Radar guarantee method and device based on fusion algorithm

Abstract

The invention provides a radar guarantee method and device based on a fusion algorithm. The method comprises the following steps: updating historical operation data based on real-time operation data; determining a real-time monitoring threshold value based on the historical operation data and an abnormal value algorithm, and comparing the real-time operation data with the real-time monitoring threshold value to obtain a real-time monitoring result; performing health assessment on the real-time operation data and the historical operation data, and applying an analytic hierarchy process to obtain a health level; and performing fault diagnosis and fault prediction based on the health level to obtain a guarantee decision of the radar. According to the method provided by the invention, effective and timely radar state monitoring is realized through the real-time operation data and the historical operation data obtained by updating the real-time operation data. Radar guarantee tasks are divided into real-time monitoring, health assessment, fault diagnosis and fault prediction, and corresponding algorithms are used in each task stage, so that tight connection of state monitoring and guarantee tasks is realized, and a real-time and accurate radar guarantee scheme is provided for decision makers.

Description

Technical Field

[0001] The present invention relates to the technical field of equipment support, and in particular to a radar support method and device based on a fusion algorithm. Background Art

[0002] In the information age, the equipment maintenance and support process has a highly dynamic feature. For example, in the field of aviation detection, the coordination between radar status monitoring and maintenance tasks is closer and more extensive, and the distributed characteristics of business processing are more prominent, which leads to a more complex composition of radar support forces, and its structure and elements are also changing dynamically.

[0003] However, the current maintenance support system structure is relatively rigid, its decision-making chain is long and lacks necessary and efficient coordination, it is difficult to form an integrated support capability, and it is impossible to guarantee the dynamic generation of support strategies in a time-varying environment. Summary of the invention

[0004] The present invention provides a radar support method and device based on a fusion algorithm, which are used to solve the defect that the current maintenance support system structure in the prior art is relatively rigid and cannot guarantee the dynamic generation of support strategies in a time-varying environment.

[0005] The present invention provides a radar support method based on a fusion algorithm, comprising: Acquire real-time operation data of the radar, and update historical operation data based on the real-time operation data; Determine a real-time monitoring threshold based on the historical operation data, compare the real-time operation data with the real-time monitoring threshold, and obtain a real-time monitoring result of the radar; Based on the real-time monitoring results, applying a hierarchical analysis algorithm to the real-time operation data and the historical operation data to perform health assessment to obtain a health level of the radar; Fault diagnosis and fault prediction are performed on the radar based on the health level to obtain a radar protection decision.

[0006] According to a radar support method based on a fusion algorithm provided by the present invention, the real-time monitoring threshold is determined based on the historical operation data and the outlier algorithm, including: When the amount of the historical operation data reaches a preset condition, based on the data type of the real-time operation data, matching the historical operation data to obtain reference historical data; Based on the reference historical data and the outlier algorithm, a real-time monitoring threshold is calculated; The data type of the real-time operation data includes at least one of a parameter value, a parameter trend value, and a parameter change rate value.

[0007] According to a radar support method based on a fusion algorithm provided by the present invention, based on the real-time monitoring results, a hierarchical analysis algorithm is applied to the real-time operation data and the historical operation data to obtain the health level of the radar, including: In the case where the real-time monitoring result is abnormal, the real-time operation data is quantitatively scaled based on the historical operation data to obtain real-time key operation data; Applying the analytic hierarchy process to the real-time key operation data and the historical operation data to determine the influencing factors of each real-time key operation data; The health level is determined based on the influencing factors and the real-time key operating data.

[0008] According to a radar assurance method based on a fusion algorithm provided by the present invention, the radar fault diagnosis based on the health level and the long short-term memory neural network algorithm includes: Extracting data features of the real-time operation data; Based on the data features and the fault diagnosis model, performing fault diagnosis on the radar; The fault diagnosis model is constructed based on the long short-term memory neural network algorithm.

[0009] According to a radar assurance method based on a fusion algorithm provided by the present invention, the radar fault prediction based on the health level and the autoregressive differential moving average algorithm includes: generating predicted key operating parameters of the radar based on the historical operating data and the fault prediction model; Based on the comparison between the predicted key operating parameters and the real-time monitoring threshold, a fault prediction is performed on the radar; The fault prediction model is constructed based on the autoregressive difference moving average algorithm.

[0010] According to a radar support method based on a fusion algorithm provided by the present invention, the radar is diagnosed for fault based on the health level and the long short-term memory neural network algorithm, and the radar is predicted for fault based on the health level and the autoregressive difference moving average algorithm to obtain a radar support decision, and then the method includes: Performing performance analysis on the real-time monitoring results, the health level, and the support decision of the radar to obtain a performance analysis result; The performance analysis result includes at least one type of power, accuracy, anti-interference, functional status, and false alarm.

