index of cooperative nodes , In the formula, It is the weighted moving standard deviation. Representing the terminal abrupt change degree, this method combines the global volatility and terminal local abrupt change degree of the candidate nodes within the sliding window to characterize the estimated position jump situation of the cooperative nodes, optimizing node selection from a stability perspective. Specifically, the calculation steps for the jump stability index of the cooperative nodes are as follows:

[0021] a) Define and compute the step-by-step drift sequence within the sliding window. In the formula This represents the positional difference between two adjacent moments. Represents the current discrete time;

[0022] b) Calculate global volatility within the time window , In the formula, It is the total length of the sliding window. The weights represent linearly increasing weights corresponding to the current time step, and are closer to the current time step. The sample signal is more sensitive;

[0023] c) Calculate the terminal abrupt change degree In the formula, It is the first-order difference of the gradual drift amplitude. It is the set length of the end of the sliding window, and ;

[0024] (4) Ranging link quality indicators of collaborative nodes In the formula, The ranging reliability index represents the current link and is used to characterize the ranging quality with candidate nodes, optimizing node selection from the perspective of collaborative information quality.

[0025] (5) Reliability indicators of the body sensors of the collaborative node In the formula Representative candidate nodes The position accuracy factor of the local sensor satellite receiver is used to characterize the quality of other measurement information of the candidate node.

[0026] Existing methods often rely solely on a single indicator such as geometric configuration or covariance confidence, making it difficult to comprehensively measure node quality. The steps described above include five quantitative indicators that together form a comprehensive node evaluation model matrix. Compared to existing node selection methods that rely on only a single indicator or a few indicators, this method achieves a comprehensive quantitative evaluation of the collaborative node's sensing capabilities, stability, communication quality, and configuration rationality, and is closer to the actual system requirements.

[0027] Furthermore, in S4, inspired by fish swarming and selective feeding behaviors, and drawing on the spatial proximity and signal perception intensity characteristics of fish selectively swarming, a distance attenuation factor is introduced. With signal fitness factor Establishing a biomimetic enhancement operator In the formula, The relative distance between node i and node j is obtained by the cooperative ranging sensor. This represents the average relative distance between the current node and all candidate nodes. Let be the signal suppression intensity factor, taken as... Traditional optimization methods often neglect the spatial relationships and behavioral response mechanisms between nodes, failing to reflect the collaborative characteristics of a group. This paper introduces biomimetic inspiration to demonstrate the "near-distance priority and signal selection" intelligent collaborative behavior of a fish swarm, improving the algorithm's environmental adaptability under topological changes.

[0028] Furthermore, in S5, robust node selection for a low-cost cluster cooperative navigation system in dynamic environments is achieved by utilizing the weight vector constructed from the information entropy described in S3. The scores of each node are calculated using entropy weight-driven index fusion. Combined with the biomimetic enhancement operator described in S4, the biomimetic enhancement score is calculated. ,in, Representing the strength of biomimetic-driven enhancement, the top performers are selected based on the final biomimetic enhancement score. This method employs a multi-dimensional scoring matrix and introduces a biomimetic behavior-driven mechanism to achieve robust node selection for collaborative navigation systems in dynamic environments. This comprehensive quantitative evaluation of collaborative nodes considers factors such as information quality, state stability, sensor reliability, link performance, and spatial structure. Compared to traditional methods that rely solely on geometric configuration or confidence levels, this mechanism offers advantages such as more comprehensive information fusion, more adaptive weight allocation, more accurate spatial perception, and more efficient deployment and operation. It significantly improves the accuracy of collaborative node selection and system robustness, making it suitable for real-time collaborative navigation tasks in complex dynamic environments.

[0029] Furthermore, S6 includes the following process:

[0030] Based on the aforementioned optimization process, at time t, the node i to be located is selected. With 3 cooperating nodes, the state equation for cooperative navigation is:

[0031]

[0032] Where node j is the selected collaborating node. Represents state noise. Represents the noise covariance matrix; Represents measurement noise. Represents the measurement noise covariance matrix; represent The n-dimensional state vector at time node i This represents the observation vector of node i at time t. The state prediction vector of node i at time t; Represents the state equation. The measurement equation represents node i at time t; where the measurement input is... for: , specifically and Let represent the velocity and position vectors of node i at time t, respectively, calculated by inertial navigation. and These represent the velocity and position vectors received by node i from the satellite navigation system at time t, respectively. For node i according to Estimated location information The estimated location information for node j.

