Air-ground cooperative multi-source sensing and unexploded object intelligent identification system

Abstract

The invention discloses an air-ground cooperative multi-source perception and unexploded object intelligent identification system, and relates to the technical field of intelligent identification. The method comprises the following steps: firstly, acquiring perception data of each cooperation unit corresponding to a perception target in an air-ground cooperation multi-source perception area by an unexploded object intelligent identification perception platform, and performing multi-source confidence degree processing based on heterogeneous perception data of each cooperation unit corresponding to the perception target to obtain a perception confidence degree evaluation value; carrying out conflict analysis on the perception confidence evaluation value of each cooperation unit corresponding to the perception target to obtain a decision weight vector, carrying out task urgency analysis on the operation state data of each cooperation unit corresponding to the perception target to obtain a task urgency analysis index value, and carrying out task urgency analysis according to the task urgency analysis index value to obtain a decision weight vector; and in combination with decision-making weight vector dynamic cooperative regulation and control analysis, a cooperative guidance instruction is obtained, and then dynamic reconfiguration of air-ground cooperative resources and planning configuration of an action sequence are performed, so that the reliability of task identification processing of the system is improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of intelligent identification, in particular to an air-ground collaborative multi-source perception and unexploded ordnance intelligent identification system. BACKGROUND

[0002] In the field of collaborative multi-source perception and intelligent identification, with the continuous improvement of the autonomy level of unmanned systems and the increase of task complexity, intelligent collaboration and consistent decision-making between heterogeneous platforms have become a key technical challenge to improve the overall efficiency and reliability of the system. Air-ground collaborative multi-source perception, as a core technical approach to realize wide-area coverage and accurate verification, the dynamic consistency of cross-platform collaborative decision-making is directly related to the final safety of task execution.

[0003] However, in the prior art, the focus is usually on the front-end fusion of multi-source heterogeneous perception data, and the core goal is to build a unified global situation map, and it is assumed that each platform can autonomously derive a coordinated and consistent action logic based on this consensus. This method ignores the potential risks caused by platform heterogeneity, especially at the decision-making level. When different platform local decision-making logic interacts with global collaborative rules, conflicts may arise due to differences in the perception of the same event. For example, when delineating a safe operating area, a high-altitude unmanned aerial vehicle may determine a certain area to be a low-thermal feature area and suggest passing through based on thermal imaging, while a ground robot may strongly recommend avoiding the area based on underground magnetic field detection of metal signals. Such cognitive bias is derived from the physical perception methods used by different platforms, not data errors. This bias can lead to logical conflicts in the collaborative decision-making process, resulting in decision-making deadlock at critical action nodes. This situation not only may cause delays in the task time window, but also may generate mutually contradictory instructions, exposing subsequent platforms to operational risks and affecting the execution of the entire task chain, thereby severely damaging the overall efficiency of the system. SUMMARY

[0004] To overcome the deficiencies of the prior art, the present application provides an air-ground collaborative multi-source perception and unexploded ordnance intelligent identification system, which can effectively solve the problems involved in the background art.

[0005] To achieve the above purpose, the present application realizes the following technical scheme: an air-ground collaborative multi-source perception and unexploded ordnance intelligent identification system, comprising a data acquisition module for collecting perception data of each collaborative unit corresponding to a perception target in an air-ground collaborative multi-source perception area by an unexploded ordnance intelligent identification perception platform, wherein the perception data of each collaborative unit corresponding to the perception target includes heterogeneous perception data of each collaborative unit corresponding to the perception target and operating state data of each collaborative unit corresponding to the perception target.

[0006] The confidence evaluation module is configured to perform multi-source confidence processing on the heterogeneous perception data of each cooperative unit corresponding to the perception target, to obtain a perception confidence evaluation value of each cooperative unit corresponding to the perception target.

[0007] The conflict analysis module is configured to perform conflict analysis on the perception confidence evaluation values of each cooperative unit corresponding to the perception target, to obtain a decision weight vector of the perception target.

[0008] The urgency analysis module is configured to perform task urgency analysis on the operation state data of each cooperative unit corresponding to the perception target, to obtain a task urgency analysis index value of the perception target.

[0009] The instruction analysis module is configured to dynamically and cooperatively control and analyze, according to the task urgency analysis index value of the perception target, in combination with the decision weight vector of the perception target, to obtain a cooperative guidance instruction of the perception target, and then to perform dynamic reconfiguration of air-ground cooperative resources and planning and configuration of an action sequence.

[0010] Further, the heterogeneous perception data of each cooperative unit corresponding to the perception target includes a sensing deviation coefficient, a multipath effect interference coefficient, a sensing signal signal-to-noise ratio average value, and a feature point matching rate of each cooperative unit corresponding to the perception target.

[0011] The operation state data of each cooperative unit corresponding to the perception target includes a residual safe operation duration, a target arrival estimated duration, a target exposure duration, and a historical task instruction average response delay duration of each cooperative unit corresponding to the perception target.

[0012] Further, the method for obtaining the perception confidence evaluation value of each cooperative unit corresponding to the perception target includes performing multi-source confidence processing on the heterogeneous perception data of each cooperative unit corresponding to the perception target, to obtain the perception confidence evaluation value of each cooperative unit corresponding to the perception target, wherein the perception confidence evaluation value of each cooperative unit corresponding to the perception target represents a quantitative result of a comprehensive reliability degree of the heterogeneous perception data of each cooperative unit corresponding to the perception target on perception data and identification conclusions output by each cooperative unit for the perception target.

[0013] Further, the method for performing conflict analysis on the perception confidence evaluation values of each cooperative unit corresponding to the perception target includes extracting the perception confidence evaluation values of each cooperative unit corresponding to the perception target, arranging the perception confidence evaluation values of all cooperative units corresponding to the perception target in descending order, and selecting a perception confidence evaluation value ranked first as a perception confidence evaluation reference value.

