Multi-unit distributed networking cooperative combat system in strong electromagnetic interference environment

Abstract

The invention relates to the field of electromagnetism, and discloses a multi-unit distributed networking cooperative combat system in a strong electromagnetic interference environment. The multi-unit distributed networking cooperative combat system comprises a sensor unit, a data fusion unit, a target state estimation unit, a multi-target tracking unit and a control and feedback unit, the invention also provides a multi-unit distributed networking cooperative combat method in the strong electromagnetic interference environment, which comprises the following steps of: receiving multi-sensor data and performing weighted fusion, performing target state prediction by adopting extended Kalman filtering, and performing multi-target tracking by combining joint probability data association and a multi-target Kalman filtering algorithm. And a data acquisition and fusion strategy is adjusted in real time according to the target state change, and the target tracking precision is optimized. Through multi-sensor data fusion, extended Kalman filtering, a multi-target tracking algorithm and a real-time feedback mechanism, the problems of single-sensor limitation and nonlinear target tracking are solved, the multi-target processing capability and the adaptive adjustment capability of the system are improved, and high-precision and stable target tracking is ensured.

Description

Technical Field

[0001] The present invention relates to the electromagnetic field, and in particular to a multi-unit distributed networking collaborative combat system in a strong electromagnetic interference environment. Background Art

[0002] In existing target tracking technology, a single type of sensor is usually relied upon for target monitoring and data collection. For example, although traditional electromagnetic spectrum sensors can capture the electromagnetic signals of the target at a long distance, their measurement accuracy will drop significantly in a strong electromagnetic interference environment. Acoustic sensors perform well in electromagnetic interference areas, but their effective monitoring distance is relatively short and they are greatly affected by environmental noise. Although optical sensors can provide accurate target images or thermal imaging, their performance is limited at night or in low light conditions. Existing technologies often find it difficult to address the limitations of such a single sensor and it is difficult to ensure that accurate target data can be obtained in various environments.

[0003] Traditional target state estimation methods mostly rely on the classic Kalman filter algorithm. Although it works well in some simple linear systems, the accuracy and stability of the Kalman filter algorithm are often insufficient when facing complex nonlinear target motion, resulting in large errors in the prediction and update of the target state. Existing technologies usually ignore the nonlinear factors that may appear in the target state estimation process, resulting in low tracking accuracy when processing high-speed and complex motion trajectories. In terms of multi-target tracking, most of the existing methods are optimized for a single target, and it is difficult to efficiently process the state estimation of multiple targets. In a multi-target environment, the target tracking algorithms used in the existing technology cannot effectively distinguish the trajectories between targets, and are prone to target confusion and misassociation. In particular, when the targets are close to each other or their trajectories are intertwined, traditional tracking methods usually fail, resulting in the system being unable to accurately track multiple targets. In addition, the existing control and feedback mechanism is relatively simple and lacks flexible adaptive adjustment functions; in an environment with strong electromagnetic interference, the system is often unable to timely adjust the data acquisition and processing strategies according to changes in the target state or environmental conditions; especially when the target enters a complex environment or an emergency situation, the existing technology cannot effectively perform dynamic adjustment of system parameters, which often leads to a decline in system performance and reduced target tracking accuracy; therefore, the present invention proposes a multi-unit distributed networking collaborative combat system in a strong electromagnetic interference environment to solve the shortcomings of the existing technology. Summary of the invention

[0004] In view of the deficiencies in the prior art, the present invention provides a multi-unit distributed networking collaborative combat system in a strong electromagnetic interference environment, which solves the problems of limited performance of a single sensor, low multi-target tracking accuracy and lack of system adaptability.

[0005] To achieve the above objectives, the present invention is implemented through the following technical solutions: a multi-unit distributed networking collaborative combat system under a strong electromagnetic interference environment, comprising: A sensor unit, used for collecting electromagnetic signals, acoustic signals and optical images of the target in real time, and transmitting the collected data to a data fusion unit; A data fusion unit is used to receive the measurement data of multiple sensor units, perform weighted fusion according to the signal quality of each sensor, output the position information of the target, and provide the position information to the target state estimation unit for further processing; A target state estimation unit is used to estimate the state of the target, including position, velocity and acceleration, by using a nonlinear filtering algorithm. The estimated state result will be passed as input to the multi-target tracking unit. The nonlinear filtering algorithm is an extended Kalman filter. A multi-target tracking unit is used to accurately track multiple targets through joint probabilistic data association and multi-target Kalman filtering algorithm, and feed back the tracking results to the control and feedback unit; The control and feedback unit is used to adjust the sensor data collection method and data fusion weight according to the changes in the target state, so as to optimize the target tracking accuracy and ensure the overall performance of the system.

[0006] Preferably, the sensor unit includes an electromagnetic spectrum sensor, an acoustic sensor and an optical sensor, and the sensors transmit data via a wireless communication network.

[0007] Preferably, the data fusion unit dynamically adjusts the weighting weights using an information entropy optimization method according to the entropy value of the data measured by each sensor.

[0008] Preferably, the target state estimation unit predicts the motion trajectory of each target based on a Kalman filter algorithm, and performs real-time estimation of the target position, velocity and acceleration in combination with data from different sensors.