[0033] Specifically, the prediction stage of information fusion can be solved as follows:

[0034]

[0035] In the formula, This represents the state covariance matrix of node i at time t-1. The predicted state covariance at time t. The value representing the l-th volume point is obtained by calculating the third-order spherical surface integral. Represents the estimated state at time t-1;

[0036] Furthermore, based on the aforementioned optimization process, the measurement update process is as follows:

[0037]

[0038]

[0039]

[0040]

[0041] Calculate the estimated state vector and corresponding covariance of node i:

[0042]

[0043]

[0044]

[0045] This approach enables information fusion based on selected nodes to obtain navigation information for the current node. Utilizing high-quality cooperative nodes selected through the proposed comprehensive optimization method for information fusion effectively suppresses interference from nodes with abnormal ranging, unstable states, or degraded sensors, significantly improving the overall accuracy and robustness of the cooperative navigation system. Compared to traditional methods that do not differentiate between node quality and simply fuse all neighbor information on an average or fixed basis, this method effectively controls error propagation and the accumulation of undesirable information while maintaining information contribution. It exhibits stronger stability and anti-interference capabilities, especially in scenarios with weak observation or partial node anomalies, providing crucial support for highly reliable navigation of unmanned surface vessel swarms in complex environments.

[0046] A node selection method for cluster cooperative navigation is provided. This method has program modules corresponding to the steps of the above-described method, and executes the steps in the above-described node selection method for cluster cooperative navigation when running.

[0047] The advantages of this invention compared to the prior art are:

[0048] This invention addresses the challenges of low-cost multi-unmanned surface vessel (USV) swarms by proposing a biomimetic neighborhood-aware node selection method. This method overcomes the limitations of traditional approaches that indiscriminately use node information, leading to heavy communication burdens and decreased positioning accuracy. By simultaneously considering node geometric distribution, information uncertainty, transition stability, and cooperative signal quality, a comprehensive node evaluation model integrating multiple indicators is constructed, providing a more comprehensive and objective reflection of node contributions to cooperative navigation. Furthermore, this invention introduces a biomimetic enhancement operator derived from fish swarming and selective foraging behaviors. Based on entropy-weighted scoring, it combines distance attenuation factors and signal fitness factors, effectively improving the robustness and global adaptability of node selection. Furthermore, by incorporating information from selected high-quality nodes into cooperative navigation fusion, the system's ability to identify and isolate abnormal nodes is enhanced, and positioning stability under weak observation or link degradation conditions is significantly improved. Simultaneously, the proposed algorithm is characterized by its simple structure, low computational cost, and ease of engineering implementation. It can be deployed in multi-node distributed cooperative architectures to meet the online decision-making needs in dynamic environments. Furthermore, this method supports practical conditions such as varying node numbers, autonomous collaborative networking, and limited communication, demonstrating good scalability and adaptability.

[0049] In summary, compared with existing optimization methods that rely on a single indicator or static rules, this invention can not only dynamically select high-quality cooperative nodes, reduce redundant communication, and improve information utilization efficiency, but also ensure the accuracy and stability of cluster cooperative navigation in complex dynamic environments. It provides systematic and engineering-based technical support for the efficient collaboration and robust positioning of low-cost unmanned surface vessel clusters. Attached Figure Description

[0050] Figure 1 A flowchart of a node selection method for cluster collaborative navigation in an embodiment of the present invention.

[0051] Figure 2 A schematic diagram of cluster topology changes and node collaboration in an embodiment of the present invention.

[0052] Figure 3 A schematic diagram of the simulated unmanned surface vessel swarm trajectory in an embodiment of the present invention.

[0053] Figure 4 Comparison of heading angle errors in unmanned surface vessel (USV) swarms in embodiments of this invention.

[0054] Figure 5 A comparison diagram of the horizontal position error of an unmanned surface vessel cluster in an embodiment of the present invention.

