A radar working mode recognition method and device with self-checking function

Abstract

The application provides a radar working mode recognition method and device with a self-checking function, and the method comprises the following steps: using a configured shortest path method to pre-recognize the working mode of a radar state sequence; determining a mode estimation result of matching between a candidate string and a verification string based on at least one of a configured active search matching algorithm or a normal matching algorithm; and recording a mode recognition result and a corresponding sliding window moving rule. The embodiment of the application can effectively avoid recognition errors caused by improper mode division, has strong recognition ability for the anti-reconnaissance means of 'one word with multiple modes', and the method provided by the embodiment of the application does not need to create training samples for each mode, can automatically correct recognition deviation by adaptively adjusting a matching strategy through a feedback loop, improves the calculation efficiency on the premise of ensuring the recognition effect.

Description

Technical Field

[0001] This invention relates to the field of radar operating mode recognition technology, and in particular to a radar operating mode recognition method and apparatus with self-verification function. Background Technology

[0002] In recent years, with the rise of the concept of cognitive electronic warfare and the widespread application of artificial intelligence technology, radar behavior pattern recognition has attracted increasing attention. One crucial aspect is radar operating mode recognition. This refers to the process of using radar signal parameter sequences to extract characteristic patterns of target radiation sources and classifying detected signals according to their operating modes. Radar operating mode recognition can provide a basis for estimating target threat levels and judging behavioral intentions. It is also the starting point and foundation for generating jamming strategies and analyzing their effects, and a key link in the transition from traditional confrontation to cognitive confrontation.

[0003] Radar pattern recognition is well-suited for model-driven artificial intelligence. Pattern-driven approaches require smaller training sample sizes, have clearly defined physical meanings for model parameters, and provide a clearer mapping from data extraction to behavioral cognition, facilitating the development of real-time and near-real-time pattern recognition algorithms. Visnevski N., *Syntactic Modeling of Multi-Function Radars* [D], McMaster University, 2005, proposed a syntactic structure model to describe radar waveform sequences, achieving a formal description of radar signal generation and changes. It uses a "radar word" composed of several pulses as the smallest unit of analysis and processing, with multiple "radar words" sequentially forming a "radar phrase" to achieve a basic function. Radar has several operating modes (such as search, interception, non-adaptive tracking, range resolution, and track-hold), each corresponding to a set of "radar phrases." When executing a certain operating mode, a "radar phrase" is randomly selected from the corresponding set to achieve the basic function. Thus, when the radar performs multi-functional tasks, the "radar phrases" are sequentially combined to form a "radar sentence." Radar operating mode recognition is the process of training a model using historical "radar sentences" and then inferring the operating mode through probability.

[0004] Littman M., Sutron R., Singh S., Predictive Representations of State [C]. Proc. of the Advances in Neural Information Processing Systems, 2002; James M., Wolfe BD, Singh S., Combining Memory and Landmarks with Predictive State Representations [C]. Proceedings of the International Joint Conference on Artificial Intelligence, 2005; A dynamic system model called Predictive State Representation (PSR) was proposed and improved. Compared with traditional Hidden Markov Models, PSR directly uses observable events as the object of statistical analysis, has strong representation ability, is intuitive and concise, and is very suitable for describing radar signals. Ou Jian, Research on Multifunctional Radar Behavior Identification and Prediction Technology [D], 2017, National University of Defense Technology; PSR was used for radar behavior pattern recognition, and a process algorithm from waveform sequence reception and processing to working mode recognition and estimation was designed. Chen Xiaozhou, Radar Operating Mode Recognition Algorithm Based on Predictive State Representation [Patent], 2023; This paper transforms mode estimation into a shortest path problem under the PSR framework and optimizes the training method, enabling overall training and recognition of state samples of mixed modes. However, the above algorithm has four problems, which reduce the algorithm efficiency and recognition effect, and also restrict the further practical application of the algorithm.