[0009] Preferably, the multi-target tracking unit calculates the association probability of each target by a joint probabilistic data association method, and tracks multiple targets in combination with a multi-target Kalman filtering method, and updates the position and trajectory of the target in real time.

[0010] Preferably, the data fusion unit is based on the formula: ; in: is the maximum likelihood function; Indicates the number of sensors; Indicates that from Observation data of sensors; Indicates the system parameters that need to be optimized; At a given parameter Next, The conditional probability of the observation data of each sensor.

[0011] Preferably, the target state estimation unit uses an extended Kalman filter method, through the following formula: ; in, For the moment The target state prediction, is the control input, is a nonlinear state transfer function.

[0012] Preferably, the multi-target tracking unit uses a joint probability data association method, through the following formula: ; in, Indicates at time The target state; Indicates at time The set of observation data of all sensors; Indicates that from Observation data of sensors; Indicates the number of sensors; Indicates that given the Observation data of sensors Under the condition of The conditional probability of Indicates The probability of a sensor observing data.

[0013] Preferably, the data fusion unit uses the following formula according to the information entropy optimization algorithm: ; in, is the entropy value of sensor data, The probability distribution of the observed data, the data with lower entropy value is given higher weight.

[0014] The present invention also provides a multi-unit distributed networking collaborative combat method under a strong electromagnetic interference environment, comprising the following steps: S1, receiving target data collected by multiple sensor units, including electromagnetic signals, acoustic signals and optical signals; S2. Receive data from each sensor through a data fusion unit and perform weighting, wherein the weighting is based on the entropy value of each sensor data; S3, predicting the target state using the extended Kalman filter algorithm through the target state estimation unit, and updating it in real time in combination with the sensor data; S4, using a joint probability data association and a multi-target Kalman filter algorithm through a multi-target tracking unit to perform state association and position update of multiple targets; S5. According to the changes in the target state, feedback is provided to adjust the data collection method and data fusion weight to optimize the target tracking accuracy.

[0015] The present invention provides a multi-unit distributed networking collaborative combat system in a strong electromagnetic interference environment. It has the following beneficial effects: 1. The present invention adopts a technical solution of multiple sensor data fusion, including electromagnetic, acoustic and optical sensors, and combines the information entropy optimization algorithm to perform weighted processing on the data of each sensor; this solution ensures that the system can obtain high-quality target data in a variety of environments; compared with the traditional single sensor reliance solution, the present invention solves the limitations of a single sensor in a complex environment and effectively improves the accuracy of data fusion and the overall reliability of the system.

[0016] 2. The present invention adopts the extended Kalman filter (EKF) algorithm to predict and update the target state in real time; this technical solution can handle complex nonlinear target motion, so that the system can maintain high-precision target state estimation under high speed or complex motion trajectory; compared with the traditional Kalman filter method, the present invention effectively solves the problem of inaccurate state prediction in nonlinear systems and improves the stability and accuracy of target tracking.

[0017] 3. The present invention realizes accurate tracking and position updating of multiple targets through joint probabilistic data association (JPDA) and multi-target Kalman filtering (M-KF) algorithms; this technical solution can correctly distinguish targets in a multi-target environment, avoid misassociation, and ensure that the system can efficiently handle real-time status updates of multiple targets; compared with traditional single-target tracking technology, the present invention significantly improves the multi-target processing capability in complex environments and solves the interference and misjudgment problems that may occur during target tracking.

[0018] 4. The present invention introduces a control and feedback mechanism to adjust the data collection method and data fusion weight in real time to optimize the target tracking accuracy; the solution can automatically adjust the system parameters according to the changes in the target state and the different environmental conditions to ensure accurate target tracking in complex environments; compared with the system in the prior art that lacks dynamic adaptive adjustment function, the present invention greatly improves the adaptability and stability of the system in a dynamically changing environment. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] Figure 1 is a system architecture diagram of the present invention; Figure 2 The figure is a flow chart of the method of the present invention. DETAILED DESCRIPTION

[0020] The following will be combined with the drawings in the specification of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0021] See also Figure 1 The embodiment of the present invention provides a multi-unit distributed networking collaborative combat system in a strong electromagnetic interference environment, including: A sensor unit, used for collecting electromagnetic signals, acoustic signals and optical images of the target in real time, and transmitting the collected data to a data fusion unit; The sensor unit of the present invention is a vital part of the system. Its function is to collect relevant signal data of the target in real time and provide basic data support for subsequent data fusion, target state estimation and multi-target tracking. The core task of the sensor unit is to ensure that the system can obtain reliable and accurate target information through the joint work of various types of sensors in a strong electromagnetic interference environment. This module reduces interference as much as possible in a complex electromagnetic environment through the collaborative work of different types of sensors, thereby improving the monitoring capability and response speed of the system.

[0022] In this embodiment, the sensor unit is composed of multiple sensors, including electromagnetic spectrum sensors, acoustic sensors and optical sensors. These sensors are connected to the data fusion unit through a wireless communication network. In a strong electromagnetic interference environment, different types of sensors have different advantages and scopes of application. Working together can effectively make up for the limitations of various sensors.