[0055] Figure 6 A comparison diagram of the collaborative nodes in an unmanned surface vessel (USV) swarm in an embodiment of this invention. Detailed Implementation

[0056] The present invention will be further described in detail with reference to the accompanying drawings and specific embodiments:

[0057] Example: Figure 1 As shown, the present invention provides a node selection method for cluster cooperative navigation, comprising the following steps:

[0058] S1: Collect dynamic information from other nodes within the cluster system;

[0059] S2: Establish a comprehensive evaluation model matrix for collaborative navigation nodes;

[0060] S3: Calculate the multi-index weight vector using the entropy weight method ;

[0061] S4: Based on fish aggregation and selective feeding behavior, a biomimetic enhancement operator is established, and a distance attenuation factor and a signal fitness factor are introduced.

[0062] S5: Achieve robust node selection for a low-cost cluster collaborative navigation system in dynamic environments;

[0063] S6: Based on the selected nodes, information fusion is completed to obtain the navigation information of the current node.

[0064] In S1, dynamic information of other nodes within the cluster system is collected, including: the number of cooperating nodes within the communication range. Location of cooperating nodes within communication range , Filtered estimation of location information matrix Relative distance information Ranging signal reliability index Reliability indicators of the main sensor .

[0065] In S2, a comprehensive evaluation model matrix for collaborative navigation nodes is established. Utilizing indicators such as the geometric distribution of collaborative nodes in the cluster, information uncertainty, jump stability, and collaborative signal quality, a multi-indicator fusion comprehensive node evaluation model is constructed, including the following steps:

[0066] S2-1: Define the comprehensive node evaluation model matrix ,

[0067] In the formula, i represents the current node number. , This represents the total number of nodes in the cluster and the corresponding number of each node. The set of potential cooperative nodes within the communication range of node i is called the neighboring nodes. quantity, Dimensions representing comprehensive evaluation indicators This represents the quantified value of the candidate node j on the index m.

[0068] S2-2: Calculate the quantitative indicators for each candidate node, specifically including:

[0069] (1) Geometric distribution index of collaborative nodes ,in, The direction vector of node j relative to the current node i is represented by the direction vector of node j. To simplify the calculation and improve the efficiency of node selection, this method proposes to use the observation cosine of the cooperative nodes instead of the cooperative accuracy factor. The observation direction cosine between each pair of cooperative nodes is calculated, and the minimum cosine value is used as the geometric distribution index to avoid multiple cooperative nodes from collinearity and optimize the node selection target from the perspective of topology.

[0070] (2) Information uncertainty index of filtering estimation ,in The submatrix representing the position covariance of node j during the information fusion process. The largest eigenvalue corresponding to the covariance. The amount of information is small to avoid computational anomalies. This method uses the position information matrix of the cooperative nodes to calculate the estimation uncertainty of the main direction, which is used to characterize the estimation accuracy of the cooperative nodes and optimize node selection from the perspective of information gain.

[0071] (3) Jump stability index of cooperative nodes , In the formula, It is the weighted moving standard deviation. Representing the terminal abrupt change degree, this method combines the global volatility and terminal local abrupt change degree of the candidate nodes within the sliding window to characterize the estimated position jump situation of the cooperative nodes, optimizing node selection from a stability perspective. Specifically, the calculation steps for the jump stability index of the cooperative nodes are as follows:

[0072] a) Define and compute the step-by-step drift sequence within the sliding window. In the formula This represents the positional difference between two adjacent moments. Represents the current discrete time;

[0073] b) Calculate global volatility within the time window , In the formula, It is the total length of the sliding window. The weights represent linearly increasing weights corresponding to the current time step, and are closer to the current time step. The sample signal is more sensitive;

[0074] c) Calculate the terminal abrupt change degree In the formula, It is the first-order difference of the gradual drift amplitude. It is the set length of the end of the sliding window, and ;

[0075] (4) Ranging link quality indicators of collaborative nodes In the formula, The ranging reliability index represents the current link and is used to characterize the ranging quality with candidate nodes, optimizing node selection from the perspective of collaborative information quality.

[0076] (5) Reliability indicators of the body sensors of the collaborative node In the formula Representative candidate nodes The position accuracy factor of the local sensor satellite receiver is used to characterize the quality of other measurement information of the candidate node.

