Small phased array radar data real-time processing system

Abstract

The invention relates to the technical field of data processing, in particular to a small phased array radar data real-time processing system which comprises a jump positioning module, a frame sequence reconstruction module, a window mapping module, a branch binding module and a path synchronization module. In the invention, in the processing of a received frame sampling point sequence, a starting frame site is extracted through a hopping identification threshold value, accurate calibration of a channel and a time intersection point is realized, time base alignment and sequence adjustment are carried out on an out-of-order frame sequence in combination with an index, and a frame recording structure with traceability is constructed; the hopping frequency distribution characteristics of a sliding area are judged through a mapping mechanism, the dynamic response capability to a locking window is enhanced, a resolution threshold limit is introduced to effectively recognize a multi-target state and complete task branch binding, and path synchronization offset is recognized based on real-time sampling difference between locking frames. The timeliness of monitoring of the concurrent path state and tolerance judgment is improved, and stable synchronization of target information extraction and a dynamic path in a data processing link is effectively guaranteed.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of data processing, and particularly relates to a small phased array radar data real-time processing system. BACKGROUND

[0002] The technical field of data processing relates to operations such as collection, preprocessing, analysis and storage of original data, and its core matters include data acquisition, data cleaning, data conversion, feature extraction and data fusion, and is widely applied to aspects such as image recognition, radar imaging, communication signal processing and industrial detection. Among them, the phased array radar data processing system refers to a system that processes echo signals from multiple antenna array elements to realize target detection and tracking in radar detection applications, and is aimed at extracting useful target information from array data with high-speed sampling and high-dimensional features, and usually adopts a beam forming method based on fast Fourier transform, a matched filter output energy decision method, a fixed threshold detection mechanism and a motion target state estimation means based on Kalman filtering to complete the data processing process.

[0003] In the existing small phased array radar data real-time processing process, the target features are usually identified by relying on fixed threshold and matched filtering methods, and there is a lack of dynamic extraction means for the starting point of the sampling frame, which can easily lead to the inability to perform real-time calibration and repair when the array data frame sequence is disordered or shifted, and in the presence of multiple target signal superposition or frequent hopping, the fixed strategy is difficult to cover high-frequency change characteristics, which can easily cause channel positioning deviation and path state out-of-step, especially in scenes with large pulse repetition frequency variation or fast target motion, the fixed detection mechanism response lags, resulting in a decrease in the effective target information capture rate, increasing the overall system processing delay and misjudgment risk. SUMMARY

[0004] The purpose of the present application is to solve the shortcomings in the prior art and propose a small phased array radar data real-time processing system.

[0005] In order to achieve the above-mentioned purpose, the present application adopts the following technical scheme: a small phased array radar data real-time processing system comprises:

[0006] The hopping positioning module obtains the starting sampling point sequence of each channel receiving frame in the phased array radar antenna array, sequentially performs difference operation, extracts the corresponding starting frame position parameter when three consecutive groups exceed the receiving hopping identification threshold, confirms the corresponding channel-time intersection point position, marks the sliding positioning starting point, and obtains the channel hopping starting point set;

[0007] The frame sequence reconstruction module acquires the continuous starting sampling point value sequence of the corresponding channel based on the array channel and frame index information in the channel hopping starting point set, constructs a backtracking sliding frame record index sequence, and outputs a radar sliding record structure diagram.

[0008] The window mapping module performs in-window mapping positioning on each received frame number range according to a frame order sequence in the radar sliding record structure diagram, judges a sliding locking area, and generates a data sliding mapping path diagram;

[0009] The branch binding module obtains a first frame sampling point value and a second frame sampling point value in each frame segment in the data sliding mapping path diagram and calculates a difference amplitude, and if a continuous frame segment is greater than a multi-target resolution threshold value, binds a corresponding index as a current concurrent branch input locking frame, and generates a branch task binding node table;

[0010] The path synchronization module obtains a real-time difference sequence of the concurrent branch input locking frame based on the branch task binding node table, and if the real-time difference sequence exceeds a set tolerance threshold, triggers a path state pause instruction, re-enables an initial anchor positioning path, and generates a small phased array radar data real-time processing atlas.