[0023] In general, electromagnetic spectrum sensors are mainly used to capture electromagnetic signals emitted by targets; the sensor covers multiple frequency bands from low frequency to high frequency, and can obtain the signal strength of the target in the communication frequency band, radar frequency band, etc.; the advantage of electromagnetic spectrum sensors is that they can identify targets at a long distance and analyze the signal characteristics in electromagnetic wave propagation; however, electromagnetic spectrum sensors will be significantly affected in an environment with strong electromagnetic interference, and the interference source may affect their normal operation. Therefore, electromagnetic spectrum sensors need to work with other types of sensors to make up for their shortcomings.

[0024] Specifically, in some embodiments, the electromagnetic spectrum sensor can monitor the spectrum changes of electromagnetic waves in real time, analyze the changing patterns of electromagnetic wave signals, and provide information for target identification and positioning; the working frequency band of the sensor needs to be reasonably selected according to different scenarios and interference environments; for example, in the military field, the frequency band of the spectrum sensor can be designed specifically for specific targets (such as radar, communication systems, etc.).

[0025] As an option, acoustic sensors are suitable for capturing sound wave signals emitted by targets, especially in the detection of low-flying targets or potential targets. Unlike electromagnetic signals, acoustic signals can still maintain strong stability in environments with strong electromagnetic interference. Therefore, in environments with strong electromagnetic interference, acoustic sensors can effectively supplement the deficiencies of electromagnetic spectrum sensors.

[0026] In one possible implementation, the acoustic sensor can not only capture the target's sound wave signal, but also make a preliminary estimate of the target's direction, distance and other information; the acoustic sensor is particularly important for close-range monitoring of the target, especially when the target is at a long distance or covered by other interference signals. The acoustic sensor can effectively reduce the impact of electromagnetic interference by receiving low-frequency sound wave signals from the target.

[0027] Optical sensors monitor the image or thermal imaging of the target through visible light or infrared rays. The advantage of optical sensors is that they can provide accurate spatial position information of the target. For targets that cannot effectively obtain electromagnetic signals or sound wave signals under strong electromagnetic interference, optical sensors can play a vital complementary role. The operation of optical sensors is not affected by electromagnetic interference, especially within the visible range of the target, and can provide more accurate target positioning information.

[0028] In specific implementations, optical sensors can be divided into visible light sensors and infrared sensors; visible light sensors are usually used for target capture during the day or in low-interference environments, while infrared sensors can function at night or in conditions of strong electromagnetic interference; the combination of the two sensors can obtain more comprehensive image information of the target.

[0029] Specifically, the sensor unit works by using different types of sensors to work together and complement each other, thereby maximizing the accuracy of target detection and positioning in a complex electromagnetic environment; each sensor transmits data to the data fusion unit through a wireless communication system; after the data fusion unit receives the data from each sensor, it performs weighted fusion based on the sensor's signal quality and data characteristics.

[0030] In one possible implementation, the signal quality of the sensor will be dynamically evaluated based on the real-time measurement error, signal-to-noise ratio, etc.; sensor data with better signal quality will receive a higher weight in the fusion process; through the weighted fusion strategy, the system can effectively reduce the impact of erroneous data caused by electromagnetic interference on overall target recognition, ensuring high-precision positioning of the target.

[0031] For the sensor data fusion process, the information entropy optimization method used can adjust the weight of each sensor according to the data uncertainty of each sensor; at the formula level, the information entropy is set as ,in For the sensor at time The calculation formula of information entropy of collected data is: ; in, is the entropy value of sensor data, The probability distribution of observed data, the data with lower entropy value is given higher weight; information entropy It represents the uncertainty of the sensor data. The lower the information entropy, the higher the reliability of the sensor data, and accordingly, the greater the weight. This model plays an important role in the data fusion process, helping the system to adjust the data influence of each sensor in real time.

[0032] The sensor units transmit data through a wireless communication network, and these data need to be transmitted to the data fusion unit in an efficient and low-latency manner; in a strong electromagnetic interference environment, the data transmission system needs to have a certain anti-interference ability to ensure that even in a complex electromagnetic environment, the sensor data can still be transmitted in a timely manner and will not be distorted due to signal attenuation or interference; for this reason, the sensor unit in this embodiment adopts high-frequency signal transmission technology and is equipped with effective anti-interference coding technology to improve the reliability of data transmission.

[0033] A data fusion unit is used to receive the measurement data of multiple sensor units, perform weighted fusion according to the signal quality of each sensor, output the position information of the target, and provide the position information to the target state estimation unit for further processing; In the multi-unit distributed networked collaborative combat system under the aforementioned strong electromagnetic interference environment, the data fusion unit plays a vital role; its task is to receive data from different sensor units, effectively fuse these heterogeneous data, and generate accurate location information of the target and its dynamic state; the data fusion unit not only needs to process the noise and uncertainty in the sensor data, but also weight it according to the data quality of each sensor to improve the accuracy of the fusion result and the stability of the system.

[0034] In this embodiment, the core function of the data fusion unit is the weighted fusion of multi-sensor data; after receiving the data from the electromagnetic spectrum sensor, acoustic sensor and optical sensor, the unit performs real-time weighted fusion according to the signal quality of each sensor; in order to provide more accurate data output in a strong electromagnetic interference environment, the data fusion unit adopts a weighted strategy based on information entropy optimization; the information entropy optimization algorithm can adjust the weight of the sensor in the fusion process according to the reliability of the sensor data, thereby improving the anti-interference ability of the overall system.