[0005] First, there is a poor ability to identify "one-word-multiple-mode" counter-reconnaissance sequences. "One-word-multiple-mode" refers to the overlap of radar phrase sets corresponding to different operating modes, meaning multiple operating modes share the same set of radar phrases. Taking the Mercury radar as an example, modes 2, 3, and 5 share the phrase "6666"; modes 3 and 4 share phrases such as "1666," "2666," "3666," "4666," and "5666." On the one hand, this waveform reuse helps the radar save resources and more efficiently perform diverse tasks such as search, interception, and tracking; on the other hand, "one-word-multiple-mode" can effectively confuse the opponent's electronic intelligence and electronic warfare support systems, effectively concealing the radar's true intentions and threat level. When the radar is detected using these phrases, traditional dictionary comparison alone will result in ambiguous identification results, requiring new dimensions for differentiation and identification.

[0006] Secondly, improper pattern segmentation leads to high recognition errors. When recognizing a reference sequence that spans multiple patterns, the sparse training samples can cause probability estimation bias, resulting in higher recognition errors. Therefore, the effectiveness of each recognition depends on the position of the sliding window; improper selection of the sliding window will produce recognition errors. However, when the reconnaissance party cannot know the starting point of each pattern, traditional algorithms cannot accurately determine the optimal position of the sliding window, inevitably causing recognition errors.

[0007] Third, errors are prone to accumulate during sequence recognition. Typically, recognizing the working mode at time t requires reference to time tl. H Working mode at all times because It is also an estimated value, and the estimation bias error will be passed on sequentially to... The recognition results are pending. If the algorithm lacks an automatic error correction mechanism, even small errors will accumulate and affect the recognition accuracy of subsequent sequences.

[0008] Fourth, efficiency and effectiveness are difficult to balance. To improve recognition accuracy, algorithms typically incorporate correlation, iteration, and verification steps to filter out false recognition results, which inevitably increases computational load and complexity. For example, the voting method proposed in existing technical literature only shifts the sliding window one position after each round of recognition, requiring multiple recognitions for each state. This sacrifices efficiency for improved recognition effectiveness, making it difficult to achieve both goals simultaneously. Therefore, algorithms with a higher cost-effectiveness ratio are needed. Summary of the Invention

[0009] The present invention provides a radar operating mode recognition method with self-verification function, comprising:

[0010] Step S1: Use the configured shortest path method to pre-identify the working mode of the radar state sequence to obtain the candidate string, verification string and corresponding pre-identification result in the state sequence, wherein the state sequence is a sequence composed of the radar states in the order of time steps.

[0011] Step S2: Based on at least one of the configured active search matching algorithm or normal matching algorithm, determine the matching status of the candidate string and the verification string pattern estimation results.

[0012] Step S3: Record the pattern recognition results and move the sliding window.

[0013] In one implementation, the specific process of step S1 is as follows: Figure 2 As shown, it includes:

[0014] Set the sliding window to tl H +1~t, denoted by sequence number i=1, where t is the current time step, l H This refers to the length of the sliding window;

[0015] Identifying state sequences using the shortest path method The recognition results are arranged into an ordered set based on the total path length, denoted as [Set Name]. The i-th recognition result Represented as Among them, S τ To represent the state at time step τ, use Indicates tl H A sequence of states within the time interval +1 to t, where the subscripts and superscripts represent the end and start time steps of the sequence, respectively, ω. τ For the pattern at time step τ, Indicates to The results of the recognition To An ordered set of recognition results;

[0016] if If the pattern in the data changes, record the change point k; if... If the pattern in the equation remains unchanged, then let k ← t, where k can take the following values: and Both are single-pattern sequences, and the patterns of the two sequences are different;

[0017] Define candidate strings as The verification string is and from The recognition results of the candidate string and the verification string are extracted from the data and denoted as follows: and

[0018] In one implementation, the specific process of the active search matching algorithm in step S2 is as follows: Figure 3 As shown, it includes:

[0019] Identify verification string The recognition results are arranged into an ordered set based on the total path length, denoted as [Set Name].