[0011] As a further scheme of the application, the channel jump starting point set comprises a jump frequency distribution point, a starting frame site position parameter, and a channel-time intersection point, the radar sliding record structure diagram comprises a frame number index chain, a disordered frame adjustment mark, and a time base alignment label, the data sliding mapping path diagram comprises a sliding locking area identifier, a mapping window sequence, and a period step index, the branch task binding node table comprises a concurrent branch input locking frame index, a frame segment difference amplitude record, and a multi-target resolution judgment label, and the small phased array radar data real-time processing atlas comprises a path state update record, an initial anchor positioning path index, and a real-time difference tolerance judgment label.

[0012] As a further scheme of the application, the jump positioning module comprises:

[0013] The starting point extraction submodule obtains a starting sampling point sequence of each channel received frame in the phased array radar antenna array, time-sequentially sorts each group of channel frame header sampling points according to a frame time base index, groups channel sample sets according to array channel numbers, constructs a channel-time two-dimensional coordinate relationship, performs position mapping on the sorted starting sampling point sequence and the frame time base index, and generates a channel frame header starting position sequence;

[0014] The difference identification submodule sequentially performs a first-order difference operation on adjacent frame header sampling values based on the channel frame header starting position sequence, point-by-point compares the difference value sequence with a received jump identification threshold value, and marks sampling point positions exceeding the received jump identification threshold value, and generates a jump mutation candidate position set;

[0015] The jump point confirmation submodule judges whether the three groups of continuous sampling points exceed the receiving jump identification threshold value according to the jump mutation candidate position set, extracts the starting frame position parameter corresponding to the continuous sampling point section meeting the condition, combines the array channel number and the frame time base index to position the channel-time cross point, constructs the channel jump starting point set and marks it as the sliding positioning starting point, and generates the channel jump starting point set.

[0016] As a further scheme of the present application, the frame sequence reconstruction module comprises:

[0017] The jump frame acquisition submodule obtains the array channel number and the frame index parameter marked in the channel jump starting point set, extracts the starting sampling point value sequence of the corresponding channel at the continuous multiple frame time base positions according to each group of channel numbers, arranges the extracted starting sampling values in time sequence to construct the channel frame sampling flow data set, generates the channel frame sampling value sequence set in the structure of the continuous frame sampling point values grouped by channels;

[0018] The frame time base alignment submodule calculates the starting sampling point time index difference value between the continuous frames of each channel based on the channel frame sampling value sequence set, judges whether the adjacent frame index difference value meets the time base distance consistency reference value, constructs the frame disorder marking index list, rearranges the original frame sequence data in the time base sequence according to the frame disorder marking index list, adjusts the frame index relationship of the frame data of all channels after the rearrangement, and obtains the reconstructed time base sequence matrix;

[0019] The sliding record construction submodule constructs a two-dimensional sliding index graph structure according to all the channel and frame index cross points in the reconstructed time base sequence matrix, maps the channel frame sliding window distribution on the time axis in combination with the starting point sampling value sequence, establishes the sliding frame traceable record path structure in combination with the time distribution sequence, summarizes the sliding index path information of all channels, outputs the connection relationship between nodes and the channel frame source identification, and generates the radar sliding record structure graph.

[0020] As a further scheme of the present application, the window mapping module comprises:

[0021] The frame window positioning submodule extracts the sliding step index value of each channel node on the time axis based on the frame sequence in the radar sliding record structure graph, segments each channel frame index sequence according to the window step length, identifies the period section of each frame through the period index value, constructs a frame number-window period two-dimensional mapping table for each group of mapping results, and outputs the frame index distribution table in the window;

[0022] The frequency concentration judgment submodule counts the number of jump frames in each window according to the frame index distribution table in the window, extracts the distribution of the jump frequency in each frame index interval, calculates the frame section concentration index, judges whether it exceeds the concentration threshold value and performs the concentrated frame section judgment, and obtains the concentrated jump frame section identification sequence;

[0023] The lock path construction submodule filters out a corresponding frame index interval according to the centralized jump frame segment identification sequence, extracts a time path node sequence in a radar sliding record structure diagram, reads centralized frame segment path mapping information, establishes a mapping path set of jump frame segments and window periods, and generates a data sliding mapping path diagram.