[0011] The present invention also provides a radar support device based on a fusion algorithm, comprising: An acquisition unit, which acquires real-time operation data of the radar, and updates historical operation data based on the real-time operation data; A status monitoring unit, which determines a real-time monitoring threshold based on the historical operation data and an outlier algorithm, compares the real-time operation data with the real-time monitoring threshold, and obtains a real-time monitoring result of the radar; A health assessment unit, based on the real-time monitoring result, applies a hierarchical analysis algorithm to the real-time operation data and the historical operation data to perform health assessment to obtain a health level of the radar; The support unit performs fault diagnosis on the radar based on the health level and the long short-term memory neural network algorithm, and performs fault prediction on the radar based on the health level and the autoregressive difference moving average algorithm to obtain a support decision for the radar.

[0012] The present invention also provides an electronic device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the program, a radar support method based on a fusion algorithm as described above is implemented.

[0013] The present invention also provides a non-transitory computer-readable storage medium having a computer program stored thereon. When the computer program is executed by a processor, the radar support method based on any one of the fusion algorithms described above is implemented.

[0014] The present invention also provides a computer program product, comprising a computer program, wherein when the computer program is executed by a processor, the radar support method based on any one of the fusion algorithms described above is implemented.

[0015] The radar support method and device based on the fusion algorithm provided by the present invention monitors the radar status in real time through real-time operation data and historical operation data updated based on the real-time operation data, thereby realizing effective and timely radar status monitoring. The radar support task is divided into real-time monitoring, health assessment, fault diagnosis and fault prediction, and the corresponding algorithm is used in each task stage to realize the close connection between status monitoring and support tasks, providing decision makers with real-time and accurate radar support task matching information and selection plans. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] In order to more clearly illustrate the technical solutions in the present invention or the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0017] Figure 1 It is one of the flow charts of the radar support method based on the fusion algorithm provided by the present invention; Figure 2This is the second flow chart of the radar support method based on the fusion algorithm provided by the present invention; Figure 3 It is a flow chart of the real-time status monitoring method of the radar provided by the present invention; Figure 4 It is a flow chart of a method for evaluating the health status of an antenna assembly provided by the present invention; Figure 5 It is a flow chart of the fault diagnosis method provided by the present invention; Figure 6 It is a structural schematic diagram of a radar support device based on a fusion algorithm provided by the present invention; Figure 7 It is a structural schematic diagram of the electronic device provided by the present invention. DETAILED DESCRIPTION

[0018] In order to make the purpose, technical solution and advantages of the present invention clearer, the technical solution of the present invention will be clearly and completely described below in conjunction with the drawings of the present invention. Obviously, the described embodiments are 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.

[0019] In view of the above problems, the present invention provides a radar support method based on a fusion algorithm, which closely connects status monitoring with support tasks through real-time monitoring of radar status, thereby achieving real-time and accurate radar support. Figure 1 FIG. 1 is one of the flow charts of the radar support method based on the fusion algorithm provided by the present invention, such as Figure 1 As shown, the method includes: Step 110, acquiring real-time operation data of the radar, and updating historical operation data based on the real-time operation data; Here, the real-time operation data of the radar may include the real-time operation parameters of the antenna component, the power component, and the processing component. For example, the voltage, current, temperature, and confidence of the antenna component, the power component, and the processing component. In addition, the historical operation data here is obtained by updating the acquired real-time operation data. The historical operation data here can dynamically reflect the characteristics of the historical operation data, and is not static, thereby providing a real-time reference for the current radar support mission. Moreover, the historical operation data may include normal operation data and fault operation data.

[0020] Specifically, the data type of the real-time operation data can be identified by acquiring the voltage, current, temperature, confidence and other working parameters of the antenna component, the power component and the processing component. Then, the acquired real-time operation data is used to update the stored historical operation data to obtain the current historical operation data.

[0021] Step 120, determining a real-time monitoring threshold based on the historical operation data and an outlier algorithm, comparing the real-time operation data with the real-time monitoring threshold, and obtaining a real-time monitoring result of the radar; Here, the monitoring threshold may be a threshold used to determine abnormal values ​​of real-time operation data.

[0022] Specifically, outlier monitoring can be performed through historical operating data and outlier algorithms to determine the current real-time monitoring threshold. Then, the real-time operating data can be state-monitored through the real-time monitoring threshold to obtain the real-time monitoring result of the radar. For example, the real-time operating data can be compared with the monitoring threshold, and the operating data that is not within the monitoring threshold range is determined as abnormal data, and the real-time monitoring result is abnormal; the operating data within the monitoring threshold range is determined as normal data, and the real-time monitoring result is normal. It can be understood that based on whether the real-time monitoring result is normal, the corresponding early warning signal is triggered. For example, when the radar temperature or vibration exceeds the threshold, the system can automatically send an alarm message to the maintenance personnel so that maintenance operations can be performed in time.

[0023] It should be noted that in the prior art, the monitoring threshold often relies on the experience of experts or equipment operation and maintenance personnel to specify a fixed value. This method cannot effectively reflect the dynamic changes of abnormal thresholds caused by real-time dynamic changes in state parameters, thereby causing the problem of false alarms and missed alarms. In the embodiment of the present invention, historical operation data is updated by real-time operation data, and then the monitoring threshold is obtained based on the dynamically updated historical operation data, which improves the effectiveness of the obtained monitoring threshold, and can achieve accurate and timely radar status monitoring, thereby enabling the subsequent radar support tasks to respond in a timely manner.