[0077] In S4, inspired by fish swarming and selective feeding behaviors, and drawing on the spatial proximity and signal perception intensity characteristics of fish selective swarming, a distance attenuation factor is introduced. With signal fitness factor Establishing a biomimetic enhancement operator In the formula, The relative distance between node i and node j is obtained by the cooperative ranging sensor. This represents the average relative distance between the current node and all candidate nodes. Let be the signal suppression intensity factor, taken as... ;

[0078] In S5, robust node selection for a low-cost cluster cooperative navigation system in dynamic environments is achieved by utilizing the weight vector constructed from the information entropy described in S3. The scores of each node are calculated using entropy weight-driven index fusion. Combined with the biomimetic enhancement operator described in S4, the biomimetic enhancement score is calculated. ,in, Representing the strength of biomimetic-driven enhancement, the top performers are selected based on the final biomimetic enhancement score. A number of nodes are used to achieve robust node selection for a cluster-based collaborative navigation system in dynamic environments.

[0079] S6 includes the following process:

[0080] Based on the aforementioned optimization process, at time t, the node i to be located is selected. With 3 cooperating nodes, the state equation for cooperative navigation is:

[0081]

[0082] Where node j is the selected collaborating node. Represents state noise. Represents the noise covariance matrix; Represents measurement noise. Represents the measurement noise covariance matrix; represent The n-dimensional state vector at time node i This represents the observation vector of node i at time t. The state prediction vector of node i at time t; Represents the state equation. The measurement equation represents node i at time t; where the measurement input is... for: , specifically and Let represent the velocity and position vectors of node i at time t, respectively, calculated by inertial navigation. and These represent the velocity and position vectors received by node i from the satellite navigation system at time t, respectively. For node i according to Estimated location information The estimated location information for node j.

[0083] Specifically, the prediction stage of information fusion can be solved as follows:

[0084]

[0085] In the formula, This represents the state covariance matrix of node i at time t-1. The predicted state covariance at time t. The value representing the l-th volume point is obtained by calculating the third-order spherical surface integral. Represents the estimated state at time t-1;

[0086] Furthermore, based on the aforementioned optimization process, the measurement update process is as follows:

[0087]

[0088]

[0089]

[0090]

[0091] Calculate the estimated state vector and corresponding covariance of node i:

[0092]

[0093]

[0094]

[0095] This allows for information fusion based on the selected nodes, resulting in navigation information for the current node.

[0096] like Figure 2 As shown, as the cluster topology changes in real time, the nodes that each unmanned surface vessel (USV) can collaborate with also change dynamically; from the perspective of USV 3, the candidate neighbor nodes at a certain moment... Including nodes 2, 4, 5, 6, and 7, in the next moment, the neighboring nodes change to nodes 1, 2, 4, 5, and 6. Node 2 has a positioning deviation due to its own fault, and the cooperative information it provides has a large error. Node 6 also cannot provide reliable cooperative information due to communication or ranging link abnormalities caused by occlusion or other reasons. At this time, through the robust node selection of the low-cost cluster cooperative navigation system in dynamic environment, an appropriate cooperative node can be selected for the unmanned surface vessel 3, thereby ensuring the continuous high-precision positioning of node 3.

[0097] To demonstrate the effectiveness of the node selection method for cluster cooperative navigation of the present invention, the present invention is compared with algorithms that use cooperative nodes within the entire communication range (method 1), algorithms that randomly select cooperative nodes (method 2), node selection methods based on geometric precision factors (method 3), and node selection methods based on node position confidence (method 4). All of these methods use capacitive Kalman filtering for information fusion.

[0098] The basic parameters were set as follows: a collaborative navigation simulation experiment was designed for a swarm of six unmanned surface vessels (USVs). Each USV was equipped with the same low-cost inertial measurement unit (IMU), satellite receiver, and ranging module parameters. The simulation duration was 600 seconds, the IMU frequency was 100Hz, and the gyroscope constant offset was [value missing]. Random noise is The constant bias of the accelerometer is Random noise is The satellite and ranging module operate at a frequency of 1 Hz, with a velocity accuracy of [missing information]. The information exchange cycle is 1Hz, the designed range for ranging and information exchange is 300m, and the ranging accuracy is [missing information]. The initial attitude error of the unmanned surface vessels within the cluster is set to... Initial velocity error Initial position error To meet simulation requirements, the design is as follows: Figure 3 The trajectory diagram shown indicates that all six unmanned surface vessels (USVs) are moving at varying speeds, numbered 1, 2, 3, 4, 5, and 6; the maximum number of cooperating nodes is designed to be 3.