[0024] As a further scheme of the present application, the branch binding module comprises:

[0025] The jump amplitude extraction submodule obtains sample point values of a first frame and a second frame of each frame segment in the data sliding mapping path diagram, extracts a starting sample amplitude pair between consecutive frame segments of a same channel, calculates an amplitude difference value and uniformly converts the amplitude difference value into a dB value, establishes a channel-frame segment amplitude difference value table, and generates a frame segment jump amplitude sequence;

[0026] The target distinguishable judgment submodule judges whether each group of difference values is greater than a multi-target distinguishable threshold value according to the frame segment jump amplitude sequence, marks frame segments with a difference value greater than the multi-target distinguishable threshold value as distinguishable effective frame segments, counts all index positions satisfying the condition, and obtains a distinguishable effective frame segment index set;

[0027] The task node binding submodule takes a starting index of each group of frame segments satisfying the condition as an input node index number based on the distinguishable effective frame segment index set, associates a current frame segment belonging to a sliding path ID, constructs a task concurrent channel mapping, and synchronously marks as a lock task input frame, and establishes a branch task binding node table.

[0028] As a further scheme of the present application, the path synchronization module comprises:

[0029] The sampling difference monitoring submodule obtains lock frame indexes in the branch task binding node table, extracts first sample point values of a current concurrent branch input lock frame and a corresponding initial binding frame, calculates an amplitude difference between the two groups of sample point values in a time axis order, forms a corresponding mapping of a frame pair sequence and an amplitude difference sequence, and obtains a lock frame difference amplitude sequence;

[0030] The tolerance step loss judgment submodule judges whether there is a frame pair exceeding a set tolerance threshold limit according to the lock frame difference amplitude sequence, extracts frame segment indexes greater than the tolerance threshold limit in the difference value sequence, marks a state as step loss, records a path number and a backtracking target for each group of step loss frame segments, and generates a step loss path identification list;

[0031] The state backtracking update submodule sets a related path state as pause based on the step loss path identification list, backtracks to an original binding frame segment, redefines a target path as an initial anchor position path, updates a path state information record table in a time index update period, marks a state field change flag, and establishes a radar data processing diagram structure table.

[0032] Compared with the prior art, the advantages and positive effects of the present invention are:

[0033] In the present invention, during the sequence processing of the received frame sampling points, the starting frame position is extracted by the jump recognition threshold to achieve accurate calibration of the channel and time intersection point, and the time base alignment and sequence adjustment of the disordered frame sequence are combined with the index to construct a frame record structure with traceability. The jump frequency distribution characteristics of the sliding area are determined by the mapping mechanism, and the dynamic response capability of the locking window is enhanced. The resolution threshold limit is introduced in the frame segment difference judgment, and the multi-target state is effectively identified and the task branch binding is completed. The path synchronization offset is identified based on the real-time sampling difference between the locked frames, and the timeliness of the monitoring of the concurrent path state and the tolerance judgment is improved. The backtracking reset of the sliding path and the accurate recording of the periodic state are realized, and the continuous extraction of target information in the data processing link and the stable synchronization of the dynamic path are effectively guaranteed. BRIEF DESCRIPTION OF THE DRAWINGS

[0034] Figure 1 is a system flow chart of the present invention;

[0035] Figure 2 This is a flow chart of the jump positioning module of the present invention;

[0036] Figure 3 This is a flow chart of the frame sequence reconstruction module of the present invention;

[0037] Figure 4 This is a flow chart of the window mapping module of the present invention;

[0038] Figure 5 This is a flow chart of the branch binding module of the present invention;

[0039] Figure 6 This is a flow chart of the path synchronization module of the present invention. DETAILED DESCRIPTION

[0040] In order to make the purpose, technical solutions and advantages of the present invention more clearly understood, the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0041] In the description of the present application, it should be understood that the terms "length", "width", "upper", "lower", "front", "back", "left", "right", "vertical", "horizontal", "top", "bottom", "inner", "outer" and the like indicate the orientation or positional relationship based on the orientation or positional relationship shown in the drawings, and are only for the convenience of describing the present application and simplifying the description, and do not indicate or imply that the devices or elements referred to must have a particular orientation, be constructed and operated in a particular orientation, and therefore cannot be understood as a limitation on the present application. In addition, in the description of the present application, "a plurality of" means two or more, unless otherwise explicitly specified.