[0020] If PrevMatchFail = 1, then from The first pattern selected from the samples is equal to The identification results are then sorted from shortest to longest path, and the shortest path is selected as the longest. PrevMatchFail is a flag indicating that the previous round of matching failed. Its initial value is 0. If the current round of matching fails, the value is set to 1; otherwise, it is set to 0.

[0021] if If the set is not empty, then the match is successful. Set PrevMatchFail = 0, and then... Value as tl H +1~k+1+l H The result of time step identification, that is: Move the sliding window, and let t←k+1+2l H Begin the next round of pattern recognition;

[0022] if If the set is empty, the match fails, and step S3 is executed.

[0023] In one implementation, the specific process of the normal matching algorithm in step S2 is as follows: Figure 3 As shown, it includes:

[0024] Select The first recognition result is used as the verification pattern sequence, that is, let

[0025] make If sequence If the pattern changes no more than once and the pattern changes conform to the pattern transition rules, then the match is successful;

[0026] Set PrevMatchFail = 0, then set the sequence As tl H Based on the recognition results at time steps +1 to k, move the sliding window and let t←k+2+l H Then return to start the next round of recognition. Otherwise, if the match fails, proceed to step S3.

[0027] In one implementation, the specific process of step S3 is as follows: Figure 4 As shown, it includes:

[0028] Let i=i+1;

[0029] if Switch to the next candidate string and re-identify;

[0030] if If all candidate strings cannot be matched, set PrevMatchFail = 1 and record the result at time step tl. H The recognition result of +1, that is Move the sliding window, t←t+1, and return to step S1 to start the next round of recognition.

[0031] Another aspect of the present invention provides a radar operating mode recognition device with self-verification function, such as... Figure 5 As shown, it includes:

[0032] The pre-identification unit is configured to pre-identify the working mode of the radar state sequence using the configured shortest path method, wherein the state sequence is a sequence of radar states arranged in the order of time steps, to obtain candidate strings, verification strings and corresponding pre-identification results in the state sequence.

[0033] The matching unit is configured to determine the matching status of the candidate string and the verification string pattern estimation results based on at least one of the configuration-based active search matching algorithm or normal matching algorithm.

[0034] The post-processing unit is configured to record the pattern recognition results and move the sliding window.

[0035] In one embodiment, the pre-identification unit is further configured to:

[0036] Set the sliding window to tl H +1~t, denoted by sequence number i=1, where t is the current time step, l H This refers to the length of the sliding window;

[0037] Identifying state sequences using the shortest path method The recognition results are arranged into an ordered set based on the total path length, denoted as [Set Name]. The i-th recognition result Represented as Among them, S τ To represent the state at time step τ, use Indicates tl H A sequence of states within the time interval +1 to t, where the subscripts and superscripts represent the end and start time steps of the sequence, respectively, ω. τ For the pattern at time step τ, Indicates to The results of the recognition To An ordered set of recognition results;

[0038] if If the pattern in the data changes, record the change point k; if... If the pattern in the equation remains unchanged, then let k ← t, where k can take the following values: and Both are single-pattern sequences, and the patterns of the two sequences are different;

[0039] Define candidate strings as The verification string is and from The recognition results of the candidate string and the verification string are extracted from the data and denoted as follows: and

[0040] In one embodiment, the active search matching algorithm configured in the matching unit includes:

[0041] Identify verification string The recognition results are arranged into an ordered set based on the total path length, denoted as [Set Name].

[0042] If PrevMatchFail = 1, then from The first pattern selected from the samples is equal to The identification results are then sorted from shortest to longest path, and the shortest path is selected as the longest. PrevMatchFail is a flag indicating that the previous round of matching failed. Its initial value is 0. If the current round of matching fails, the value is set to 1; otherwise, it is set to 0.

[0043] if If the set is not empty, then the match is successful. Set PrevMatchFail = 0, and then... Value as tl H +1~k+1+l H The result of time step identification, that is: Move the sliding window, and let t←k+1+2l H Begin the next round of pattern recognition;

[0044] if If the set is empty, the matching fails, and the process is transferred to the post-processing unit.