[0099] Given the complex water surface environment, the ranging noise of the six unmanned surface vessels was designed to be:

[0100]

[0101] The ranging noise distribution representing an 80% probability for all six unmanned surface vessels is as follows: There is a 20% probability that the ranging noise distribution is as follows: This situation occurs randomly throughout the simulation.

[0102] The results of the method of the present invention and four other comparative methods are shown, such as Figure 4 The figure shows the heading angle errors of six unmanned surface vessels under different methods. Figure 5 The image shows a comparison of the horizontal position errors of six unmanned surface vessels using different methods. Figure 6 The figure shows the node selection status of different methods throughout the simulation. Black dots represent selected nodes, white dots represent unselected nodes, and the vertical axis represents all nodes in the corresponding cluster. The results show that the preferred method proposed in this invention achieves the highest navigation accuracy and maintains consistently high-precision positioning. This is because, compared to other node selection methods that randomly select nodes or consider only a single influencing factor, this invention considers a combination of multiple influencing factors and enhances the selection through biomimetic techniques, resulting in more stable and effective results in the selection of cooperative nodes. Therefore, even with noise interference, it can still present stable navigation results.

[0103] The quantitative analysis of the positioning error RMSE and single-step execution time of the proposed method and four other methods is shown in Table 1. According to the results, the proposed method exhibits the best heading attitude accuracy and horizontal position accuracy in all nodes of the cluster. At the same time, the execution time is only slightly longer than that of method 2 with randomly selected nodes, and shorter than that of other methods.

[0104]

[0105] Through simulation experiments, considering both navigation accuracy and execution time, it can be found that the proposed method has better accuracy and computational performance than other methods.

[0106] The above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention in any other way. Any modifications or equivalent changes made based on the technical essence of the present invention shall still fall within the scope of protection claimed by the present invention.

Claims

1. A node selection method for cluster cooperative navigation, characterized in that: All nodes within the cluster are equipped with the same low-cost navigation equipment, including body sensors, cooperative ranging sensors, and communication devices. The node selection method includes the following steps: S1: Collect dynamic information from other nodes within the cluster system; S2: Establish a comprehensive evaluation model matrix for collaborative navigation nodes; S3: Calculate the multi-index weight vector using the entropy weight method ; S4: Based on fish aggregation and selective feeding behavior, a biomimetic enhancement operator is established, and a distance attenuation factor and a signal fitness factor are introduced. S5: Achieve robust node selection for a low-cost cluster collaborative navigation system in dynamic environments; S6: Based on the selected nodes, information fusion is completed to obtain the navigation information of the current node.

2. The node selection method for cluster cooperative navigation according to claim 1, characterized in that: In step S1, dynamic information of other nodes within the cluster system is collected, including: the number of cooperating nodes within the communication range, the location of the cooperating nodes within the communication range, the location information matrix of the filtered estimation, the relative distance information, the reliability index of the ranging signal, and the reliability index of the body sensor.