[0042] Please refer to Figure 1 The small phased array radar data real-time processing system comprises:

[0043] The jump positioning module obtains the starting sample point sequence of each channel receiving frame in the phased array radar antenna array, sequentially performs difference operation on adjacent frame header sample values, compares with the receiving jump identification threshold value (3 times the standard deviation threshold value of noise floor: a commonly used noise suppression threshold value in radar signal processing, by calculating the standard deviation (σ) of background noise, setting 3σ as the effective signal detection threshold, ensuring 99.7% confidence), and extracts the corresponding starting frame position parameter when three groups of continuous values exceed the receiving jump identification threshold value, confirms the corresponding channel-time intersection point position combining the array channel number and frame time base index, marks as the sliding positioning starting point, and obtains the channel jump starting point set;

[0044] The frame sequence reconstruction module collects the continuous starting sample point value sequence of the corresponding channel based on the array channel and frame index information marked in the channel jump starting point set, and performs time base alignment operation in sequence to adjust the out-of-order frame mark, constructs a traceable sliding frame record index sequence, and outputs a radar sliding record structure diagram;

[0045] The window mapping module performs in-window mapping positioning on the serial number range of each receiving frame according to the frame sequence sequence in the radar sliding record structure diagram, combining the window step (1 / 4 of the radar pulse repetition interval, a commonly used analysis window step in radar system design, taking 1 / 4 of the pulse repetition interval (PRI) as the basic unit of time domain analysis), and the periodic index value, judges whether the jump frequency distribution in the mapping window is concentrated in a single frame segment, and if the judgment is concentrated, the corresponding frame segment mapping window is taken as a sliding locking region, and a data sliding mapping path diagram is generated;

[0046] The branch binding module obtains the first frame sample point value and the second frame sample point value in each frame segment in the data sliding mapping path diagram, and calculates a difference amplitude, if there is a continuous frame segment difference amplitude greater than a multi-target resolution threshold (-3dB power point: radar target detection standard parameter, when the power difference of two target echoes is less than 3dB (i.e. the power ratio is less than 2), it is considered to be indistinguishable), the corresponding index is bound as the current concurrent branch input locking frame, and a branch task binding node table is generated;

[0047] The path synchronization module obtains the real-time difference sequence of the first sample point of the current concurrent branch input locking frame and the initial binding frame sample point based on the locking frame index in the branch task binding node table, judges whether the two exceed the set tolerance threshold (5% of the pulse repetition frequency: radar system synchronization tolerance standard, the allowed timing error range is ±5% of the pulse repetition frequency, and if it exceeds this range, it is determined to be out of step), if it is determined to exceed, a path state pause instruction is triggered and the original binding frame segment is traced back, the initial anchor positioning path is re-enabled, and the path state of each period is updated and recorded, and a small phased array radar data real-time processing atlas is generated.

[0048] The channel jump starting point set includes a jump frequency distribution point, a starting frame site position parameter, and a channel-time intersection point, the radar sliding record structure diagram includes a frame serial number index chain, a disordered frame adjustment mark, and a time base alignment label, the data sliding mapping path diagram includes a sliding locking area identifier, a mapping window sequence, and a period step index, the branch task binding node table includes a concurrent branch input locking frame index, a frame segment difference amplitude record, and a multi-target resolution judgment label, and the small phased array radar data real-time processing atlas includes a path state update record, an initial anchor positioning path index, and a real-time difference tolerance judgment label.