[0035] In general, the data fusion process includes the following steps: Data reception and preprocessing: The data from each sensor will first go through the preprocessing stage, including denoising, filtering and other operations to reduce interference and errors in the data; the preprocessed data will enter the weighted fusion stage.

[0036] Information entropy calculation: The information entropy value is calculated for the measurement data of each sensor to evaluate the reliability and uncertainty of its data; the lower the information entropy, the higher the reliability of the sensor data.

[0037] Weighted fusion: During the weighted fusion process, sensor data will be weighted according to its information entropy value. Sensor data with larger weights will have a greater impact on the final result. The weighted data are summed to obtain the final target location information. Specifically, each time the data is fused, the weighting coefficient in the fusion process is calculated based on the entropy value of each sensor, and the data of each sensor are weighted and summed to obtain the accurate location of the target.

[0038] ; in, Represents the final target position information output by the data fusion unit; Indicates the number of sensors involved in data fusion; Indicates The weight of each sensor; Indicates The raw data of the sensor.

[0039] In some embodiments, the weighting strategy not only considers the signal quality of the sensor, but also incorporates the spatial distribution of the sensor; as an option, if some sensors are located in front of the target and other sensors are located to the side or behind the target, the data from the front sensors will receive a higher weight to ensure the accuracy of target positioning.

[0040] Specifically, the weight of the sensor It can be calculated by the following formula: ; in, Indicates The entropy value of sensor data, is a small constant to avoid division by zero errors. The meaning of this formula is that the smaller the entropy value, the greater the weight of the sensor data, and vice versa. Through this weighting strategy, the system can maintain a high fusion accuracy when facing sensor data of different qualities.

[0041] In a strong electromagnetic interference environment, the working state of the sensor may change, resulting in a decrease in the data quality of some sensors; for this reason, the data fusion unit in this embodiment can dynamically adjust its weight according to the real-time state of the sensor; in one possible implementation, when the signal-to-noise ratio of a sensor decreases or a failure occurs, the system will automatically reduce the weight of the sensor to avoid it affecting the data fusion result; this strategy can effectively prevent the overall performance of the system from being affected when a sensor is abnormal.

[0042] For example, if the data of an electromagnetic spectrum sensor is unreliable due to electromagnetic interference, the weight of the sensor will be dynamically adjusted to a lower value to reduce the impact on the final data fusion result; accordingly, other sensors with more reliable signals (such as acoustic sensors or optical sensors) will receive higher weights to ensure the stability and accuracy of the system.

[0043] The target position information after weighted fusion will be passed to the subsequent target state estimation unit; specifically, the data fusion unit will output the target's precise position, velocity, acceleration and other state information, and use this information as input for further processing and updating by the multi-target tracking unit.

[0044] A target state estimation unit is used to estimate the state of the target, including position, velocity and acceleration, by using a nonlinear filtering algorithm. The estimated state result will be passed as input to the multi-target tracking unit. The nonlinear filtering algorithm is an extended Kalman filter. In a multi-unit distributed networked collaborative combat system under a strong electromagnetic interference environment, the target state estimation unit plays a vital role; the core task of this unit is to estimate the target state based on the target information received from the data fusion unit, including key parameters such as the target's position, velocity, and acceleration; the target state estimation unit requires not only accurate prediction of the target's motion trajectory, but also real-time update of the target state based on actual observations; this process needs to ensure the high accuracy and stability of the target state estimation, especially in an environment with strong electromagnetic interference.

[0045] In this embodiment, the target state estimation unit adopts the extended Kalman filter (EKF) method to estimate the target state; the extended Kalman filter can effectively handle the target state prediction and update problems in nonlinear systems and is an important tool for estimating the target state in this system.

[0046] In general, the workflow of the target state estimation unit includes the following steps: State prediction: Based on the current state of the target and the control input, the future state of the target is predicted through the state transfer function; the state transfer function is usually a nonlinear model that describes the motion law of the target; assuming that the state of the target at time for , the control input is , then the target state prediction formula is: ; in, For the moment The target state prediction, is the control input, is a nonlinear state transfer function.

[0047] Forecast covariance update: In order to evaluate the uncertainty of the forecast state, it is necessary to calculate the forecast covariance matrix; the forecast covariance matrix Represents the error of the target state prediction; the error is transmitted through the state transfer matrix To update: ; in, time The prediction error covariance matrix represents the uncertainty of the error in the target state prediction; time The state transition matrix describes the transition from time To time The state change is usually determined by the target's motion model; time The estimated error covariance matrix represents the uncertainty of the target state estimate at the previous moment; State transition matrix The transposed matrix of The process noise covariance matrix represents the noise or uncertainty introduced by the target state in the prediction process.