[0045] In one implementation, the normal matching algorithm configured in the matching unit includes:

[0046] Select The first recognition result is used as the verification pattern sequence, that is, let

[0047] make If sequence If the pattern changes no more than once and the pattern changes conform to the pattern transition rules, then the match is successful;

[0048] Set PrevMatchFail = 0, then set the sequence As tl H Based on the recognition results at time steps +1 to k, move the sliding window and let t←k+2+l H Then, return to the starting point of the next round of recognition. Otherwise, if the match fails, proceed to the post-processing unit.

[0049] In one embodiment, the post-processing unit is further configured to:

[0050] Let i=i+1;

[0051] if Switch to the next candidate string and re-identify;

[0052] if If all candidate strings cannot be matched, set PrevMatchFail = 1 and record the result at time step tl. H The recognition result of +1, that is Move the sliding window, t←t+1, and return to step S1 to start the next round of recognition.

[0053] By adopting the above technical solution, the present invention has at least the following advantages:

[0054] 1) This invention adopts a "trial-test-adjust" approach to dynamically adjust the pattern division method, which can effectively avoid recognition errors caused by improper pattern division; it makes full use of the information matching of the preceding pattern to screen the pre-identification results, and has a strong recognition capability against "one word multiple patterns" anti-reconnaissance methods.

[0055] 2) The method provided by this invention does not require creating training samples separately for each mode. It adaptively adjusts the matching strategy through the "pre-identification-matching-sliding window movement" feedback loop, automatically corrects the identification deviation, and improves the computational efficiency while ensuring the identification effect. Attached Figure Description

[0056] Figure 1 This is a schematic flowchart of a radar operating mode recognition method with self-verification function according to an embodiment of the present invention;

[0057] Figure 2 This is a schematic diagram of the specific process of step S1 according to an embodiment of the present invention;

[0058] Figure 3 This is a schematic diagram of the specific process of the active search and matching algorithm in step S2 according to an embodiment of the present invention;

[0059] Figure 4 This is a schematic diagram of the specific process of the normal matching algorithm in step S2 according to an embodiment of the present invention;

[0060] Figure 5 This is a schematic diagram of the specific process of step S3 according to an embodiment of the present invention;

[0061] Figure 6 This is a structural diagram of a radar operating mode recognition device with self-verification function according to an embodiment of the present invention.

[0062] Figure 7 This is a schematic diagram of an electronic device structure according to an embodiment of the present invention. Detailed Implementation

[0063] To further illustrate the technical means and effects of the present invention in achieving its intended purpose, the present invention will be described in detail below with reference to the accompanying drawings and preferred embodiments.

[0064] In the accompanying drawings, the thickness, size, and shape of the objects have been slightly exaggerated for ease of illustration. The drawings are for illustrative purposes only and are not drawn to scale.

[0065] It should also be understood that the terms "comprising," "including," "having," "containing," and / or "comprising," when used in this specification, indicate the presence of the stated features, integrals, steps, operations, elements, and / or components, but do not exclude the presence or addition of one or more other features, integrals, steps, operations, elements, components, and / or combinations thereof. Furthermore, when expressions such as "at least one of..." appear after a list of listed features, they modify the entire listed feature, not individual elements in the list. Additionally, when describing embodiments of this application, the word "may" is used to mean "one or more embodiments of this application." And the term "exemplary" is intended to refer to an example or illustration.

[0066] As used herein, the terms “basically,” “approximately,” and similar terms are used as terms of approximation rather than terms of degree, and are intended to describe inherent biases in measured or calculated values ​​that will be recognized by those skilled in the art.

[0067] Unless otherwise specified, all terms used herein (including technical and scientific terms) shall have the same meaning as commonly understood by one of ordinary skill in the art to which this application pertains. It should also be understood that terms (e.g., those defined in common dictionaries) shall be interpreted as having the meaning consistent with their meaning in the context of the relevant art and shall not be interpreted in an idealized or overly formal sense unless expressly so specified herein.

[0068] It should be noted that, unless otherwise specified, the embodiments and features described in this application can be combined with each other. This application will now be described in detail with reference to the accompanying drawings and embodiments.