[0049] Please refer to Figure 2 , the jump positioning module includes:

[0050] The starting point extraction submodule obtains the starting sample point sequence of each channel receiving frame in the phased array radar antenna array, time-sequentially sorts each group of channel frame head sample points according to the frame time base index, groups the channel sample set according to the array channel number, constructs a channel-time two-dimensional coordinate relationship, maps the sorted starting sample point sequence and the frame time base index according to the position, and generates a channel frame head starting position sequence;

[0051] In the process of acquiring the starting sample point sequence of each channel receiving frame in the phased array radar antenna array, first, the radar antenna array is divided into multiple channels, and the receiving end of each channel collects echo signals to form a frame structure. Assuming that there are 4 channels, each channel contains 1024 sample points per frame, the sample point extraction operation is performed on the 1st frame of each channel, that is, the sampling starting position index value is extracted, for example, the 1st channel frame head sample point is 20, the 2nd channel is 18, the 3rd channel is 22, and the 4th channel is 19. The mapping index relationship between the corresponding channel and the frame head sample point is established respectively, and the sample points are further time-sequenced. Taking the frame time base index as the reference, the frame head sample points of each channel are sequentially mapped to the time axis according to the frame serial number to form an initial time line. Assuming that the frame time base index is 0, 1, 2, …, N, the channel-time two-dimensional coordinate relationship matrix is formed by sequentially mapping the above channel sample points. The sample set is grouped by channel number, for example, the 5 frame sample points under channel 1 are [20, 21, 23, 22, 24], and the sample point curve of the group is displayed to visualize the trend. Further, a data structure is constructed to organize the sample point index corresponding to the channel number and frame index in the form of a two-dimensional list, such as channel 1→[20, 21, 23, 22, 24], channel 2→[18, 19, 19, 21, 22]. Then, the above data structure is standardized to unify the frame head position data format of each channel to an integer array structure. Finally, the channel frame head sample point sequence extraction and mapping operation is completed, and the channel frame head starting position sequence is obtained.

[0052] The difference identification sub-module sequentially performs a first-order difference operation on adjacent frame head sample values based on the channel frame head starting position sequence, compares the difference value sequence with the receiving jump identification threshold point by point, and marks the sample point positions that exceed the receiving jump identification threshold to generate a jump mutation candidate position set.

[0053] In the process of sequentially performing a first-order difference operation on adjacent frame head sample values based on the channel frame head starting position sequence, first, the frame head sample point index array of each channel in the previous result is selected, and a first-order difference operation is performed, that is, a subtraction calculation is performed on adjacent frame index positions, such as the frame head sequence of channel 1 is [20, 21, 23, 22, 24], and the first-order difference is [1, 2, -1, 2]. The difference results corresponding to each channel are recorded as a difference value sequence and stored as a two-dimensional array structure with channel dimension. Next, the receiving jump identification threshold needs to be set, which is calculated based on the standard deviation of background noise and set to 3σ. The background noise data samples are collected from the area without echo signals in the idle frame of the receiving channel, for example, the sample value set is [1.2, 1.4, 1.3, 1.1, 1.5, 1.6, 1.3], and the standard deviation σ is calculated. The calculation process is: where , then According to the setting, the receiving jump identification threshold is set as Each item in the channel difference sequence is compared with the threshold, and the sampling points with an absolute value greater than 0.501 are marked. Assuming that the channel 1 difference is [1, 2, -1, 2], all the sampling points are marked as mutation points, and the point set exceeding the jump threshold in each channel is obtained after screening as the candidate point position, for example, frames 1, 2, 3 and 4 in channel 1 all meet the condition, and finally the jump mutation candidate position set is obtained.

[0054] The jump point confirmation submodule judges whether the three consecutive groups of sampling points all exceed the receiving jump identification threshold according to the jump mutation candidate position set, extracts the starting frame position parameter corresponding to the continuous sampling point section meeting the condition, combines the array channel number and the frame time base index to position the channel-time intersection point, constructs the channel jump starting point set and marks it as the sliding positioning starting point, and generates the channel jump starting point set;