[0048] Observation update: When the sensor observation data arrives, the target state estimation unit will update the target state; at this time, the system uses the Kalman gain to adjust the difference between the predicted state and the actual observation; assuming that the observation value received from the data fusion unit is , then the observation update formula is: ; in, time The updated target state estimate includes the target's position information, velocity, acceleration, etc. time The predicted target state estimation is based on the time The state estimate and predicted control input are calculated; Kalman gain, which represents the trade-off coefficient between the predicted value and the actual observed value, is used to update the estimated value and is usually calculated based on the predicted covariance matrix and the observed covariance matrix; time The observations are measured by sensors and contain actual information about the target, such as position and speed; The observation model represents the target state according to the prediction Compute the expected observation, which maps the state of the target to the observation space.

[0049] Update the covariance matrix: After completing the state update, the estimated error covariance matrix also needs to be updated; the updated covariance matrix Represents the uncertainty after the target state is updated: ; in, is the updated target state estimation error covariance matrix, at time The error of the estimated result; is the unit matrix, which is the same size as the matrix and represents the unchanged part; is the Kalman gain matrix, which represents the trade-off between observation information and prediction results. It determines the weight of the measured data in the estimation process. The observation matrix represents the linear relationship between the target state and the observation value (Jacobian matrix in Kalman filter); The predicted target state estimation error covariance matrix is The prediction error for the target state.

[0050] Specifically, in the target state estimation process, the key to the extended Kalman filter lies in the processing of nonlinear state transfer functions and observation models; since the target motion model usually has nonlinear characteristics, the extended Kalman filter handles nonlinear problems by linearizing the target state transfer function and observation model at each estimation.

[0051] In this embodiment, the nonlinear model of the target state is generally based on the motion equation of the target. For example, for a high-speed target, the following nonlinear motion model may be used: ; in, and The targets are -Axles and - axis position information, and For the purpose -Axles and -axis; the update formula of the target state takes into account nonlinear factors such as the acceleration or movement pattern of the target.

[0052] In the process of target state estimation, it is necessary not only to estimate the target motion, but also to consider the influence of process noise and observation noise. To this end, the process noise covariance matrix is ​​introduced in this embodiment: and the observation noise covariance matrix , represent the uncertainty in the target motion and the error in sensor observation, respectively.

[0053] In general, process noise It can be obtained through experiments or empirical estimates, reflecting the error of the target motion model; while the observation noise Adjustments are made according to different types of sensors, taking into account the measurement accuracy of different sensors.

[0054] In some embodiments, the target state estimation unit not only has to process the state estimation of a single target, but also needs to support simultaneous estimation of multiple targets; in multi-target tracking, the extended Kalman filter will maintain an independent state estimation process for each target; through the joint probabilistic data association (JPDA) method, the system can effectively distinguish between multiple targets and avoid the cross-influence of target state estimation.

[0055] A multi-target tracking unit is used to accurately track multiple targets through joint probabilistic data association and multi-target Kalman filtering algorithm, and feed back the tracking results to the control and feedback unit; In a multi-unit distributed networked collaborative combat system under a strong electromagnetic interference environment, the function of the multi-target tracking unit is crucial; this unit is responsible for simultaneously tracking multiple targets monitored by the system to ensure that the status of each target is updated in real time; after the data fusion unit and the target state estimation unit provide the target state information, the multi-target tracking unit uses this information to distinguish different targets, and matches the target state with the corresponding observation data through appropriate algorithms, thereby accurately tracking the motion trajectories of multiple targets.

[0056] In this embodiment, the multi-target tracking unit mainly relies on the joint probabilistic data association (JPDA) method and the multi-target Kalman filter (M-KF) algorithm to achieve accurate tracking of multiple targets; through these methods, target separation can be effectively performed between multiple targets to avoid target misassociation or loss.

[0057] Generally speaking, the states of multiple targets may be observed by multiple sensors at the same time. Especially when there is overlap in the trajectories of the targets or signal confusion, it becomes a challenge to correctly match the observed data with the target states. In this case, the JPDA method solves this problem by calculating the association probability between each target and the observation data of multiple sensors.

[0058] The key to JPDA is that the system calculates the probability of each possible target matching the observed data based on the sensor's observations and the target's predicted state. and observation set , the JPDA method calculates the joint probability of each pair of target and observation: ; in, Indicates at time The target state; Indicates at time The set of observation data of all sensors; Indicates that from Observation data of sensors; Indicates the number of sensors; Indicates that given the Observation data of sensors Under the condition of The conditional probability of Indicates The probability of a sensor observing data.

[0059] As an alternative, the advantage of the JPDA method is that it can simultaneously consider the association between multiple targets and multiple sensors, thereby effectively reducing the mismatch caused by trajectory overlap or similarity between targets; through this method, the system can independently and accurately track multiple targets in complex environments.

[0060] When multiple targets exist at the same time, the multi-target Kalman filter method can perform state estimation for each target independently; unlike traditional Kalman filtering, multi-target Kalman filtering assigns an independent state estimation process to each target; specifically, each target has an independent Kalman filter, which is updated by combining the target's state prediction and observation data.

[0061] In the specific implementation, the state estimation of each target includes information such as position, velocity, acceleration, etc., and the system is updated in real time based on the data from each sensor; specifically, the state update formula of each target can be obtained through the observation update formula in the previous steps: ; in, time The updated target state estimate includes the target's position information, velocity, acceleration, etc. time The predicted target state estimation is based on the time The state estimate and predicted control input are calculated; Kalman gain, which represents the trade-off coefficient between the predicted value and the actual observed value, is used to update the estimated value and is usually calculated based on the predicted covariance matrix and the observed covariance matrix; time The observations are measured by sensors and contain actual information about the target, such as position and speed; The observation model represents the target state according to the prediction Compute the expected observation, which maps the state of the target to the observation space.