[0069] The first embodiment of the present invention provides a radar operating mode recognition method with self-verification function, such as... Figure 1 As shown, it includes the following steps:

[0070] Step S1: Use the configured shortest path method to pre-identify the working mode of the radar state sequence to obtain the candidate string, verification string and corresponding pre-identification result in the state sequence, wherein the state sequence is a sequence composed of the radar states in the order of time steps.

[0071] Step S2: Based on at least one of the configured active search matching algorithm or normal matching algorithm, determine the matching status of the candidate string and the verification string pattern estimation results.

[0072] Step S3: Record the pattern recognition results and move the sliding window.

[0073] The method provided by the present invention will be described in detail step by step below. For ease of understanding, the terms used in this embodiment will be explained below.

[0074] Time step τ: Each basic time unit is called a time step. In this algorithm, the time it takes for the radar to complete one radar word is called one time step.

[0075] The state S at time step τ τ At a certain time step, the observable and measurable parameter labels emitted by the radar are also referred to as observations. In this article, each radar term represents a state.

[0076] State sequence: A sequence of states arranged in order of time steps, where the subscripts and superscripts indicate the end and start time steps of the sequence, respectively. For example...

[0077] The pattern of the τth time step ω τ Each working mode corresponds to a radar phrase, which is randomly selected from the set of phrases corresponding to that mode.

[0078] Pattern sequence: A sequence of patterns over a period of time. The subscripts and superscripts represent the end and start time steps of the sequence, respectively.

[0079] Reference sequence: Calculate mode transition probabilities The state sequence on which the time is based

[0080] Sliding window: The time step corresponding to one identification sequence is called a "sliding window". In this application example, the length of the sliding window is l. H Therefore, a single point in time can be used to describe the range of the sliding window. For example, recognizing sequences. The result is Sliding windows can be described as tl H +1 to t+1, where t is used directly. After each round of recognition, the sliding window moves forward several time steps. In this application example, there are corresponding sliding windows in both the pre-recognition and verification string recognition processes, with the same meaning.

[0081] In this embodiment, reference Figure 2 Step S1 specifically includes:

[0082] Set the sliding window to tl H+1~t, denoted by sequence number i=1, where τ is the time step, and the time it takes for the radar to complete one radar word is called one time step. H t is the length of the sliding window and t is the preset time;

[0083] Identification using the shortest path method The recognition results are arranged into an ordered set based on the total path length, denoted as [Set Name]. The i-th recognition result Represented as Among them, S τ Let ω represent the state at time step τ. The state sequence is a sequence of states arranged in order of time steps, with its subscripts and superscripts indicating the end and start time steps of the sequence, respectively. τ Let τ be the pattern at time step τ. The pattern sequence is a sequence of patterns within a preset time period, and its subscripts and superscripts represent the end and start time steps of the sequence, respectively.

[0084] if If the pattern in the data changes, record the change point k; if... If the pattern in the equation remains unchanged, then let k ← t, where k can take the following values: and Both are single-pattern sequences, and the patterns of the two sequences are different;

[0085] Define candidate strings as The verification string is and from The recognition results of the candidate string and the verification string are extracted from the data and denoted as follows: and

[0086] In this embodiment, reference Figure 3 The active search and matching algorithm in step S2 specifically includes:

[0087] Identify verification string The recognition results are arranged into an ordered set based on the total path length, denoted as [Set Name].

[0088] If PrevMatchFail = 1, then from The first pattern selected from the samples is equal to The identification results are then sorted from shortest to longest path, and the shortest path is selected as the longest. Otherwise, proceed to step S3;

[0089] if If the set is not empty, then the match is successful. Set PrevMatchFail = 0, and then... Value as tl H +1~k+1+l HThe result of time step identification, that is to say Move the sliding window, and let t←k+1+2l H Begin the next round of pattern recognition;

[0090] if If the set is empty, the match fails, and step S3 is executed.