[0014] The difference between the perception confidence assessment value of each cooperative unit corresponding to the perception target and the perception confidence assessment reference value is calculated, and the absolute value is obtained to obtain the difference absolute value of the perception confidence assessment value of each cooperative unit corresponding to the perception target.

[0015] The difference absolute value of the perception confidence assessment value of each cooperative unit corresponding to the perception target is compared with a preset difference threshold to obtain the conflict determination result of each cooperative unit corresponding to the perception target. The preset difference threshold includes a preset first difference threshold and a preset second difference threshold. The preset first difference threshold is greater than the preset second difference threshold.

[0016] Further, the method for obtaining the conflict determination result of each cooperative unit corresponding to the perception target includes that the conflict determination result of each cooperative unit corresponding to the perception target includes existence of direct conflict, existence of potential conflict and non-existence of conflict.

[0017] If the difference absolute value of the perception confidence assessment value of a certain cooperative unit corresponding to the perception target is higher than the preset first difference threshold, it is determined that the cooperative unit exists direct conflict, and the conflict determination result of the cooperative unit corresponding to the perception target is marked as existence of direct conflict.

[0018] If the difference absolute value of the perception confidence assessment value of a certain cooperative unit corresponding to the perception target is lower than or equal to the preset first difference threshold, and the difference absolute value of the perception confidence assessment value of the cooperative unit corresponding to the perception target is higher than the preset second difference threshold, it is determined that the cooperative unit exists potential conflict, and the conflict determination result of the cooperative unit corresponding to the perception target is marked as existence of potential conflict.

[0019] If the difference absolute value of the perception confidence assessment value of a certain cooperative unit corresponding to the perception target is lower than or equal to the preset second difference threshold, it is determined that the cooperative unit does not exist conflict, and the conflict determination result of the cooperative unit corresponding to the perception target is marked as non-existence of conflict.

[0020] Further, the method for obtaining the decision weight vector of the perception target is that the conflict determination result of each cooperative unit corresponding to the perception target is extracted. If the conflict determination result of a certain cooperative unit corresponding to the perception target is marked as existence of direct conflict or existence of potential conflict, the conflict factor of the cooperative unit corresponding to the perception target is matched according to the difference absolute value of the perception confidence assessment value of the cooperative unit corresponding to the perception target, and the product of the conflict factor of the cooperative unit corresponding to the perception target and the perception confidence assessment value is recorded as the decision weight value of the cooperative unit corresponding to the perception target.

[0021] If the conflict determination result of the certain coordination unit corresponding to the perception target is marked as no conflict, the perception confidence evaluation value of the coordination unit corresponding to the perception target is directly recorded as the decision weight value of the coordination unit corresponding to the perception target, and the decision weight values of all coordination units corresponding to the perception target are sequentially arranged according to the number of the coordination units to form a decision weight vector of the perception target.

[0022] Further, the method for performing task urgency analysis to obtain a task urgency analysis index value of the perception target is: performing task urgency analysis on the running state data of each coordination unit corresponding to the perception target to obtain a task urgency analysis index value of the perception target, wherein the task urgency analysis index value of the perception target represents a quantitative result of the running state data of each coordination unit corresponding to the perception target on the emergency degree of the task of the perception target.

[0023] Further, the method for dynamically analyzing coordination and regulation to obtain a coordination guidance instruction of the perception target is: extracting a maximum value of the decision weight values from the decision weight vector of the perception target, and recording the maximum value as a maximum decision weight value of the perception target.

[0024] Comparing the task urgency analysis index value of the perception target and the maximum decision weight value with a preset task urgency analysis index threshold value and a preset maximum decision weight threshold value respectively, to obtain a coordination guidance instruction of the perception target.

[0025] Further, the method for obtaining the coordination guidance instruction of the perception target is: the coordination guidance instruction of the perception target includes a dominant verification instruction, a coordination re-verification instruction and a continuous observation instruction.

[0026] If the task urgency analysis index value of the perception target is higher than the preset task urgency analysis index threshold value, and the maximum decision weight value of the perception target is higher than the preset maximum decision weight threshold value, the coordination guidance instruction of the perception target is marked as a dominant verification instruction.

[0027] If the task urgency analysis index value of the perception target is higher than the preset task urgency analysis index threshold value, and the maximum decision weight value of the perception target is lower than or equal to the preset maximum decision weight threshold value, the coordination guidance instruction of the perception target is marked as a coordination re-verification instruction.

[0028] If the task urgency analysis index value of the perception target is lower than or equal to the preset task urgency analysis index threshold value, the coordination guidance instruction of the perception target is marked as a continuous observation instruction.

[0029] Further, the method for further carrying out dynamic reconfiguration of air-ground collaborative resources and planning configuration of action sequence is: extracting a collaborative guidance instruction of a perception target, carrying out dynamic reconfiguration of air-ground collaborative resources and planning configuration of an action sequence, dynamically scheduling and reconfiguring the air-ground collaborative resources, optimizing task allocation of each collaborative unit corresponding to the perception target, and generating task allocation and an action sequence of the perception target.

[0030] The present application has the following advantages: (1) The present application reduces the decision deadlock problem caused by cognitive bias in the multi-source perception system. The prior art stops at generating a unified global situation map, but cannot make a judgment when multiple platforms based on different physical perception principles draw contradictory conclusions. Usually, the system will be stuck before making a conflict instruction to pass or avoid. The present application introduces a conflict analysis module and a decision weight vector to convert cognitive conflicts into quantifiable and calculable decision weight values and generate clear collaborative guidance instructions. According to the objective analysis of data reliability and task urgency, the most efficient and safest action path can be selected to ensure the continuity and decisiveness of the task chain at critical moments.