[0062] Through multi-target Kalman filtering, the state of each target can be accurately estimated. Even when there is cross-interference or complex motion trajectories between multiple targets, the system can effectively distinguish each target and track and update them independently.

[0063] In some embodiments, in order to ensure that the tracking status of each target is correctly matched, the system uses a data association-based method to perform target allocation; specifically, when multiple sensors observe multiple targets at the same time, the system calculates the correlation between each sensor's observation data and the target state to ensure that the observation data of each target is correctly assigned to the corresponding target.

[0064] Generally speaking, data association methods use the predicted information of the target state and the observation results of the sensor to calculate the matching probability between each pair of target and observation data, and match the target and observation data according to the principle of maximum probability. Through this method, the system can effectively avoid misassociation and ensure that the state of each target is correctly estimated in a multi-target environment.

[0065] During multi-target tracking, as the target's motion trajectory is continuously updated, the system will adjust the tracking process of each target in real time to ensure that the target's status is continuously updated and corrected over time; for example, when a target undergoes drastic motion changes, the system will quickly update the target's status through the Kalman filter; when the target enters an area of ​​strong electromagnetic interference, the system will automatically make predictions based on existing data and adjust the target's tracking strategy.

[0066] In one possible implementation, the system will perform real-time multi-target status updates based on the predicted state and observed data of the target through a combination of joint probabilistic data association and Kalman filtering; specifically, the system can track multiple targets simultaneously and can handle complex situations such as sudden changes in target states and occlusions, thereby ensuring stable tracking of multiple targets.

[0067] As an option, when faced with complex multi-target scenarios, the system can also optimize the tracking process based on the dynamic changes of the target; for example, if the system finds that the movement pattern of a target has changed, or the trajectories of targets overlap, the system will adjust the Kalman filter parameters and data association strategy in real time to optimize the target tracking accuracy.

[0068] In addition, the system can also adjust the weight distribution of the target state estimation process according to real-time observations through a feedback mechanism; for example, when the data quality of a certain sensor is poor, the system will automatically reduce the weight of the sensor in the target tracking process and enhance the role of other sensors, thereby ensuring the stability and accuracy of the overall tracking effect.

[0069] The control and feedback unit is used to adjust the sensor data collection method and data fusion weight according to the change of the target state, so as to optimize the target tracking accuracy and ensure the overall performance of the system; In a multi-unit distributed networked collaborative combat system under a strong electromagnetic interference environment, the role of the control and feedback unit is crucial; it is responsible for dynamically adjusting the working strategy of the entire system according to the target status information provided by the multi-target tracking unit, ensuring that the system can operate stably and efficiently under complex and changing environmental conditions; in an environment with strong electromagnetic interference, the effectiveness and data quality of the sensor may be affected. The control and feedback unit needs to adjust the sensor weights in the data fusion process and optimize the target tracking strategy in real time according to the changes in the target status, so as to ensure the reliability and accuracy of the system.

[0070] In this embodiment, the control and feedback unit receives target state information from the multi-target tracking unit in real time and makes control decisions based on the dynamic changes of the target. The system continuously performs feedback adjustments during this process to ensure that the target tracking algorithm and sensor data processing can make the most optimized response under different environmental conditions.

[0071] In general, the working principle of the control and feedback unit includes the following main steps: Monitoring target status: The control and feedback unit receives target status data from the multi-target tracking unit, including the target's position, velocity, and acceleration; the system analyzes this data to determine the target's motion pattern, velocity changes, or possible interference effects.

[0072] Sensor data weighting adjustment: In a strong electromagnetic interference environment, some sensors may be greatly interfered with, resulting in a decrease in the quality of the data they collect; the control and feedback unit will dynamically adjust the weight of each sensor in the data fusion process to ensure that sensors with better signal quality have a greater impact in data fusion; for example, if the data quality of the electromagnetic sensor decreases, the system will reduce the data weight of the sensor and increase the weight of other sensors (such as acoustic or optical sensors).

[0073] Optimize data collection strategy: The control and feedback unit will also optimize the data collection strategy according to the state changes of the target. If the system detects that the target enters a complex or heavily interfered area, it may select a different sensor combination to enhance data reliability and tracking accuracy. Specifically, the system can temporarily shut down certain inefficient sensors or adjust the operating mode of the sensor to maximize the trackability of the target.

[0074] In this embodiment, the control and feedback unit adjusts the weight of the sensor data in real time based on the change of the target state; during the data fusion process, the system evaluates and adjusts the weight of each sensor according to the signal quality of the sensor, the dynamic change of the target and the change of the environment; for example, the system adjusts the weight according to the following factors: Sensor validity: If the sensor's measurement results are subject to significant interference (such as strong electromagnetic interference), the sensor's data will be given a lower weight.

[0075] Changes in target motion: If the target's velocity or trajectory changes significantly, the system will increase its reliance on sensors best suited to monitor that target (such as high-precision optical or acoustic sensors), thereby improving the accuracy of the target state estimate.