[0024] Step 130, based on the real-time monitoring result, applying a hierarchical analysis algorithm to the real-time operation data and the historical operation data to perform health assessment to obtain the health level of the radar; Here, the hierarchical analysis algorithm clarifies the goal of the decision-making problem and its related influencing factors, and divides the decision-making problem into a decision layer, an influencing factor layer and a quantitative analysis layer; the decision layer refers to the final decision goal, the influencing factor layer can have multiple levels to store multiple influencing factors that affect the decision goal, and the quantitative analysis layer quantifies the specific situation of the lowest-level influencing factors into numerical values.

[0025] Specifically, when there are risks in the real-time monitoring results, the health level of the radar can be obtained by applying the hierarchical analysis algorithm to the real-time operation data and the historical operation data for health assessment. The health level here can be divided into multiple levels, thereby providing auxiliary decision-making for the use and maintenance of the radar. It can be understood that the health level here can reflect the probability of abnormalities in each component of the radar. For example, the higher the health level, the lower the probability of abnormal conditions in the real-time operation data; the lower the health level, the higher the probability of abnormal conditions in the real-time operation data.

[0026] Step 140, performing fault diagnosis on the radar based on the health level and the long short-term memory neural network algorithm, and performing fault prediction on the radar based on the health level and the autoregressive difference moving average algorithm to obtain a radar protection decision.

[0027] Specifically, the radar can be diagnosed and predicted for faults through health levels and long short-term memory neural network algorithms. For example, when the health level is lower than a certain level, it means that the real-time operation data corresponding to the health level is abnormal, and then the fault diagnosis can be performed through the real-time operation data, historical operation data, and long short-term memory neural network algorithm. When the health level is within a certain range, it means that the real-time operation data corresponding to the health level may be abnormal in the future, and then the fault prediction can be performed through the real-time operation data, historical operation data, and autoregressive difference moving average algorithm. After obtaining the fault prediction result or the fault diagnosis result, the fault prediction result or the corresponding guarantee decision can be obtained based on the fault prediction and fault diagnosis results and the health level, and the fault diagnosis strategy of the radar system and the strategy for preventing the occurrence of faults can be obtained.

[0028] It should be noted that in practical applications, fault location and maintenance require a quick response. If the knowledge fusion process is too complex or time-consuming, it may affect the real-time and response speed of the support method, and thus fail to meet the needs of rapid maintenance. At the same time, it is impossible to predict the status of the equipment and respond quickly. The embodiment of the present invention improves the response speed of radar support by performing fault diagnosis based on a long short-term memory neural network algorithm after performing a health status assessment, and performing fault prediction based on an autoregressive difference moving average algorithm.

[0029] In one embodiment, Figure 2 FIG. 2 is a flow chart of the radar support method based on the fusion algorithm provided by the present invention. Figure 2As shown, the method includes: first, collecting radar equipment operation data for radar equipment to obtain real-time operation data. Then, radar status monitoring is performed based on the real-time operation data, and an early warning signal is output according to the real-time monitoring results. In addition, the radar historical usage record, radar historical fault data, and radar historical maintenance data can be used as radar basic historical data. Among them, the radar historical usage record can include information such as startup time, shutdown time, power, and surrounding environment; the radar historical fault data can include information such as fault time, fault type, fault cause, impact degree, fault part name, and fault part quantity; the radar historical maintenance data can include information such as repair data, repair data, return data, and repair return data.

[0030] Then, the radar operation data and radar historical data can be input into the radar health status assessment model, and the radar health status can be output through the radar health status assessment model. Then, radar fault prediction and radar fault diagnosis are performed based on the radar operation data and radar basic historical data. And according to the radar health status, radar fault prediction results and radar fault diagnosis results, radar maintenance suggestions are obtained.

[0031] It should be noted that, compared with the prior art, the radar support system is composed of multiple levels, and each level needs to be closely integrated and work together. As a result, the complexity of the support system increases the difficulty of development, deployment and maintenance, especially in achieving seamless integration between different equipment and technical standards. The radar support method based on the fusion algorithm provided by the present invention selects four dimensions: real-time status monitoring, health status assessment, fault diagnosis and fault prediction, to ensure the radar system, based on the real-time operation data of the radar, and uses the radar basic historical data as a benchmark for prediction, providing a radar predictive maintenance support method that integrates status monitoring warning signals, radar health status assessment and radar maintenance suggestions. In addition, the corresponding algorithms are applied in real-time status monitoring, health status assessment, fault diagnosis and fault prediction, respectively, to achieve radar support that integrates multiple algorithms, that is, to achieve easy-to-use, predictable radar status, fast response and clear levels of radar system support.