[0031] (2) The present application can automatically switch between three different collaborative modes of dominant verification, collaborative re-verification and continuous observation according to the dynamic changes of the task urgency analysis index value and the decision weight vector through the linkage of urgency analysis and instruction analysis, and drive the resource scheduling strategy and action sequence generation rule that accurately match it. This dynamic adaptation capability greatly optimizes the overall resource utilization efficiency and task execution agility of the system in complex and variable environments.

[0032] (3) The present application improves the accuracy and task success rate of the entire air-ground collaborative system. The prior art lacks fault tolerance processing for single-point sensor failure or transient environmental interference. Local anomalies can easily be amplified into global task interruption through the collaborative rule network. In the present application, even if part of the unit data deviates due to interference, the system can identify and suppress its influence through conflict analysis, guide other reliable units to supplement through matching smaller conflict factors. This self-correction and resource elasticity reorganization capability based on group credibility evaluation enables the system to maintain stable output and continuous advancement of task goals when facing internal uncertainties such as equipment occasional failure and external dynamic challenges, thereby improving the task success rate. BRIEF DESCRIPTION OF DRAWINGS

[0033] Figure 1 It is a schematic diagram of the system module connection of the present application. DETAILED DESCRIPTION

[0034] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0035] Please see Figure 1 As shown, this embodiment of the invention provides a technical solution: an air-ground coordinated multi-source sensing and unexploded ordnance intelligent identification system, including a data acquisition module for the unexploded ordnance intelligent identification and sensing platform to collect sensing data of each coordinated unit corresponding to the sensing target in the air-ground coordinated multi-source sensing area. The sensing data of each coordinated unit corresponding to the sensing target includes heterogeneous sensing data of each coordinated unit corresponding to the sensing target and operating status data of each coordinated unit corresponding to the sensing target.

[0036] It should be added that the aforementioned unexploded ordnance intelligent identification and perception platform refers to the central computing and information fusion hub mounted on a large air vehicle. Its core function is to act as the intelligent brain of the entire air-ground collaborative system. It gathers heterogeneous perception data from all front-end collaborative units in real time through high-speed data links, runs intelligent processing software that integrates multi-source information fusion, feature extraction and pattern recognition algorithms, automatically detects, classifies and identifies suspicious targets in a wide-area perception area, and finally generates a unified situation map containing the type, geographical location and threat level of unexploded ordnance. It also provides authoritative and consistent perception target information input for subsequent collaborative decision-making and resource scheduling.

[0037] It needs to be explained that the collaborative units corresponding to the sensing targets in the air-ground collaborative multi-source sensing area refer to multiple heterogeneous platforms operating around the same target to be identified, namely suspected unexploded ordnance, within the air-ground collaborative multi-source sensing area. These platforms include high-altitude reconnaissance UAVs, low-altitude detailed reconnaissance UAVs, and ground-based bomb disposal robots. Each unit acts as an independent data acquisition and front-end processing node. Its core role is to use its own specific types of sensors, such as optical, infrared, radar, and magnetometers, to conduct complementary observations of the sensing target from different spatial locations and different physical dimensions. This allows it to acquire raw, multi-view, and multi-modal sensing data, which, along with its own status information, is reported in real time. Together, these data provide a distributed and multi-source data foundation for the back-end platform's fusion identification and collaborative decision-making.

[0038] The confidence assessment module is used to perform multi-source confidence processing based on the heterogeneous sensing data of each collaborative unit corresponding to the sensing target, and obtain the sensing confidence assessment value of each collaborative unit corresponding to the sensing target.

[0039] The conflict analysis module is used to perform conflict analysis on the perception confidence evaluation values ​​of each collaborative unit corresponding to the perception target, and obtain the decision weight vector of the perception target.

[0040] The urgency analysis module is used to perform task urgency analysis on the operational status data of each collaborative unit corresponding to the sensing target to obtain the task urgency analysis index value of the sensing target.

[0041] The instruction analysis module is used to analyze indicator values ​​based on the task urgency of the perceived target, combine them with the decision weight vector of the perceived target, and dynamically coordinate and control the analysis to obtain the coordinated guidance instructions for the perceived target, thereby carrying out dynamic reallocation of air-ground coordinated resources and planning and configuration of action sequences.

[0042] Specifically, the heterogeneous sensing data of each cooperative unit corresponding to the sensing target includes the sensing deviation coefficient, multipath effect interference coefficient, average signal-to-noise ratio of sensing signal, and feature point matching rate of each cooperative unit corresponding to the sensing target.

[0043] It should be noted that the sensing deviation coefficient reflects the inherent systematic error of the sensors in the collaborative unit after calibration. This is achieved by repeatedly measuring the sensors of each collaborative unit using a high-precision reference device in a controlled calibration field, recording the output value of each measurement, calculating the arithmetic mean of all output values, comparing this arithmetic mean with the true reference value provided by the reference device, calculating the difference between the two, and averaging the absolute values ​​of this difference. The average signal-to-noise ratio (SNR) represents the strength of the effective signal relative to the background noise when the unit senses a target. This is achieved by the spectrum analysis unit sampling multiple times within the sensing period, calculating the ratio of signal power to noise power, and converting it to decibels. The average signal-to-noise ratio of the sensor signal is obtained by taking the arithmetic mean of the decibel values ​​of the sampling points. The multipath interference coefficient quantifies the interference level caused by signal distortion due to electromagnetic waves propagating through multiple paths to the sensor data of this unit. The multipath interference coefficient is obtained by analyzing the time delay spread power spectrum of the received signal through the signal processing unit and calculating the power ratio of the direct signal to the multipath signal. The feature point matching rate refers to the quantitative value of the consistency of image features of the cooperative unit, mainly for units equipped with optical sensors, when continuously observing the target or synchronously observing with other units. It is obtained by the image processing unit detecting feature points in the sequence of images and performing cross-image matching, and dividing the number of successfully matched feature points by the total number of detected feature points.