[0055] In the process of judging whether the three consecutive groups of sampling points all exceed the receiving jump identification threshold according to the jump mutation candidate position set, first, the index sequence of the mutation points in each channel is checked by a sliding window, the window length is set to 3, and the mutation point set of the channel is traversed, for example, the mutation points of channel 1 are frames 1, 2, 3, 4 and 5, the three windows of frames 1-3, 2-4 and 3-5 are judged by sliding, the number of mutation points in each window is calculated, and if each point is marked as exceeding the threshold, it is judged as a continuous jump group. The starting frame index meeting the continuous 3-point condition is extracted, for example, the starting position frames 1 and 2 are extracted when the windows frames 1-3 and 2-4 meet the condition, and the corresponding sampling point positions are further obtained, for example, the sampling points of frame 1 are 20 and the sampling points of frame 2 are 21. They are marked as two-dimensional point coordinates in combination with the channel number (such as channel 1) and the frame time base index (such as time base index 0 and 1). Each group of channel-time intersection points meeting the condition is numbered and summarized to construct a sliding positioning point matrix structure. Finally, the frame header starting point set meeting the continuous mutation determination condition in all channels is summarized and processed, and the channel number and the corresponding time axis mutation starting point position meeting the condition are output. A two-dimensional mapping structure is constructed and the position index of each point is recorded to form a sliding positioning mapping table, and the channel jump starting point set is generated.

[0056] Table 1 Channel jump starting point set example table

[0057] Channel number Start frame time base index Start sample point position Whether continuous mutation group 1 0 20 Yes 1 1 21 Yes 2 2 18 No 3 3 22 Yes

[0058] As shown in Table 1, multiple jump starting points are identified in the continuous frames of some channels, indicating that the channels have stable mutation behavior and have the sliding positioning feature.

[0059] Please refer to Figure 3 The frame sequence reconstruction module includes:

[0060] The jump frame acquisition submodule obtains the array channel number and frame index parameters marked in the channel jump start point set, extracts the starting sample point value sequence of the corresponding channel at the continuous frame time base position according to each group of channel numbers, arranges the extracted starting sample values in time sequence to construct a channel frame sampling flow data set, and generates a channel frame sampling value sequence set according to the continuous frame sampling point value structure of the channel grouping;

[0061] In the process of obtaining the array channel number and frame index parameters marked in the channel jump start point set, first, the frame index positions marked as sliding positioning start points of each channel are extracted from the existing channel jump start point set, and are corresponded to the original received data frame sequence. It is assumed that the channel number is 1 to 4, the frame index range is 0 to 15, and the starting frame index of channel 1 is 2, 4, and 6. The sample point position values are extracted from the point positions, such as the starting sample values of channel 1 in frames 2, 4, and 6 are 125, 128, and 126 respectively, and the sample value unit is amplitude unit dB. The sequence array structure of the sample points corresponding to each channel is formed by repeating the operation for each channel. The sequence structure is sorted in ascending order according to the frame time base index to establish the channel frame sampling timeline. For example, the sequence of channel 2 is [124, 125, 126], and the frame index is 3, 4, and 5. The time-sample value mapping table is constructed after sorting. Then, consistency judgment is performed on the sample value sequence of each channel. The short-time interference identification threshold is set to ±3 units. If the sample point value and its adjacent two points differ by more than the threshold, it is determined to be an interference frame and is deleted. For example, the frame value sequence of channel 3 is [126, 145, 127], and the difference between the middle point 145 and the adjacent two points is 19 and 18, which are both greater than the set threshold 3. Therefore, the frame index corresponding frame is marked as an abnormal frame and deleted. Finally, the sample point value sequence after screening is mapped back to the channel-time matrix structure according to the frame index, and a multi-channel continuous sampling frame structure is formed. The time-ordered sample point array of each channel is constructed and the structure is saved to the memory. Finally, a unified time-ordered frame sampling set grouped by channels is established, and a channel frame sampling value sequence set is obtained.

[0062] The frame time base alignment submodule calculates the starting sample point time index difference between the continuous frames of each channel based on the channel frame sampling value sequence set, judges whether the adjacent frame index difference satisfies the time base spacing consistency reference value, constructs a frame out-of-order marking index list, and rearranges the original frame sequence data in time base order according to the frame out-of-order marking index list. After adjustment, the frame index relationship of all channel frame data is reconstructed to obtain a reconstructed time base sequence matrix.