[0032] The method provided by the embodiment of the present invention performs real-time monitoring of the radar status through real-time operation data and historical operation data updated based on the real-time operation data, thereby realizing effective and timely radar status monitoring. The radar support task is divided into real-time monitoring, health assessment, fault diagnosis and fault prediction, and corresponding algorithms are used in each task stage to achieve a close connection between status monitoring and support tasks, providing decision makers with real-time and accurate radar support task matching information and selection plans.

[0033] Based on any of the above embodiments, in step 120, determining the real-time monitoring threshold based on the historical operation data and the outlier algorithm includes: When the amount of the historical operation data reaches a preset condition, based on the data type of the real-time operation data, matching the historical operation data to obtain reference historical data; Based on the reference historical data and the outlier algorithm, a real-time monitoring threshold is calculated; The data type of the real-time operation data includes at least one of a parameter value, a parameter trend value, and a parameter change rate value.

[0034] Here, the data type of the real-time operation data includes at least one of a parameter value, a parameter trend value, and a parameter change rate value.

[0035] Specifically, first, when the amount of historical operation data reaches a preset condition, the data type of the real-time operation data can be used to search for historical operation data of this data type from a database storing historical operation data, and the reference historical data can be obtained by matching. It should be noted that if the historical operation data of this data type does not exist in the database, the real-time operation data that is input will be searched and matched again after reaching the set amount of data, and the reference historical data can be obtained.

[0036] Then, by referring to the historical data and the outlier algorithm, the threshold interval for outlier determination under the data type corresponding to each real-time operation data can be calculated, that is, the real-time monitoring threshold can be obtained. The outlier algorithm here can be a 3sigma outlier algorithm.

[0037] It should be noted that the 3Sigma outlier algorithm is adopted to update the threshold adaptively. The sampling quantity of the 3Sigma algorithm is controlled by defining the “lower limit of sample size”. As the operation data is input in real time, the latest outlier threshold is calculated in real time by the 3Sigma algorithm to achieve the adaptive update of the real-time monitoring threshold, thereby realizing effective and timely radar status monitoring.

[0038] In one embodiment, Figure 3 It is a flow chart of the real-time status monitoring method of the radar provided by the present invention, such as Figure 3As shown, the method includes: first, when the radar system is turned on, start the detection task of a certain parameter. Then, read the configuration information and detection category of the parameter. Then, determine whether there is historical data of the parameter in the database. If it does not exist, wait until there is enough data, and then enter the adaptive threshold update process; if it exists, directly enter the adaptive threshold update process. In the adaptive threshold update process, first, read the current data. Then, according to the historical data and the outlier detection algorithm, update the threshold interval of the outlier. Then, through the threshold comparison, determine whether the current monitoring value is an outlier; if it is not an outlier, continue to read the current data and monitor the next current data; if it is an outlier, output the warning information, and then continue to read the current data and monitor the next current data. Until the real-time status monitoring is turned off, for example, the radar system can be stopped, or a stop status monitoring instruction is received. Then, save the parameter data of this operation, replace and update the historical data, and the program terminates.

[0039] Based on any of the above embodiments, step 130 includes: In the case where the real-time monitoring result is abnormal, the real-time operation data is quantitatively scaled based on the historical operation data to obtain real-time key operation data; Applying the analytic hierarchy process to the real-time key operation data and the historical operation data to determine the influencing factors of each real-time key operation data; The health level is determined based on the influencing factors and the real-time key operating data.

[0040] Specifically, when there are abnormalities in the real-time monitoring results, first, the key operating data reflecting the radar health status can be selected from the real-time operating data. Then, the current real-time key operating data can be scaled by applying the quantization function through the historical operating data to avoid the impact of the numerical differences between parameters of different dimensions on the health status assessment model. Usually, a semi-trapezoidal normalized quantization function is used to obtain the real-time key operating data.

[0041] Next, based on the hierarchical analysis algorithm and expert experience, the importance levels of each real-time key operation data can be compared, and an importance judgment matrix can be constructed to determine the weight of each real-time key operation data, that is, to obtain the influence factor of each real-time key operation data. Further, according to the influence factor and real-time key operation data of each real-time key operation data, the time correction function is used to correct them, and the corrected influence factor and real-time key operation data of each real-time key operation data are input into the health status assessment model, and the health status assessment model uses the health status level membership distribution function to calculate the membership of the overall health status level.

[0042] Finally, the confidence level is determined, and the health level of the equipment is determined according to the maximum membership principle through the basic credibility distribution function, so as to realize the health status assessment of the antenna component, power component, processing component and the radar as a whole. The health level here can be divided into five levels: I, II, III, IV, and V.

[0043] In one embodiment, Figure 4 is a flow chart of the method for evaluating the health status of an antenna assembly provided by the present invention, such as Figure 4 As shown, the method includes: first, determining the antenna component health assessment identification framework, that is, selecting a health status assessment model. Then, selecting key parameters that reflect the health status of the antenna component. Next, normalizing the parameter data, and then determining the weights of the key parameters after normalization based on the hierarchical analysis algorithm and normalizing them, and correcting them according to the time correction function to obtain the corrected key parameters. In addition, for the normalized key parameters, determining the health status level membership distribution function, and calculating the health status level membership. Then, by correcting the key parameters and the health status level membership, determine the confidence level, and then determine the health status level of the antenna component.