[0044] The operational status data of each collaborative unit corresponding to the perceived target includes the remaining safe operation time, estimated time to reach the target, target exposure time, and average response delay time of historical task instructions for each collaborative unit corresponding to the perceived target.

[0045] It should be noted that the remaining safe operating time refers to the estimated remaining time that the collaborative unit can sustainably perform its mission based on its current energy reserves and real-time power consumption. This is calculated by reading the current battery charge or remaining fuel level from the onboard power management system, combining this with historical average power consumption data for the unit's current mission mode, and dividing the remaining energy by the average power consumption value. The estimated time to reach the target refers to the estimated time required for the collaborative unit to move from its current location to the location of the detected target. This is calculated by obtaining the coordinates of the unit and the detected target using a satellite positioning module, calculating the straight-line distance between them, and then dividing this distance by the unit's preset average moving speed in the current terrain. The target exposure time refers to the accumulated time since the detected target was first detected and continuously tracked by the system. This is calculated by recording the first timestamp of the detected target's identification using the central mission clock, and then calculating the difference between this timestamp and the current system timestamp. The average response delay of historical task instructions refers to the average delay of the collaborative unit in responding to system instructions over a period of time. The unit records the timestamp of receiving each instruction and the timestamp of starting to execute the corresponding action through the instruction log module, calculates the response delay of each instruction, and then takes the arithmetic mean of the delays of all instructions within a set period to obtain the average response delay of historical task instructions.

[0046] Specifically, the method for obtaining the perception confidence evaluation value of each collaborative unit corresponding to the perception target is as follows: multi-source confidence processing is performed based on the heterogeneous perception data of each collaborative unit corresponding to the perception target to obtain the perception confidence evaluation value of each collaborative unit corresponding to the perception target. The perception confidence evaluation value of each collaborative unit corresponding to the perception target represents the quantitative result of the comprehensive reliability of the perception data and recognition conclusion output by each collaborative unit for the perception target by the heterogeneous perception data of each collaborative unit corresponding to the perception target.

[0047] In this embodiment, the perception confidence evaluation value of each collaborative unit corresponding to the perceived target can be obtained through the following analysis method, with the specific analysis conditions as follows: ; In the formula, This represents the perception confidence assessment value of the i-th collaborative unit corresponding to the perceived target. This represents the sensing deviation coefficient of the i-th cooperative unit corresponding to the perceived target. This represents the confidence assessment factor corresponding to the set unit sensing deviation coefficient. This represents the multipath interference coefficient of the i-th cooperative unit corresponding to the perceived target. This represents the confidence assessment factor corresponding to the set multipath effect interference coefficient. This represents the average signal-to-noise ratio of the sensing signal of the i-th cooperative unit corresponding to the perceived target. This represents the confidence assessment factor corresponding to the average signal-to-noise ratio of the set sensor signal. This represents the feature point matching rate of the i-th collaborative unit corresponding to the perceived target. The confidence evaluation factor corresponding to the set feature point matching rate is represented by i, where i represents the number of each collaborative unit, i=1, 2, 3, ..., n, and n represents the total number of collaborative units.

[0048] It should be added that, in this embodiment, the confidence evaluation factors corresponding to the preset unit sensing deviation coefficient, multipath effect interference coefficient, average signal-to-noise ratio of sensing signal, and feature point matching rate are obtained from the intelligent recognition and perception database.

[0049] It should be explained that the confidence evaluation factors corresponding to the unit sensing deviation coefficient, multipath effect interference coefficient, average signal-to-noise ratio of the sensing signal, and feature point matching rate are used to adjust the importance of the data in the heterogeneous sensing data of each cooperative unit corresponding to the sensing target in the process of analyzing and obtaining the sensing confidence evaluation value. For example, a mapping relationship between the heterogeneous sensing data of each cooperative unit corresponding to the sensing target and the confidence evaluation factors is set in the intelligent recognition sensing database. Through the pre-set mapping relationship, the confidence evaluation factors corresponding to the heterogeneous sensing data of each cooperative unit corresponding to the sensing target in real time can be obtained. By matching the heterogeneous sensing data of each cooperative unit corresponding to the sensing target with the pre-set mapping relationship, the confidence evaluation factors corresponding to the unit sensing deviation coefficient, multipath effect interference coefficient, average signal-to-noise ratio of the sensing signal, and feature point matching rate are obtained.

[0050] In this implementation scheme, the sensing deviation coefficient, multipath interference coefficient, average signal-to-noise ratio of the sensing signal, and feature point matching rate of each cooperative unit corresponding to the sensing target are correlated and do not exist independently. For example, the sensing deviation coefficient directly affects the accuracy of feature point matching. When the inherent deviation of the sensor is large, even under ideal conditions, the extracted feature point positions may experience systematic drift, resulting in a decrease in the feature point matching rate across images or platforms. At the same time, an increase in the multipath interference coefficient will significantly degrade the quality of the received signal, which will not only directly lead to a decrease in the average signal-to-noise ratio of the sensing signal but also cause signal waveform distortion. This reduces the performance of recognition algorithms that rely on subtle signal features, indirectly affecting the reliability of the final recognition conclusion. A higher average signal-to-noise ratio of the sensor signal usually provides cleaner input for the feature extraction algorithm, helping to obtain a higher and more stable feature point matching rate. Comprehensive analysis of the perception confidence evaluation value of each collaborative unit corresponding to the perceived target can quantitatively evaluate the overall perception reliability of the collaborative unit under multiple factors such as comprehensive hardware error, environmental interference and signal quality. This helps to reasonably allocate weights and resolve conflicts of data provided by different units that may have inherent contradictions in subsequent collaborative decision-making.

[0051] Specifically, the method for conflict analysis of the perception confidence assessment values ​​of each collaborative unit corresponding to the perception target is as follows: extract the perception confidence assessment values ​​of each collaborative unit corresponding to the perception target, arrange the perception confidence assessment values ​​of all collaborative units corresponding to the perception target in descending order, and select the perception confidence assessment value ranked first as the perception confidence assessment benchmark value.