[0076] In this way, the control and feedback unit can adjust the data weight of each sensor according to the real-time target dynamic changes, environmental conditions and sensor performance to ensure the optimal performance of the system in different environments.

[0077] In some embodiments, the control and feedback unit adopts a strategy based on dynamic optimization to adapt to complex and unstable environmental conditions; when the target state changes, the control and feedback unit optimizes the target tracking process by adjusting system parameters and sensor selection; for example: Optimization adjustment when the target state changes: When the target enters an area with strong electromagnetic interference, the control and feedback unit will switch the sensor's operating mode, reduce the use of electromagnetic spectrum sensors, and increase the use of optical or acoustic sensors to ensure that the target can still be accurately tracked in an interference environment.

[0078] Response to emergencies: In some cases, when the speed or trajectory of a target changes suddenly, the system will dynamically adjust the parameters of the Kalman filter or the target tracking algorithm based on the target's changing trend to ensure that the system responds to the target more quickly and accurately.

[0079] The control and feedback unit in this embodiment has real-time feedback capability, and it can immediately adjust the system's operating strategy after the multi-target tracking unit provides a target status update; for example, the control and feedback unit will track the motion trajectory of each target in real time to ensure that the target's status estimate in the system can be updated in time; if a target has a large position prediction error, the control and feedback unit will adjust the parameters of the target tracking algorithm according to the target's motion changes and quickly correct the target state.

[0080] The control and feedback unit not only relies on simple parameter adjustments, but can also make complex adaptive adjustments based on real-time data and environmental feedback; in some embodiments, the control and feedback unit optimizes the system's working strategy in real time through an adaptive algorithm; for example, during the estimation of the target state, the system automatically adjusts the gain of the Kalman filter based on the target's motion changes and its estimation error, thereby improving the system's real-time adaptability to the target.

[0081] In an environment with strong electromagnetic interference, the stability of the system is particularly important. The control and feedback unit is designed with full consideration of various emergencies (such as sensor failure, target disappearance, sudden acceleration of the target, etc.). Through the real-time feedback mechanism, the control unit can maintain the stability of the system in these emergencies and avoid problems such as loss and misassociation during target tracking. Specifically, when the system encounters extreme situations, it will re-evaluate the sensor weights and data collection strategies based on real-time data to ensure continuous tracking of the target.

[0082] See also Figure 2 , a multi-unit distributed networking collaborative combat method under a strong electromagnetic interference environment includes the following steps: S1, receiving target data collected by multiple sensor units, including electromagnetic signals, acoustic signals and optical signals; In this step, the system receives data collected from different types of sensor units; the sensor units include electromagnetic spectrum sensors, acoustic sensors and optical sensors. The combination of these sensors enables the system to comprehensively obtain target information from multiple dimensions; the electromagnetic spectrum sensor is responsible for monitoring the electromagnetic signals emitted by the target and is particularly suitable for detecting long-range targets; the acoustic sensor captures the target's sound wave signals and can take advantage in environments with strong electromagnetic interference; the optical sensor provides accurate spatial position information for the target by acquiring images or thermal images of the target; through these three types of sensors, the system can obtain comprehensive and accurate target data in complex environments.

[0083] S2. Receive data from each sensor through a data fusion unit and perform weighting, wherein the weighting is based on the entropy value of each sensor data; After receiving the data from each sensor, the data fusion unit will weight the data according to the data quality of each sensor; the weighting basis is the entropy value of each sensor data. The lower the entropy value, the smaller the uncertainty of the data and the higher the data quality. Therefore, the data of this sensor will obtain a greater weight during fusion; the entropy value is calculated by evaluating the probability distribution of the sensor data. The lower the data entropy, the higher the reliability of the data. Through this weighting method, the system can give priority to the use of more reliable sensor data in the fusion process, thereby ensuring the accuracy and efficiency of the fusion results.

[0084] S3, predicting the target state using the extended Kalman filter algorithm through the target state estimation unit, and updating it in real time in combination with the sensor data; The target state estimation unit uses the extended Kalman filter (EKF) algorithm to predict the state of the target; the extended Kalman filter is suitable for nonlinear systems and can estimate the target's position information, velocity, acceleration and other states in real time; the process first predicts based on the current target state and motion model, and updates the prediction results in combination with data from the sensor; after each update, the target state estimation unit will output the updated target state, including the target's precise position and motion trajectory; this process can improve the system's tracking accuracy of the target state and ensure that the target's state can be accurately reflected even in a dynamic environment.

[0085] S4, using a joint probability data association and a multi-target Kalman filter algorithm through a multi-target tracking unit to perform state association and position update of multiple targets; In the multi-target tracking unit, the system uses joint probabilistic data association (JPDA) and multi-target Kalman filtering (M-KF) algorithms to deal with the state association problem of multiple targets; the JPDA method ensures that the observation data of multiple targets can be correctly matched and avoids misassociation between targets by calculating the association probability between each target and the observation data; for each target, the system independently estimates the state of the target through the multi-target Kalman filtering algorithm, and updates it in real time based on data from different sensors; in this way, the system can track multiple targets simultaneously and accurately update the position, speed and state of each target, ensuring effective monitoring of all targets in complex scenarios.