[0044] It should be noted that a hierarchical analysis is conducted on real-time key operating data and historical operating data. This theory clarifies the goals of decision-making problems and their related influencing factors, and divides decision-making problems into decision-making layer, influencing factor layer and quantitative analysis layer; the decision-making layer refers to the final decision goal, the influencing factor layer can have multiple levels to store multiple influencing factors that affect the decision goal, and the quantitative analysis layer quantifies the specific situation of the lowest-level influencing factors into numerical values.

[0045] In addition, by combining the historical operation data of the radar, considering multiple aspects such as uncertainty, multi-source information and mission requirements, and using various reasoning algorithms and intelligent models, the product health status is evaluated and predicted, and auxiliary decision-making is provided for radar use and maintenance.

[0046] Based on any of the above embodiments, the fault diagnosis of the radar based on the health level and the long short-term memory neural network algorithm includes: Extracting data features of the real-time operation data; Based on the data features and the fault diagnosis model, performing fault diagnosis on the radar; The fault diagnosis model is constructed based on the long short-term memory neural network algorithm.

[0047] Specifically, first, the data features of the real-time operation data can be extracted. Then, the data features are input into the fault diagnosis model, and the fault diagnosis results of the radar are output through the fault diagnosis model. Then, based on the fault diagnosis results, the guarantee decision corresponding to the fault diagnosis results can be determined. It should be noted that radar fault diagnosis includes three parts: antenna assembly, power supply assembly and processing assembly. Real-time monitoring of the working parameters such as voltage, current, temperature and CAN interface of the antenna assembly, power supply assembly and processing assembly, output component fault diagnosis results of the antenna assembly, power supply assembly and processing assembly, and the results include two parts: fault location and fault mode.

[0048] Here, the fault diagnosis model can be constructed based on the long short-term memory neural network algorithm, and the initial fault diagnosis model is trained by constructing a sample data set to obtain a fault diagnosis model in the application stage. It should be noted that the long short-term memory neural network theory follows the radar structure tree division and is divided into three parts: antenna component failure mode, power component failure mode, and processing component failure mode. The antenna component failure parts include the antenna power supply and the DC amplifier; the power supply component failure parts include the extension power supply and the thyristor rectifier; the processing component failure parts include the frequency synthesizer, signal processing unit, RF interface, automatic gain control, mixer, pulse modulator, and frequency synthesizer.

[0049] In the training phase, first, a sample data set is constructed. The historical operation data or simulation data can be obtained by acquiring and collecting the sensor data, usage records and maintenance history data of the radar. For example, the transmit power, receive sensitivity, noise coefficient, transmit / receive switching time, chip temperature, working environment temperature, output waveform quality, working voltage, working current, number of abnormal parameters, fault mode, normal state, and the initial sample data are obtained. Then, the initial sample data is preprocessed, including data cleaning, data labeling, etc., to obtain a sample data set. It should be noted that the data of normal operation and the data of failure in the initial sample data are removed by data cleaning to remove the dirty data with null values ​​or interference in the initial sample data, and the fault information in the data of failure is labeled by data. Finally, the initial fault diagnosis model is trained based on the constructed sample data set to obtain the final fault diagnosis model. During the model training process, the prediction effect of the model on the test data set after training can be evaluated, and the reasoning effect of the current fault diagnosis model can be evaluated using various classification model evaluation indicators such as accuracy, recall rate, confusion matrix, etc. After the fault diagnosis model reasoning effect reaches the expected level, the fault diagnosis model will be deployed on the equipment health management software to calculate the real-time operation data transmitted in real time to infer whether the current radar system has a fault, as well as the fault location and fault mode.

[0050] In one embodiment, Figure 5It is a flow chart of the fault diagnosis method provided by the present invention, such as Figure 5 As shown, the method includes a model training process and a model reasoning process. In the model training stage, first, data cleaning is performed through the historical operation data of the antenna component (marked fault mode). Then, feature extraction is performed on the data after data cleaning, and time series data processing is performed on the extracted data features. The processed data is used for model training. During the training process, the trained fault diagnosis model is evaluated. If the model capability does not meet expectations, the fault diagnosis model is continuously trained until the model capability meets expectations, and then the fault diagnosis model is applied to the reasoning stage.

[0051] In the model reasoning stage, the real-time operation data (antenna component BIT data) is input into the fault diagnosis model, and the fault diagnosis model outputs the current fault location and fault mode.

[0052] It should be noted that in the training phase, by systematically collecting radar sensor data, usage records and maintenance history data, the features related to the fault are extracted according to the fault mode and existing knowledge, and the appropriate feature subset is selected. The fault diagnosis model is established and verified and tuned using machine learning or expert system methods. In the application phase, the fault diagnosis is performed based on the real-time radar operation data using the trained fault diagnosis model, and the corresponding diagnosis report and suggestions are generated.