[0052] The difference between the perception confidence assessment value of each collaborative unit corresponding to the perception target and the perception confidence assessment benchmark value is calculated, and the absolute value of the difference is obtained by taking the absolute value.

[0053] The absolute value of the difference between the perception confidence evaluation values ​​of each collaborative unit corresponding to the perception target is compared with a preset difference threshold to obtain the conflict determination result of each collaborative unit corresponding to the perception target. The preset difference threshold includes a preset first difference threshold and a preset second difference threshold, and the preset first difference threshold is greater than the preset second difference threshold.

[0054] Specifically, the method for obtaining the conflict determination results of each collaborative unit corresponding to the perception target is as follows: the conflict determination results of each collaborative unit corresponding to the perception target include direct conflict, potential conflict, and no conflict.

[0055] If the absolute value of the difference between the perception confidence assessment values ​​of a certain collaborative unit corresponding to the perception target is higher than the preset first difference threshold, it is determined that there is a direct conflict between the collaborative units, and the conflict determination result of the collaborative unit corresponding to the perception target is marked as having a direct conflict.

[0056] If the absolute value of the difference in the perception confidence assessment value of a certain collaborative unit corresponding to the perception target is lower than or equal to a preset first difference threshold, and the absolute value of the difference in the perception confidence assessment value of the collaborative unit corresponding to the perception target is higher than a preset second difference threshold, then it is determined that the collaborative unit has a potential conflict, and the conflict determination result of the collaborative unit corresponding to the perception target is marked as having a potential conflict.

[0057] If the absolute value of the difference between the perception confidence assessment values ​​of a certain collaborative unit corresponding to the perception target is lower than or equal to a preset second difference threshold, it is determined that there is no conflict in the collaborative unit, and the conflict determination result of the collaborative unit corresponding to the perception target is marked as no conflict.

[0058] Specifically, the method for obtaining the decision weight vector of the perception target is as follows: extract the conflict judgment results of each collaborative unit corresponding to the perception target. If the conflict judgment result of a certain collaborative unit corresponding to the perception target is marked as having a direct conflict or a potential conflict, then the conflict factor of the collaborative unit corresponding to the perception target is matched according to the absolute value of the difference between the perception confidence evaluation values ​​of the collaborative unit corresponding to the perception target. The product of the conflict factor of the collaborative unit corresponding to the perception target and the perception confidence evaluation value is recorded as the decision weight value of the collaborative unit corresponding to the perception target.

[0059] It should be noted that the conflict factor of the collaborative unit corresponding to the perception target is obtained by matching the absolute value of the difference between the perception confidence evaluation values ​​of the collaborative unit corresponding to the perception target. Specifically, the absolute value of the difference between the perception confidence evaluation values ​​of the collaborative unit corresponding to the perception target is matched with the conflict factors corresponding to the absolute values ​​of the difference between each perception confidence evaluation value stored in the intelligent recognition perception database. The conflict factor corresponding to the absolute value of the difference between the perception confidence evaluation values ​​is queried and obtained, and is recorded as the conflict factor of the collaborative unit corresponding to the perception target.

[0060] It should be explained that the absolute value of the difference between the perceived confidence assessment values ​​directly quantifies the degree of deviation between the perceived conclusion of the collaborative unit and the conclusion of the most reliable collaborative unit. The larger the absolute value of the difference, the more significant the discrepancy between the perceived data and identification conclusion provided by this collaborative unit and the judgment of the most reliable source. When such units participate in collaborative decision-making, their data should be assigned a lower confidence weight, that is, matched with a smaller conflict factor, to reduce the influence of their original perceived confidence assessment value in the final decision, thereby preventing the abnormality or local cognition of individual units from excessively affecting the global judgment and maintaining the robustness of the system decision. Conversely, if the absolute value of the difference is smaller, it indicates that the perceived conclusion of the unit is highly consistent with the most reliable source. This consistency is a strong testament to the reliability of the unit's perception, even if its own absolute perceived confidence assessment value is not the highest. Therefore, this consensus should be fully trusted and utilized, and a larger conflict factor, usually close to 1, should be matched for it, so that its perceived data can retain most or even all of its influence in subsequent decisions, thereby strengthening the dominant position of high-quality consensus in collaborative decision-making and improving decision-making efficiency.

[0061] If the conflict determination result of a certain collaborative unit corresponding to the perception target is marked as no conflict, then the perception confidence evaluation value of the collaborative unit corresponding to the perception target is directly recorded as the decision weight value of the collaborative unit corresponding to the perception target. The decision weight values ​​of all collaborative units corresponding to the perception target are arranged in the order of the collaborative unit numbers to form the decision weight vector of the perception target.

[0062] Specifically, the method for obtaining the task urgency analysis index value of the sensing target by performing task urgency analysis is as follows: perform task urgency analysis on the operation status data of each collaborative unit corresponding to the sensing target to obtain the task urgency analysis index value of the sensing target. The task urgency analysis index value of the sensing target represents the quantitative result of the urgency of the task handling of the sensing target by the operation status data of each collaborative unit corresponding to the sensing target.

[0063] In this embodiment, the task urgency analysis index value of the perceived target can be obtained through the following analysis method, with the specific analysis conditions as follows: ; In the formula, The index value represents the task urgency analysis of the perceived target. This represents the remaining safe operation time for the i-th collaborative unit corresponding to the perceived target. This represents the urgency analysis factor corresponding to the set remaining safe working time per unit. This represents the estimated time for the i-th coordinating unit corresponding to the perceived target to reach the target. This indicates the urgency analysis factor corresponding to the estimated time for a unit to reach the target. This represents the target exposure duration of the i-th collaborative unit corresponding to the perceived target. This represents the urgency analysis factor corresponding to the set unit target exposure duration. This represents the average response time of historical task instructions for the i-th coordinating unit corresponding to the perceived target. The urgency analysis factor represents the average response delay time of the set unit historical task instructions. i represents the number of each collaborative unit, i=1, 2, 3, ..., n, n represents the total number of collaborative units, and e represents the natural constant.