[0063] In the process of calculating the starting sample point time index difference value between consecutive frames of each channel based on the channel frame sample value sequence set, first, the time index corresponding to the sample points of two frames in each channel is subjected to difference operation, and the difference is defined as the time index of the next frame minus the time index of the previous frame. For example, the channel 1 frame time base index sequence is [2, 4, 6], and the difference is 2 and 2. The difference sequence is compared with the set time base spacing consistency reference value. The reference value is set to 2, and the judgment condition is whether the difference is equal to 2. If the two consecutive difference values are equal to the reference value, it is determined that the frame sequence is normal, otherwise it is marked as a disordered frame segment. If the channel 2 frame sequence is [3, 5, 8], the difference is 2 and 3, and the second segment is 3, which exceeds the reference value 2, so the corresponding frame 6 is marked as a disordered frame. All disordered frame indexes are recorded into the disordered marking index table. The index table fields include channel number, frame index, difference, whether disordered and the like. As shown in Table 2, after the construction is completed, the disordered index table is called to perform adjustment operation on the original frame sequence data, that is, the disordered frame is inserted into the adjacent sequence at a suitable position, for example, frame 6 is inserted after frame 5 and renumbered as frame 5.5. After the rearrangement is completed, the timeline of each channel is revalued, and the updated time base index of all channel frame data is established, and finally the reconstructed time base sequence matrix is obtained.

[0064] Table 2: Disordered frame marking index table

[0065] Channel number Original frame index Difference value Whether out of order 1 4 2 No 1 6 2 No 2 5 2 No 2 8 3 Yes

[0066] As shown in Table 2, the frame index 8 of channel 2 is marked as a disordered frame because the spacing with the previous frame is 3, which is greater than the set value 2.

[0067] The sliding record construction submodule constructs a two-dimensional sliding index graph structure according to all channel and frame index intersection points in the reconstructed time base sequence matrix, maps the channel frame sliding window distribution on the time axis combined with the starting point sample value sequence, establishes a sliding frame traceable record path structure combined with the time distribution sequence, summarizes the sliding index path information of all channels, outputs the node connection relationship and channel frame source identification, and generates a radar sliding record structure diagram;

[0068] According to the process of constructing the two-dimensional sliding index graph structure in all channel and frame index intersection points in the reconstructed time base sequence matrix, first, each effective frame index in each channel is mapped as a node in the graph structure, and the node is marked as "channel number-frame index", for example, frame 2, 4 and 6 of channel 1 are mapped as nodes 1-2, 1-4 and 1-6, and the nodes are arranged in time sequence to construct a sliding path with the frame index difference as the edge weight, and the edge weight is defined as the time difference between the current frame and the next frame, assuming that the time index of channel 1 is 2, 4 and 6, then the edge weight is 2 and 2, and the constructed edge set is (1-2→1-4, weight 2), (1-4→1-6, weight 2), and the paths of all channels are combined to construct a unified sliding index graph structure, which is in the form of a multi-path graph, each path represents a time sliding record track of a channel, and then according to the time stamp value of each node, the starting point sampling value sequence is called to construct its time distribution mapping, for example, the node 1-2 of channel 1 is mapped to the sampling value 125, and then the key-value pair (1-2, 125) is generated, and finally a retracable frame path structure is constructed according to the connection order of the nodes, and all channel path sets are integrated into a global path index table, the node connection relationship is recorded in a two-dimensional adjacency matrix, and the structure graph of all node numbers, edge weights and starting point amplitude values is output, and a radar sliding record structure graph is generated.

[0069] Referring to Figure 4 , the window mapping module comprises:

[0070] The frame window positioning submodule extracts the sliding step index value of each channel node on the time axis based on the frame sequence in the radar sliding record structure graph, segments each channel frame index sequence according to the window step, identifies the period segment of each frame through the period index value, constructs a frame number-window period two-dimensional mapping table for each mapping result, and outputs the frame index distribution table in the window;

[0071] Based on the frame sequence in the radar sliding record structure graph, first, the frame index information of each channel in the sliding record structure graph is extracted, and the pulse repetition interval PRI parameter value set by the system is obtained, assuming that PRI is 400 μs, and then the window step is calculated as 100 μs, then according to the window step, the frame index sequence is segmented and mapped according to the time interval, for example, the frame index sequence is [1, 2, 3, 4, 5, 6], and three windows are obtained by grouping every 100 μs: [1, 2], [3, 4], [5, 6], then the time stamp corresponding to each frame is collected and classified combined with the period index value, assuming that the period index is 0~2, then window one corresponds to period 0, window two is 1, and window three is 2, and finally a two-dimensional mapping structure table containing frame index, starting time and period number is constructed, as shown in Table 3.