[0053] Based on any of the above embodiments, the fault prediction of the radar based on the health level and the autoregressive differential moving average algorithm includes: generating predicted key operating parameters of the radar based on the historical operating data and the fault prediction model; Based on the comparison between the predicted key operating parameters and the real-time monitoring threshold, a fault prediction is performed on the radar; The fault prediction model is constructed based on the autoregressive difference moving average algorithm.

[0054] The fault prediction model here is constructed based on the autoregressive differential moving average algorithm. Time series preprocessing includes two aspects of testing, stationarity test and white noise test. The time series that can be used for analysis and prediction of the fault prediction model constructed based on the autoregressive differential moving average algorithm must meet the condition that it is a stationary non-white noise series.

[0055] In addition, the predicted key operating parameters here can be used to reflect the values ​​that the historical operating data may reach in the future.

[0056] Specifically, first, a fault prediction model that matches the given time series process can be selected from the known fault prediction models. In addition, the Bayesian information criterion method is used to determine the model order of the fault diagnosis model and select the optimal model order. Then, the historical operation data is input into the fault prediction model, and the predicted key operating parameters of the radar are predicted by the fault prediction model. Next, the predicted key operating parameters are compared with the real-time monitoring threshold to determine whether there is an abnormality in the predicted key operating parameters and obtain the fault prediction results. Finally, based on the fault prediction results, the guarantee decision corresponding to the fault prediction results can be determined. The fault prediction results here include fault location and fault mode.

[0057] It should be noted that predictive maintenance decisions are based on radar operating parameters and fault diagnosis prediction models to predict when specific components will fail, prepare supporting support resources in advance, and arrange predictive maintenance support in a timely manner to ensure that the radar is always in good condition. Intelligent maintenance decisions can identify early signs of failure in advance and only perform maintenance when the radar is really needed, which can reduce maintenance costs while extending the service life of the radar, correctly predict possible catastrophic failures, and formulate appropriate strategies based on the prediction results to improve system safety.

[0058] Based on any of the above embodiments, step 140 includes: Performing performance analysis on the real-time monitoring results, the health level, and the support decision of the radar to obtain a performance analysis result; The performance analysis result includes at least one type of power, accuracy, anti-interference, functional status, and false alarm.

[0059] Specifically, real-time monitoring results, health levels, and radar support decisions can be input for performance analysis. The inputs here include main lobe rate gain, horizontal side lobe level, vertical side lobe level, horizontal beam width, vertical beam width, S-curve azimuth difference slope, S-curve pitch difference slope, and the number of abnormal parameters.

[0060] In addition, the performance analysis here mainly analyzes the performance of the radar from five indicators: power, accuracy, anti-interference, functional status, and false alarm. In detail, the power coefficient of the radar can be calculated by inputting the parameter S curve azimuth difference slope; the detection distance of the radar system is calculated by the radar equation, and the distance accuracy is calculated based on the detection distance; by calculating the anti-interference coefficient of the radar, the coefficient includes the suppression of ground clutter and the suppression of active interference. For ground clutter, the influence of antenna scanning, ground clutter and meteorological clutter is calculated respectively; for active interference, the suppression ability of narrowband active interference is calculated; the false alarm rate is dynamically adjusted through the constant false alarm rate algorithm to find a balance between "false alarm" and "missed detection", thereby improving the stability of the radar system. Finally, the performance analysis results of power, accuracy, anti-interference, false alarm and functional status are obtained.

[0061] It should be noted that a radar performance analysis model is established based on the current status detection value of the radar system, the number of abnormal parameters of the subsystem, the number of subsystem failures, and the subsystem health assessment score. Based on the analysis method of multi-performance dimension algorithm fusion, the performance analysis results of the radar in five dimensions: power, accuracy, anti-interference, functional status, and false alarm are obtained.

[0062] Based on any of the above embodiments, Figure 6 is a schematic diagram of the structure of the radar support device based on the fusion algorithm provided by the present invention, such as Figure 6 As shown, the device comprises: An acquisition unit 610 acquires real-time operation data of the radar, and updates historical operation data based on the real-time operation data; The state monitoring unit 620 determines a real-time monitoring threshold based on the historical operation data and an outlier algorithm, compares the real-time operation data with the real-time monitoring threshold, and obtains a real-time monitoring result of the radar; A health assessment unit 630 performs health assessment on the real-time operation data and the historical operation data by applying a hierarchical analysis algorithm based on the real-time monitoring result to obtain a health level of the radar; The protection unit 640 performs fault diagnosis on the radar based on the health level and the long short-term memory neural network algorithm, and performs fault prediction on the radar based on the health level and the autoregressive difference moving average algorithm to obtain a protection decision for the radar.