[0064] It should be explained that the longer the remaining safe operation time, the more safe operation time the platform has. The less time pressure there is, the lower the task urgency analysis index value should be. The longer the estimated time to reach the target, the longer it takes for the platform to reach the target, which increases the risk of delay. The higher the task urgency analysis index value is because the platform needs to start earlier or the task may not be able to keep up. The longer the target exposure time, the longer the target has been discovered. The environment may have changed, the threat may have increased, and the task may need to be dealt with as soon as possible. Therefore, the task urgency analysis index value increases. The longer the average response delay time of historical task instructions, the slower the platform's response is. This may slow down the overall task progress. To compensate for this delay, the task should be arranged earlier or considered high-risk, and the task urgency analysis index value increases.

[0065] It should be added that, in this embodiment, the urgency analysis factors corresponding to the unit's remaining safe operation time, the unit's estimated time to reach the target, the unit's target exposure time, and the unit's historical average response delay time for task instructions are obtained from the intelligent identification and perception database. The urgency analysis factors corresponding to the unit's remaining safe operation time, the unit's estimated time to reach the target, the unit's target exposure time, and the unit's historical average response delay time for task instructions are used to adjust the importance of the data in the operation status data of each collaborative unit corresponding to the perceived target in the process of analyzing and obtaining the task urgency analysis index value.

[0066] In this implementation plan, the remaining safe operation time, estimated arrival time, target exposure time, and average response delay of historical task instructions for each collaborative unit corresponding to the perceived target are correlated and not independent. For example, the estimated arrival time is directly affected by the unit's current location and mobility, but its practical significance needs to be judged in conjunction with the remaining safe operation time. If the estimated arrival time is long and the unit's remaining safe operation time is relatively tight, it means that the effective time window for completing the task is extremely limited, and the urgency will increase sharply. At the same time, the target exposure time imposes time pressure from the task background level. The longer this time is, the greater the potential accumulation of uncertainty in the target's situation, which requires the system to respond faster. The comprehensive analysis of the task urgency index value of the perceived target can quantitatively assess the comprehensive timeliness pressure faced by the system in completing the identification and handling of the perceived target under the current dynamic constraints. This helps to provide data science basis for the collaborative system in scenarios with limited resources or concurrent targets, so as to dynamically determine the handling priority and resource input intensity among different targets and optimize the overall task efficiency.

[0067] Specifically, the method for obtaining the collaborative guidance instructions of the sensing target through dynamic collaborative regulation analysis is as follows: extract the maximum value of the decision weight from the decision weight vector of the sensing target, and denote it as the maximum decision weight value of the sensing target.

[0068] The task urgency analysis index value and maximum decision weight value of the perceived target are compared with the preset task urgency analysis index threshold and the preset maximum decision weight threshold, respectively, to obtain the collaborative guidance instruction for the perceived target.

[0069] Specifically, the method for obtaining the collaborative guidance instructions for the perceived target is as follows: the collaborative guidance instructions for the perceived target include the leading verification instruction, the collaborative re-verification instruction, and the continuous observation instruction.

[0070] If the task urgency analysis index value of the perceived target is higher than the preset task urgency analysis index threshold, and the maximum decision weight value of the perceived target is higher than the preset maximum decision weight threshold, then the collaborative guidance instruction of the perceived target will be marked as the dominant verification instruction.

[0071] If the task urgency analysis index value of the perceived target is higher than the preset task urgency analysis index threshold, and the maximum decision weight value of the perceived target is lower than or equal to the preset maximum decision weight threshold, then the collaborative guidance instruction of the perceived target will be marked as a collaborative verification instruction.

[0072] If the task urgency analysis index value of the perceived target is lower than or equal to the preset task urgency analysis index threshold, then the collaborative guidance instruction of the perceived target will be marked as a continuous observation instruction.

[0073] Specifically, the method for dynamically reconfiguring air-ground collaborative resources and planning action sequences is as follows: extract the collaborative guidance instructions of the sensing target, dynamically reconfigure air-ground collaborative resources and plan action sequences, dynamically schedule and reconfigure air-ground collaborative resources, optimize the task allocation of each collaborative unit corresponding to the sensing target, and generate the task allocation and action sequence of the sensing target.

[0074] It should be added that if the collaborative guidance command for the perceived target is marked as the dominant verification command, it indicates that the task is highly urgent and there is a relatively prominent conflict. Resource allocation and task planning should be carried out for the collaborative unit with the highest decision weight value. In terms of dynamic reallocation of air-ground collaborative resources, priority should be given to allocating various key resources to this collaborative unit. For example, if the collaborative unit is a high-altitude reconnaissance UAV, a preset number of high-performance optical lenses should be added, and the preset data transmission bandwidth should be increased to ensure rapid data return and ensure that it can stably execute the task. In terms of action sequence planning, the collaborative unit should be arranged to prioritize the execution of a comprehensive and in-depth verification task of the perceived target. For example, the high-altitude reconnaissance UAV should be arranged to quickly fly to the area of ​​the perceived target and use the newly added high-performance optical lenses to collect high-precision images of suspected unexploded ordnance. At the same time, relevant analysis software should be launched to process the collected data in real time and quickly determine the accurate information of the unexploded ordnance.