3. The node selection method for cluster cooperative navigation according to claim 1, characterized in that: In step S2, a comprehensive evaluation model matrix for collaborative navigation nodes is established. This model utilizes indicators such as the geometric distribution of collaborative nodes in the cluster, information uncertainty, transition stability, and collaborative signal quality to construct a multi-indicator integrated evaluation model. The steps include: S2-1: Define the comprehensive node evaluation model matrix , In the formula, i represents the current node number. , This represents the total number of nodes in the cluster and the corresponding number of each node. The set of candidate cooperating nodes within the communication range of node i. quantity, Dimensions representing comprehensive evaluation indicators This represents the quantified value of the candidate node j on index m. S2-2: Calculate the quantitative indicators for each candidate node, specifically including: (1) Geometric distribution index of collaborative nodes ,in, Let $j$ represent the direction vector of node $j$ relative to the current node $i$. A method is proposed that uses the observation cosine of cooperative nodes instead of the cooperative accuracy factor. The observation direction cosines between each pair of cooperative nodes are calculated, and the minimum cosine value is used as the geometric distribution index to avoid collinearity among multiple cooperative nodes. This optimizes node selection from a topological perspective. (2) Information uncertainty index of filtering estimation ,in The submatrix representing the position covariance of node j during the information fusion process. The largest eigenvalue corresponding to the covariance. The amount is small to avoid calculation errors; (3) Jump stability index of cooperative nodes , In the formula, It is the weighted moving standard deviation. Represents the terminal abrupt change degree; the calculation steps for the jump stability index of the cooperative node are as follows: a) Define and compute the step-by-step drift sequence within the sliding window. In the formula This represents the positional difference between two adjacent moments. Represents the current discrete time; b) Calculate global volatility within the time window , In the formula, It is the total length of the sliding window. The weights represent linearly increasing weights corresponding to the current time step, and are closer to the current time step. The sample signal is more sensitive; c) Calculate the terminal abrupt change degree In the formula, It is the first-order difference of the gradual drift amplitude. It is the set length of the end of the sliding window, and ; (4) Ranging link quality indicators of collaborative nodes In the formula, The ranging reliability index represents the current link and is used to characterize the ranging quality with candidate nodes, optimizing node selection from the perspective of collaborative information quality. (5) Reliability indicators of the body sensors of the collaborative node In the formula Representative candidate nodes The position accuracy factor of the local sensor satellite receiver is used to characterize the quality of other measurement information of the candidate node.

4. The node selection method for cluster cooperative navigation according to claim 1, characterized in that: In S4, a distance attenuation factor is introduced. With signal fitness factor Establishing a biomimetic enhancement operator In the formula, The relative distance between node i and node j is obtained by the cooperative ranging sensor. This represents the average relative distance between the current node and all candidate nodes. Let be the signal suppression intensity factor, taken as... .

5. The node selection method for cluster cooperative navigation according to claim 1, characterized in that: In step S5, robust node selection for a low-cost cluster cooperative navigation system under dynamic environments is achieved using a weight vector constructed from the information entropy in step S3. The scores of each node are calculated using entropy weight-driven index fusion. Combined with the biomimetic enhancement operator in S4, the biomimetic enhancement score is calculated. ,in, Representing the strength of biomimetic-driven enhancement, the top performers are selected based on the final biomimetic enhancement score. A number of nodes are used to achieve robust node selection for a cluster-based collaborative navigation system in dynamic environments.

6. The node selection method for cluster cooperative navigation according to claim 1, characterized in that: S6 includes the following process: At time t, select the node i to be located. With 3 cooperating nodes, the state equation for cooperative navigation is: Where node j is the selected collaborating node. Represents state noise. Represents the noise covariance matrix; Represents measurement noise. Represents the measurement noise covariance matrix; represent The n-dimensional state vector at time node i This represents the observation vector of node i at time t. The state prediction vector of node i at time t; Represents the state equation. The measurement equation represents node i at time t; where the measurement input is... for: , specifically and Let represent the velocity and position vectors of node i at time t, respectively, calculated by inertial navigation. and These represent the velocity and position vectors received by node i from the satellite navigation system at time t, respectively. For node i according to Estimated location information The estimated location information for node j. The prediction phase of information fusion is solved as follows: In the formula, This represents the state covariance matrix of node i at time t-1. The predicted state covariance at time t. The value representing the l-th volume point is obtained by calculating the third-order spherical surface integral. Represents the estimated state at time t-1; The measurement update process is as follows: Calculate the estimated state vector and corresponding covariance of node i: This allows for information fusion based on the selected nodes, resulting in navigation information for the current node.

7. The node selection method for cluster cooperative navigation according to claim 1, characterized in that, The method has a program module corresponding to the steps of the method described in any one of claims 1 to 6, and executes the steps in the above-described method for selecting nodes in a cluster cooperative navigation when it is run.

8. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the program, it implements a node selection method for cluster cooperative navigation as described in any one of claims 1 to 6.

9. A computer-readable storage medium storing computer instructions thereon, characterized in that, When the computer instructions are executed by the processor, they implement a node selection method for cluster cooperative navigation as described in any one of claims 1-6.

Images