[0063] The device provided by the embodiment of the present invention performs real-time monitoring of the radar status through real-time operation data and historical operation data updated based on the real-time operation data, thereby realizing effective and timely radar status monitoring. The radar support task is divided into real-time monitoring, health assessment, fault diagnosis and fault prediction, and corresponding algorithms are used in each task stage to achieve a close connection between status monitoring and support tasks, providing decision makers with real-time and accurate radar support task matching information and selection plans.

[0064] Based on any of the above embodiments, the status monitoring unit is specifically used for: When the amount of the historical operation data reaches a preset condition, based on the data type of the real-time operation data, matching the historical operation data to obtain reference historical data; Based on the reference historical data and the outlier algorithm, a real-time monitoring threshold is calculated; The data type of the real-time operation data includes at least one of a parameter value, a parameter trend value, and a parameter change rate value.

[0065] Based on any of the above embodiments, the health assessment unit is specifically used for: In the case where the real-time monitoring result is abnormal, the real-time operation data is quantitatively scaled based on the historical operation data to obtain real-time key operation data; Applying the analytic hierarchy process to the real-time key operation data and the historical operation data to determine the influencing factors of each real-time key operation data; The health level is determined based on the influencing factors and the real-time key operating data.

[0066] Based on any of the above embodiments, the guarantee unit is specifically used for: Extracting data features of the real-time operation data; Based on the data features and the fault diagnosis model, performing fault diagnosis on the radar; The fault diagnosis model is constructed based on the long short-term memory neural network algorithm.

[0067] Based on any of the above embodiments, the guarantee unit is further specifically used for: generating predicted key operating parameters of the radar based on the historical operating data and the fault prediction model; Based on the comparison between the predicted key operating parameters and the real-time monitoring threshold, a fault prediction is performed on the radar; The fault prediction model is constructed based on the autoregressive difference moving average algorithm.

[0068] Based on any of the above embodiments, the device further includes an optimization unit, and the optimization unit is specifically used for: Performing performance analysis on the real-time monitoring results, the health level, and the support decision of the radar to obtain a performance analysis result; The performance analysis result includes at least one type of power, accuracy, anti-interference, functional status, and false alarm.

[0069] Figure 7 An example of a physical structure diagram of an electronic device is shown in FIG. Figure 7 As shown, the electronic device may include: a processor 710, a communication interface 720, a memory 730 and a communication bus 740, wherein the processor 710, the communication interface 720 and the memory 730 communicate with each other through the communication bus 740. The processor 710 may call the logic instructions in the memory 730 to execute the radar guarantee method based on the fusion algorithm, the method comprising: obtaining the real-time operation data of the radar, updating the historical operation data based on the real-time operation data; determining the real-time monitoring threshold based on the historical operation data and the outlier algorithm, comparing the real-time operation data with the real-time monitoring threshold, and obtaining the real-time monitoring result of the radar; based on the real-time monitoring result, applying the hierarchical analysis algorithm to the real-time operation data and the historical operation data to perform health assessment, and obtain the health level of the radar; performing fault diagnosis on the radar based on the health level and the long short-term memory neural network algorithm, and performing fault prediction on the radar based on the health level and the autoregressive difference moving average algorithm to obtain the guarantee decision of the radar.

[0070] In addition, the logic instructions in the above-mentioned memory 730 can be implemented in the form of a software functional unit and can be stored in a computer-readable storage medium when it is sold or used as an independent product. Based on this understanding, the technical solution of the present invention can be essentially or partly embodied in the form of a software product that contributes to the prior art. The computer software product is stored in a storage medium, including several instructions to enable a computer device (which can be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present invention. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), disk or optical disk, etc. Various media that can store program codes.

[0071] On the other hand, the present invention also provides a computer program product, which includes a computer program. The computer program can be stored on a non-transitory computer-readable storage medium. When the computer program is executed by a processor, the computer can execute the radar support method based on the fusion algorithm provided by the above methods, and the method includes: acquiring real-time operation data of the radar, and updating historical operation data based on the real-time operation data; determining a real-time monitoring threshold based on the historical operation data and an outlier algorithm, and comparing the real-time operation data with the real-time monitoring threshold to obtain a real-time monitoring result of the radar; based on the real-time monitoring result, applying a hierarchical analysis algorithm to perform a health assessment on the real-time operation data and the historical operation data to obtain a health level of the radar; diagnosing faults of the radar based on the health level and a long short-term memory neural network algorithm, and predicting faults of the radar based on the health level and an autoregressive difference moving average algorithm to obtain a support decision for the radar.

[0072] On the other hand, the present invention also provides a non-transitory computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, is implemented to execute the radar support method based on the fusion algorithm provided by the above-mentioned methods, the method comprising: acquiring real-time operation data of the radar, and updating historical operation data based on the real-time operation data; determining a real-time monitoring threshold based on the historical operation data and an outlier algorithm, and comparing the real-time operation data with the real-time monitoring threshold to obtain a real-time monitoring result of the radar; based on the real-time monitoring result, applying a hierarchical analysis algorithm to perform a health assessment on the real-time operation data and the historical operation data to obtain a health level of the radar; diagnosing faults of the radar based on the health level and a long short-term memory neural network algorithm, and predicting faults of the radar based on the health level and an autoregressive difference moving average algorithm to obtain a support decision for the radar.