[0075] When the collaborative guidance instruction for the perceived target is a collaborative verification instruction, it means that the task is urgent but the degree of conflict is relatively low. At this time, the emphasis is on the cooperation between multiple collaborative units. When dynamically reallocating air-to-ground collaborative resources, resources are allocated relatively evenly according to the capabilities and needs of each collaborative unit. For example, more sophisticated detection equipment is provided for low-altitude detailed survey UAVs, and protective upgrades are provided for ground-based bomb disposal robots. At the same time, communication equipment for data interaction is allocated to each collaborative unit to ensure real-time information sharing. In terms of action sequence planning, each collaborative unit is arranged to verify the perceived target in turn. For example, the low-altitude detailed survey UAV first uses sophisticated detection equipment to conduct close-range detection of unexploded ordnance and shares the data with the ground-based bomb disposal robot in real time. Based on the received data and combined with its own protective upgrade equipment, the ground-based bomb disposal robot further approaches the unexploded ordnance for verification, and together they complete the verification task of the perceived target.

[0076] If the collaborative guidance command for the perceived target is a continuous observation command, it indicates that the mission urgency is low, and the focus is on the continuous monitoring of the perceived target. In terms of dynamic reallocation of air-ground collaborative resources, appropriate resources are allocated to each collaborative unit to maintain routine observation. For example, fuel is allocated to each UAV to meet its daily patrol needs, and a stable power supply and corresponding data storage devices are provided for ground equipment. In terms of action sequence planning, each collaborative unit is arranged to perform observation tasks periodically according to the routine observation mode. For example, high-altitude reconnaissance UAVs fly to the perceived target area at a fixed time every day to conduct routine patrols and record relevant data; ground monitoring equipment collects environmental data around the perceived target at regular intervals to continuously observe the perceived target and grasp its status changes.

[0077] It should be noted that the air-ground collaborative multi-source sensing and unexploded ordnance intelligent identification system also includes an intelligent identification and sensing database, which stores the first parameter set, the second parameter set, the third parameter set, and the fourth parameter set obtained by analyzing historical data.

[0078] The first parameter set includes the confidence evaluation factor corresponding to the unit sensing deviation coefficient, the confidence evaluation factor corresponding to the multipath effect interference coefficient, the confidence evaluation factor corresponding to the average signal-to-noise ratio of the sensing signal, and the confidence evaluation factor corresponding to the feature point matching rate.

[0079] The second parameter set includes a difference threshold, a first difference threshold, a preset second difference threshold, and a conflict factor corresponding to the absolute value of the difference between each perceived confidence assessment value.

[0080] The third parameter set includes the urgency analysis factor corresponding to the unit remaining safe operation time, the urgency analysis factor corresponding to the unit estimated time to reach the target, the urgency analysis factor corresponding to the unit target exposure time, and the urgency analysis factor corresponding to the unit average response delay time of historical task instructions.

[0081] The fourth parameter set includes the threshold for task urgency analysis indicators and the threshold for maximum decision weight.

[0082] It should be noted that, in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article, or apparatus.

[0083] The preferred embodiments of the present invention disclosed above are merely illustrative of the invention. These preferred embodiments do not exhaustively describe all details, nor do they limit the invention to the specific implementations described. Clearly, many modifications and variations can be made based on the content of this specification. This specification selects and specifically describes these embodiments to better explain the principles and practical applications of the invention, thereby enabling those skilled in the art to better understand and utilize the invention.

Claims

1. An air-ground collaborative multi-source sensing and unexploded ordnance intelligent identification system, characterized in that, include: The data acquisition module is used by the intelligent identification and sensing platform for unexploded ordnance to collect the sensing data of each collaborative unit corresponding to the sensing target in the air-ground collaborative multi-source sensing area. The sensing data of each collaborative unit corresponding to the sensing target includes the heterogeneous sensing data of each collaborative unit corresponding to the sensing target and the operating status data of each collaborative unit corresponding to the sensing target. The confidence assessment module is used to perform multi-source confidence processing based on the heterogeneous sensing data of each collaborative unit corresponding to the sensing target, and to obtain the sensing confidence assessment value of each collaborative unit corresponding to the sensing target. The conflict analysis module is used to perform conflict analysis on the perception confidence evaluation values ​​of each collaborative unit corresponding to the perception target, and obtain the decision weight vector of the perception target. The urgency analysis module is used to perform task urgency analysis on the operational status data of each collaborative unit corresponding to the sensing target to obtain the task urgency analysis index value of the sensing target. The instruction analysis module is used to analyze indicator values ​​based on the task urgency of the perceived target, combine them with the decision weight vector of the perceived target, and dynamically coordinate and control the analysis to obtain the coordinated guidance instructions for the perceived target, thereby carrying out dynamic reallocation of air-ground coordinated resources and planning and configuration of action sequences.

2. The air-ground collaborative multi-source sensing and unexploded ordnance intelligent identification system according to claim 1, characterized in that: The heterogeneous sensing data of each cooperative unit corresponding to the sensing target includes the sensing deviation coefficient, multipath effect interference coefficient, average signal-to-noise ratio of the sensing signal, and feature point matching rate of each cooperative unit corresponding to the sensing target. The operational status data of each collaborative unit corresponding to the perceived target includes the remaining safe operation time, estimated time to reach the target, target exposure time, and average response delay time of historical task instructions for each collaborative unit corresponding to the perceived target.

3. The air-ground collaborative multi-source sensing and unexploded ordnance intelligent identification system according to claim 1, characterized in that: The method for obtaining the perception confidence evaluation value of each collaborative unit corresponding to the perceived target is as follows: Multi-source confidence processing is performed on the heterogeneous sensing data of each collaborative unit corresponding to the sensing target to obtain the sensing confidence evaluation value of each collaborative unit corresponding to the sensing target. The sensing confidence evaluation value of each collaborative unit corresponding to the sensing target represents the quantitative result of the comprehensive reliability of the sensing data and recognition conclusion output by each collaborative unit for the sensing target by the heterogeneous sensing data of each collaborative unit corresponding to the sensing target.