[0091] In this embodiment, reference Figure 4 The normal matching algorithm in step S2 specifically includes:

[0092] Select The first recognition result is used as the verification pattern sequence, that is, let

[0093] make If sequence If the pattern changes no more than once and the pattern changes conform to the pattern transition rules, then the match is successful;

[0094] Set PrevMatchFail = 0, then set the sequence As tl H Based on the recognition results at time steps +1 to k, move the sliding window and let t←k+2+l H Then return to start the next round of recognition. Otherwise, if the match fails, proceed to step S3.

[0095] In this embodiment, reference Figure 5 Step S3 specifically includes:

[0096] Let i=i+1;

[0097] if Switch to the next candidate string and re-identify;

[0098] if If all candidate strings cannot be matched, set PrevMatchFail = 1 and record the result at time step tl. H The recognition result of +1, that is Move the sliding window and return to step S1 to begin the next round of recognition.

[0099] Compared with the prior art, this embodiment has at least the following advantages:

[0100] 1) This invention adopts a "trial-test-adjust" approach to dynamically adjust the pattern division method, which can effectively avoid recognition errors caused by improper pattern division; it makes full use of the information matching of the preceding pattern to screen the pre-identification results, and has a strong recognition capability against "one word multiple patterns" anti-reconnaissance methods.

[0101] 2) The method provided by this invention does not require creating training samples separately for each mode. It adaptively adjusts the matching strategy through the "pre-identification-matching-sliding window movement" feedback loop, automatically corrects the identification deviation, and improves the computational efficiency while ensuring the identification effect.

[0102] The second embodiment of the present invention corresponds to the first embodiment, as described above. Figure 6 This embodiment introduces a radar operating mode recognition device with self-verification function, which includes the following components:

[0103] The pre-identification unit is configured to pre-identify the working mode of the radar state sequence using the configured shortest path method, wherein the state sequence is a sequence of radar states arranged in the order of time steps, to obtain candidate strings, verification strings and corresponding pre-identification results in the state sequence.

[0104] The matching unit is configured to determine the matching status of the candidate string and the verification string pattern estimation results based on at least one of the configuration-based active search matching algorithm or normal matching algorithm.

[0105] The post-processing unit is configured to record the pattern recognition results and move the sliding window.

[0106] In this embodiment, the pre-identification unit is further configured as follows:

[0107] Set the sliding window to tl H +1~t, denoted by sequence number i=1, where t is the current time step, l H This refers to the length of the sliding window;

[0108] Identifying state sequences using the shortest path method The recognition results are arranged into an ordered set based on the total path length, denoted as [Set Name]. The i-th recognition result Represented as Among them, S τ To represent the state at time step τ, use Indicates tl H A sequence of states within the time interval +1 to t, where the subscripts and superscripts represent the end and start time steps of the sequence, respectively, ω. τ For the pattern at time step τ, Indicates to The results of the recognition To An ordered set of recognition results;

[0109] if If the pattern in the data changes, record the change point k; if... If the pattern in the equation remains unchanged, then let k ← t, where k can take the following values: and Both are single-pattern sequences, and the patterns of the two sequences are different;

[0110] Define candidate strings as The verification string is and from The recognition results of the candidate string and the verification string are extracted from the data and denoted as follows: and

[0111] In this embodiment, the active search matching algorithm configured in the matching unit includes:

[0112] Identify verification string The recognition results are arranged into an ordered set based on the total path length, denoted as [Set Name].

[0113] If PrevMatchFail = 1, then from The first pattern selected from the samples is equal to The identification results are then sorted from shortest to longest path, and the shortest path is selected as the longest. PrevMatchFail is a flag indicating that the previous round of matching failed. Its initial value is 0. If the current round of matching fails, the value is set to 1; otherwise, it is set to 0.

[0114] if If the set is not empty, then the match is successful. Set PrevMatchFail = 0, and then... Value as tl H +1~k+1+l H The result of time step identification, that is: Move the sliding window, and let t←k+1+2l H Begin the next round of pattern recognition;

[0115] if If the set is empty, the matching fails, and the process is transferred to the post-processing unit.