[0086] S5. According to the change of target state, feedback is provided to adjust the data collection method and data fusion weight to optimize the target tracking accuracy; Finally, the system adjusts the data collection method and data fusion weight in real time according to the changes in the target state and the target state information provided by the multi-target tracking unit; if the motion pattern of a target changes significantly or the target enters an area with strong interference, the control and feedback unit will adjust the data collection strategy and select sensors that are more suitable for the current environment and target characteristics for data collection; in addition, the system will dynamically adjust the weights of sensors during the data fusion process to optimize the target tracking accuracy; for example, in an area of ​​strong electromagnetic interference, the system may reduce the weight of electromagnetic spectrum sensors and increase the weight of optical sensors to ensure that the target can be tracked stably and accurately; this feedback adjustment mechanism ensures that the system can maintain efficient target tracking capabilities under different environmental conditions.

[0087] Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. A multi-unit distributed networking collaborative combat system under strong electromagnetic interference environment, characterized by: include: A sensor unit, used for collecting electromagnetic signals, acoustic signals and optical images of the target in real time, and transmitting the collected data to a data fusion unit; A data fusion unit is used to receive the measurement data of multiple sensor units, perform weighted fusion according to the signal quality of each sensor, output the position information of the target, and provide the position information to the target state estimation unit for further processing; A target state estimation unit is used to estimate the state of the target, including position, velocity and acceleration, by using a nonlinear filtering algorithm. The estimated state result will be passed as input to the multi-target tracking unit. The nonlinear filtering algorithm is an extended Kalman filter. A multi-target tracking unit is used to accurately track multiple targets through joint probabilistic data association and multi-target Kalman filtering algorithm, and feed back the tracking results to the control and feedback unit; The control and feedback unit is used to adjust the sensor data collection method and data fusion weight according to the changes in the target state, so as to optimize the target tracking accuracy and ensure the overall performance of the system.

2. The multi-unit distributed networking collaborative combat system in a strong electromagnetic interference environment according to claim 1 is characterized in that: The sensor unit includes an electromagnetic spectrum sensor, an acoustic sensor and an optical sensor, and the sensors transmit data via a wireless communication network.

3. The multi-unit distributed networking collaborative combat system in a strong electromagnetic interference environment according to claim 1 is characterized in that: The data fusion unit dynamically adjusts the weighting weights using an information entropy optimization method according to the entropy value of the data measured by each sensor.

4. The multi-unit distributed networking collaborative combat system in a strong electromagnetic interference environment according to claim 1, characterized in that: The target state estimation unit predicts the motion trajectory of each target based on the Kalman filter algorithm, and combines data from different sensors to perform real-time estimation of the target position, speed and acceleration.

5. The multi-unit distributed networking collaborative combat system in a strong electromagnetic interference environment according to claim 1, characterized in that: The multi-target tracking unit calculates the association probability of each target through a joint probability data association method, and tracks multiple targets in combination with a multi-target Kalman filtering method, and updates the position and trajectory of the target in real time.

6. The multi-unit distributed networking collaborative combat system in a strong electromagnetic interference environment according to claim 1, characterized in that: The data fusion unit is based on the formula: ； in: is the maximum likelihood function; Indicates the number of sensors; Indicates that from Observation data of sensors; Indicates the system parameters that need to be optimized; At a given parameter Next, The conditional probability of the observation data of each sensor.

7. The multi-unit distributed networking collaborative combat system in a strong electromagnetic interference environment according to claim 1, characterized in that: The target state estimation unit uses the extended Kalman filter method, through the following formula: ； in, For the moment The target state prediction, is the control input, is a nonlinear state transfer function.

8. The multi-unit distributed networking collaborative combat system in a strong electromagnetic interference environment according to claim 1, characterized in that: The multi-target tracking unit uses a joint probability data association method, through the following formula: ； in, Indicates at time The target state; Indicates at time The set of observation data of all sensors; Indicates that from Observation data of sensors; Indicates the number of sensors; Indicates that given the Observation data of sensors Under the condition of The conditional probability of Indicates The probability of a sensor observing data.

9. The multi-unit distributed networking collaborative combat system in a strong electromagnetic interference environment according to claim 1, characterized in that: The data fusion unit uses the following formula according to the information entropy optimization algorithm: ； in, is the entropy value of sensor data, The probability distribution of the observed data, the data with lower entropy value is given higher weight.

10. A multi-unit distributed networked collaborative combat method under a strong electromagnetic interference environment, applied to a multi-unit distributed networked collaborative combat system under a strong electromagnetic interference environment as claimed in any one of claims 1 to 9, characterized in that: The following steps are involved: S1, receiving target data collected by multiple sensor units, including electromagnetic signals, acoustic signals and optical signals; S2. Receive data from each sensor through a data fusion unit and perform weighting, wherein the weighting is based on the entropy value of each sensor data; S3, predicting the target state using the extended Kalman filter algorithm through the target state estimation unit, and updating it in real time in combination with the sensor data; S4, using a joint probability data association and a multi-target Kalman filter algorithm through a multi-target tracking unit to perform state association and position update of multiple targets; S5. According to the changes in the target state, feedback is provided to adjust the data collection method and data fusion weight to optimize the target tracking accuracy.

Images