4. The air-ground collaborative multi-source sensing and unexploded ordnance intelligent identification system according to claim 1, characterized in that: The method for conflict analysis of the perception confidence evaluation values ​​of each collaborative unit corresponding to the perception target is as follows: Extract the perception confidence assessment value of each collaborative unit corresponding to the perception target, arrange the perception confidence assessment values ​​of all collaborative units corresponding to the perception target in descending order, and select the perception confidence assessment value ranked first as the perception confidence assessment benchmark value. The difference between the perception confidence assessment value of each collaborative unit corresponding to the perception target and the perception confidence assessment benchmark value is calculated, and the absolute value of the difference is obtained by taking the absolute value. The absolute value of the difference between the perception confidence evaluation values ​​of each collaborative unit corresponding to the perception target is compared with a preset difference threshold to obtain the conflict determination result of each collaborative unit corresponding to the perception target. The preset difference threshold includes a preset first difference threshold and a preset second difference threshold, and the preset first difference threshold is greater than the preset second difference threshold.

5. The air-ground collaborative multi-source sensing and unexploded ordnance intelligent identification system according to claim 4, characterized in that: The method for obtaining the conflict determination results of each cooperative unit corresponding to the perceived target is as follows: The conflict determination results of each collaborative unit corresponding to the perception target include direct conflict, potential conflict, and no conflict. If the absolute value of the difference between the perception confidence assessment values ​​of a certain collaborative unit corresponding to the perception target is higher than the preset first difference threshold, it is determined that there is a direct conflict between the collaborative unit and the conflict determination result of the collaborative unit corresponding to the perception target is marked as having a direct conflict. If the absolute value of the difference between the perception confidence assessment values ​​of a certain collaborative unit corresponding to the perception target is lower than or equal to a preset first difference threshold, and the absolute value of the difference between the perception confidence assessment values ​​of the collaborative unit corresponding to the perception target is higher than a preset second difference threshold, then it is determined that the collaborative unit has a potential conflict, and the conflict determination result of the collaborative unit corresponding to the perception target is marked as having a potential conflict. If the absolute value of the difference between the perception confidence assessment values ​​of a certain collaborative unit corresponding to the perception target is lower than or equal to a preset second difference threshold, it is determined that there is no conflict in the collaborative unit, and the conflict determination result of the collaborative unit corresponding to the perception target is marked as no conflict.

6. The air-ground collaborative multi-source sensing and unexploded ordnance intelligent identification system according to claim 5, characterized in that: The method for obtaining the decision weight vector of the perceived target is as follows: Extract the conflict determination results of each collaborative unit corresponding to the perception target. If the conflict determination result of a certain collaborative unit corresponding to the perception target is marked as having a direct conflict or a potential conflict, then match the conflict factor of the collaborative unit corresponding to the perception target based on the absolute value of the difference between the perception confidence evaluation values ​​of the collaborative unit corresponding to the perception target. The product of the conflict factor of the collaborative unit corresponding to the perception target and the perception confidence evaluation value is recorded as the decision weight value of the collaborative unit corresponding to the perception target. If the conflict determination result of a certain collaborative unit corresponding to the perception target is marked as no conflict, then the perception confidence evaluation value of the collaborative unit corresponding to the perception target is directly recorded as the decision weight value of the collaborative unit corresponding to the perception target. The decision weight values ​​of all collaborative units corresponding to the perception target are arranged in the order of the collaborative unit numbers to form the decision weight vector of the perception target.

7. The air-ground collaborative multi-source sensing and unexploded ordnance intelligent identification system according to claim 1, characterized in that: The method for obtaining the task urgency analysis index value of the perceived target through task urgency analysis is as follows: The task urgency analysis index value of the sensing target is obtained by performing task urgency analysis on the operational status data of each collaborative unit corresponding to the sensing target. The task urgency analysis index value of the sensing target represents the quantitative result of the urgency of the task handling of the sensing target by the joint operational status data of each collaborative unit corresponding to the sensing target.

8. The air-ground collaborative multi-source sensing and unexploded ordnance intelligent identification system according to claim 1, characterized in that: The method for obtaining the collaborative guidance instructions for the perceived target through dynamic collaborative regulation analysis is as follows: Extract the maximum value of the decision weight from the decision weight vector of the perceived target, and denote it as the maximum decision weight value of the perceived target; The task urgency analysis index value and maximum decision weight value of the perceived target are compared with the preset task urgency analysis index threshold and the preset maximum decision weight threshold, respectively, to obtain the collaborative guidance instruction for the perceived target.

9. The air-ground collaborative multi-source sensing and unexploded ordnance intelligent identification system according to claim 8, characterized in that: The method for obtaining the collaborative guidance instructions for the perceived target is as follows: The collaborative guidance instructions for the perceived target include the leading verification instruction, the collaborative re-verification instruction, and the continuous observation instruction; If the task urgency analysis index value of the perceived target is higher than the preset task urgency analysis index threshold, and the maximum decision weight value of the perceived target is higher than the preset maximum decision weight threshold, then the collaborative guidance instruction of the perceived target will be marked as the dominant verification instruction. If the task urgency analysis index value of the perceived target is higher than the preset task urgency analysis index threshold, and the maximum decision weight value of the perceived target is lower than or equal to the preset maximum decision weight threshold, then the collaborative guidance instruction of the perceived target will be marked as a collaborative verification instruction. If the task urgency analysis index value of the perceived target is lower than or equal to the preset task urgency analysis index threshold, then the collaborative guidance instruction of the perceived target will be marked as a continuous observation instruction.

10. The air-ground collaborative multi-source sensing and unexploded ordnance intelligent identification system according to claim 9, characterized in that: The method for dynamically reallocating air-ground collaborative resources and planning and configuring action sequences is as follows: Extract the collaborative guidance instructions of the perceived target, dynamically reconfigure the air-ground collaborative resources and plan the action sequence, dynamically schedule and reconfigure the air-ground collaborative resources, optimize the task allocation of each collaborative unit corresponding to the perceived target, and generate the task allocation and action sequence of the perceived target.

Images