[0116] In this embodiment, the normal matching algorithm configured in the matching unit includes:

[0117] Select The first recognition result is used as the verification pattern sequence, that is, let

[0118] make If sequence If the pattern changes no more than once and the pattern changes conform to the pattern transition rules, then the match is successful;

[0119] Set PrevMatchFail = 0, then set the sequence As tl H Based on the recognition results at time steps +1 to k, move the sliding window and let t←k+2+l H Then, return to the starting point of the next round of recognition. Otherwise, if the match fails, proceed to the post-processing unit.

[0120] In this embodiment, the post-processing unit is further configured as follows:

[0121] Let i=i+1;

[0122] if Switch to the next candidate string and re-identify;

[0123] if If all candidate strings cannot be matched, set PrevMatchFail = 1 and record the result at time step tl. H The recognition result of +1, that is Move the sliding window, t←t+1, and return to step S1 to start the next round of recognition.

[0124] A third embodiment of the present invention provides an electronic device, such as... Figure 7 As shown, it can be understood as a physical device, including a processor and a memory storing processor-executable instructions. When the instructions are executed by the processor, the following operations are performed:

[0125] Step S1: Use the configured shortest path method to pre-identify the working mode of the radar state sequence to obtain the candidate string, verification string and corresponding pre-identification result in the state sequence, wherein the state sequence is a sequence composed of the radar states in the order of time steps.

[0126] Step S2: Based on at least one of the configured active search matching algorithm or normal matching algorithm, determine the matching status of the candidate string and the verification string pattern estimation results.

[0127] Step S3: Record the pattern recognition results and move the sliding window.

[0128] In the fourth embodiment of the present invention, the process of the radar operating mode recognition method with self-verification function is the same as that of the first, second, or third embodiments. The difference lies in the engineering implementation: this embodiment can be implemented using software plus necessary general-purpose hardware platforms. While hardware implementation is also possible, the former is often a better approach. Based on this understanding, the method of the present invention can be embodied in the form of a computer software product stored in a storage medium (such as ROM / RAM, magnetic disk, optical disk), including several instructions to cause a device to execute the method of the present invention.

[0129] Through the description of specific embodiments, a more in-depth and specific understanding should be gained of the technical means and effects adopted by the present invention to achieve the intended purpose. However, the accompanying drawings are only for reference and illustration and are not intended to limit the present invention.

Claims

1. A radar operating mode recognition method with self-verification function, characterized in that, include: Step S1: Use the configured shortest path method to pre-identify the working mode of the radar state sequence to obtain the candidate string, verification string and corresponding pre-identification result in the state sequence, wherein the state sequence is a sequence composed of the radar states in the order of time steps. Step S2: Based on the configured active search matching algorithm or normal matching algorithm, determine the matching status of the candidate string and verification string pattern estimation results; Step S3: Record the pattern recognition results and move the sliding window; Step S1 includes: Set the sliding window to Record the serial number ,in, For the current time step, This refers to the length of the sliding window; Identifying state sequences using the shortest path method The recognition results are arranged into an ordered set based on the total path length, denoted as . Among them, the first One identification result Represented as Among them, using Indicates the first The state of the time step, using express A sequence of states within a time interval, where the subscripts and superscripts represent the end and start time steps of the sequence, respectively. Indicates to The results of the recognition To An ordered set of recognition results; if If the pattern changes, record the point of change. ;if If the pattern in the middle does not change, then let ,in, The value satisfies: and Both are single-pattern sequences, and the patterns of the two sequences are different; Define candidate strings as The verification string is and from The recognition results of the candidate string and the verification string are extracted from the data and denoted as follows: and .

2. The radar operating mode recognition method with self-verification function according to claim 1, characterized in that, In step S2, the active search matching algorithm includes: Identify verification string The recognition results are arranged into an ordered set based on the total path length, denoted as . ; If PrevMatchFail = 1, then from The first pattern selected from the samples is equal to The identification results are then sorted from shortest to longest path, and the shortest path is selected as the longest. PrevMatchFail is a flag indicating that the previous round of matching failed. Its initial value is 0. If the current round of matching fails, the value is set to 1; otherwise, it is set to 0. if If the set is not empty, the match is successful. Set PrevMatchFail=0 and change the sequence. Value as The result of time step identification, that is: Move the sliding window to make Begin the next round of pattern recognition; if If the set is empty, the match fails, and step S3 is executed.