[0073] The device embodiments described above are merely illustrative, wherein the units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they may be located in one place, or they may be distributed on multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the scheme of this embodiment. Ordinary technicians in this field can understand and implement it without paying creative labor.

[0074] Through the description of the above implementation methods, those skilled in the art can clearly understand that each implementation method can be implemented by means of software plus a necessary general hardware platform, and of course, can also be implemented by hardware. Based on this understanding, the above technical solution is essentially or the part that contributes to the prior art can be embodied in the form of a software product, and the computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, a disk, an optical disk, etc., including a number of instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods described in each embodiment or some parts of the embodiments.

[0075] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A radar support method based on fusion algorithm, characterized in that: include: Acquire real-time operation data of the radar, and update historical operation data based on the real-time operation data; Determine a real-time monitoring threshold based on the historical operation data and an outlier algorithm, compare the real-time operation data with the real-time monitoring threshold, and obtain a real-time monitoring result of the radar; Based on the real-time monitoring results, applying a hierarchical analysis algorithm to the real-time operation data and the historical operation data to perform health assessment to obtain a health level of the radar; The radar is diagnosed for fault based on the health level and the long short-term memory neural network algorithm, and the radar is predicted for fault based on the health level and the autoregressive difference moving average algorithm to obtain a radar protection decision.

2. The radar support method based on fusion algorithm according to claim 1, characterized in that: The determining of the real-time monitoring threshold based on the historical operation data and the outlier algorithm includes: When the amount of the historical operation data reaches a preset condition, based on the data type of the real-time operation data, matching the historical operation data to obtain reference historical data; Based on the reference historical data and the outlier algorithm, a real-time monitoring threshold is calculated; The data type of the real-time operation data includes at least one of a parameter value, a parameter trend value, and a parameter change rate value.

3. The radar support method based on fusion algorithm according to claim 2 is characterized in that: Based on the real-time monitoring results, applying a hierarchical analysis algorithm to the real-time operation data and the historical operation data to perform health assessment to obtain the health level of the radar includes: In the case where the real-time monitoring result is abnormal, the real-time operation data is quantitatively scaled based on the historical operation data to obtain real-time key operation data; Applying the analytic hierarchy process to the real-time key operation data and the historical operation data to determine the influencing factors of each real-time key operation data; The health level is determined based on the influencing factors and the real-time key operating data.

4. The radar support method based on fusion algorithm according to claim 3 is characterized in that: The radar fault diagnosis based on the health level and the long short-term memory neural network algorithm includes: Extracting data features of the real-time operation data; Based on the data features and the fault diagnosis model, performing fault diagnosis on the radar; The fault diagnosis model is constructed based on the long short-term memory neural network algorithm.

5. The radar support method based on fusion algorithm according to claim 3 is characterized in that: The radar fault prediction based on the health level and the autoregressive differential moving average algorithm includes: generating predicted key operating parameters of the radar based on the historical operating data and the fault prediction model; Based on the comparison between the predicted key operating parameters and the real-time monitoring threshold, a fault prediction is performed on the radar; The fault prediction model is constructed based on the autoregressive difference moving average algorithm.

6. The radar support method based on fusion algorithm according to any one of claims 1 to 5, characterized in that: The method further comprises: performing fault diagnosis on the radar based on the health level and the long short-term memory neural network algorithm, and performing fault prediction on the radar based on the health level and the autoregressive difference moving average algorithm to obtain a radar security decision, and then comprising: Performing performance analysis on the real-time monitoring results, the health level, and the support decision of the radar to obtain a performance analysis result; The performance analysis result includes at least one type of power, accuracy, anti-interference, functional status, and false alarm.

7. A radar support device based on a fusion algorithm, characterized in that: include: An acquisition unit, which acquires real-time operation data of the radar, and updates historical operation data based on the real-time operation data; A status monitoring unit, which determines a real-time monitoring threshold based on the historical operation data and an outlier algorithm, compares the real-time operation data with the real-time monitoring threshold, and obtains a real-time monitoring result of the radar; A health assessment unit, based on the real-time monitoring result, applies a hierarchical analysis algorithm to the real-time operation data and the historical operation data to perform health assessment to obtain a health level of the radar; The support unit performs fault diagnosis on the radar based on the health level and the long short-term memory neural network algorithm, and performs fault prediction on the radar based on the health level and the autoregressive difference moving average algorithm to obtain a support decision for the radar.

8. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein: When the processor executes the computer program, the radar support method based on the fusion algorithm as claimed in any one of claims 1 to 6 is implemented.

9. A non-transitory computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the radar support method based on the fusion algorithm as claimed in any one of claims 1 to 6 is implemented.

10. A computer program product, comprising a computer program, characterized in that When the computer program is executed by a processor, the radar support method based on the fusion algorithm as claimed in any one of claims 1 to 6 is implemented.

Images