3. The radar operating mode recognition method with self-verification function according to claim 2, characterized in that, In step S2, the normal matching algorithm includes: Select The first recognition result is used as the verification pattern sequence, that is, let ; make If the sequence If the pattern changes no more than once and the pattern changes conform to the pattern transition rules, then the match is successful; Set PrevMatchFail=0, then set the sequence As The time step recognition results are used to move the sliding window, allowing... If the match fails, proceed to step S3.

4. The radar operating mode recognition method with self-verification function according to claim 3, characterized in that, Step S3 includes: make ; if Switch to the next candidate string and re-identify; if If all candidate strings cannot be matched, set PrevMatchFail=1 and record the result at time step. The recognition result, i.e. Sliding window Then return to step S1 and begin the next round of recognition.

5. A radar operating mode recognition device with self-verification function, characterized in that, include: The pre-identification unit is configured to pre-identify the working mode of the radar state sequence using the configured shortest path method, wherein the state sequence is a sequence of radar states arranged in the order of time steps, to obtain candidate strings, verification strings and corresponding pre-identification results in the state sequence. The matching unit is configured to use either an active search matching algorithm or a normal matching algorithm based on the configuration to determine the matching status of the candidate string and the validation string pattern estimation results. The post-processing unit is configured to record the pattern recognition results and move the sliding window; The pre-identification unit is further configured as follows: Set the sliding window to Record the serial number ,in, For the current time step, This refers to the length of the sliding window; Identifying state sequences using the shortest path method The recognition results are arranged into an ordered set based on the total path length, denoted as . Among them, the first One identification result Represented as Among them, using Indicates the first The state of the time step, using express A sequence of states within a time interval, where the subscripts and superscripts represent the end and start time steps of the sequence, respectively. Indicates to The results of the recognition To An ordered set of recognition results; if If the pattern changes, record the point of change. ;if If the pattern in the middle does not change, then let ,in, The value satisfies: and Both are single-pattern sequences, and the patterns of the two sequences are different; Define candidate strings as The verification string is and from The recognition results of the candidate string and the verification string are extracted from the data and denoted as follows: and .

6. The radar operating mode recognition device with self-verification function according to claim 5, characterized in that, The active search matching algorithm configured in the matching unit includes: Identify verification string The recognition results are arranged into an ordered set based on the total path length, denoted as . ; If PrevMatchFail = 1, then from The first pattern selected from the samples is equal to The identification results are then sorted from shortest to longest path, and the shortest path is selected as the longest. PrevMatchFail is a flag indicating that the previous round of matching failed. Its initial value is 0. If the current round of matching fails, the value is set to 1; otherwise, it is set to 0. if If the set is not empty, the match is successful. Set PrevMatchFail=0 and change the sequence. Value as The result of time step identification, that is: Move the sliding window to make Begin the next round of pattern recognition; if If the set is empty, the matching fails, and the process is transferred to the post-processing unit.

7. The radar operating mode recognition device with self-verification function according to claim 6, characterized in that, The normal matching algorithm configured in the matching unit includes: Select The first recognition result is used as the verification pattern sequence, that is, let ; make If the sequence If the pattern changes no more than once and the pattern changes conform to the pattern transition rules, then the match is successful; Set PrevMatchFail=0, then set the sequence As The time step recognition results are used to move the sliding window, allowing... If the matching fails, the process returns to the beginning of the next round of identification; otherwise, the process returns to the post-processing unit.

8. The radar operating mode recognition device with self-verification function according to claim 7, characterized in that, The post-processing unit is further configured to: make ; if Switch to the next candidate string and re-identify; if If all candidate strings cannot be matched, set PrevMatchFail=1 and record the result at time step. The recognition result, i.e. Sliding window Then return to step S1 and begin the next round of recognition.

Images