[0072] Table 3 Window frame mapping table

[0073] Frame index Start time (μs) Period number 1 100 0 2 200 0 3 300 1 4 400 1 5 500 2 6 600 2

[0074] As shown in Table 3, the correspondence between the frame index and its timestamp and period has been explicitly mapped, laying the foundation for subsequent frequency judgment, and ultimately obtaining the frame index distribution table within the window.

[0075] The frequency concentration judgment submodule counts the number of jump frames in each window according to the frame index distribution table within the window, extracts the distribution of jump frequency in each frame index interval, and sets the frame index in a certain window as [1, 2, 3, 4], and the corresponding jump frequency as [3, 0, 0, 0]. The jump is concentrated in frame 1. Set the concentration judgment threshold to 90% proportion. If the jump frequency ratio exceeds the concentration judgment threshold, it is determined to be a concentrated frame segment. The formula is:

[0076] ;

[0077] The frame segment concentration index is calculated to determine whether it exceeds the concentration threshold and make a concentrated frame segment judgment to obtain the concentrated jump frame segment identification sequence; wherein, represents the jump concentration index of the first frame, represents the jump frequency of the first frame in the window, represents the number of frames in the window, represents the mean of the jump frequency, represents the jump frequency fluctuation of the first frame;

[0078] The frame index distribution table within the window is called. First, the jump frequency of each frame segment under each window is counted, and the jump count value of each frame is extracted. For example, the jump frequency of each frame in the window [1, 2, 3, 4] is [3, 0, 0, 0], and the total frequency is 3, which is concentrated in frame 1. Then calculate the frequency mean and standard deviation of all frames in the frame segment. Set the frame frequency array as [3, 0, 0, 0], the mean is 0.75, and the standard deviation is about 1.5. Calculate the frequency deviation of the concentrated frame from the mean of all frames, and then combine the fluctuation value before and after the frame to construct the innovation concentration index. The frame 1 concentration is calculated by the formula:

[0079] ;

[0080] ;

[0081] ;

[0082] ;

[0083] ;

[0084] Since the concentration judgment threshold is 0.25, 0.326>0.25, it is determined that frame 1 is a concentrated frame, and finally all window frame segments meeting the condition are screened out to obtain the concentrated frame segment identification sequence. The results show that the jump concentration index effectively reveals the concentration of the frame jump event in the time distribution dimension.

[0085] The operation logic of the formula is to comprehensively measure the concentration characteristics of a frame in the jump frequency, and at the same time to suppress the misjudgment caused by local mutation, so a variety of mathematical operations are adopted to enhance the discrimination ability of the index. The numerator part represents the jump frequency of the first frame, and the deviation between the jump frequency of the frame and the average jump frequency of all frames in the window is adopted. The absolute value operation can prevent the result from being offset due to the negative deviation, so that the deviation of concentration and dispersion has a unified dimension in the magnitude; the denominator part is the standard deviation form of the frequency distribution in the window, which is used to reflect the dispersion degree of the frequency in the whole group of frames, and the square deviation is restored to the original dimension by the square root operation, which enhances the comparability of the index; the additional term is the local fluctuation of the first frame, which is defined as the absolute value sum of the difference between the adjacent frame jump frequencies, and is used to describe the stability of the frame in its neighborhood. If the value is too large, the overall denominator value will be increased, thereby reducing the concentration index of the frame, avoiding misjudgment of individual mutation as a concentrated frame; the whole formula is calculated by the ratio of the numerator and the denominator, and the concentration degree is output in a relative standardized way, which not only highlights the frequency anomaly of the current frame, but also considers the global distribution and local fluctuation factors, realizing the judgment of the jump concentrated frame segment under multi-dimensional constraints.

[0086] The jump concentration index is used to measure the concentration degree of a single frame in the jump event distribution in its window, and reflects the jump dominance of the frame in the time domain. The index not only measures whether the jump frequency of the frame is significantly higher than the average level of other frames in the same window, but also introduces the frequency fluctuation of the whole group of frames and the local change amplitude of the frame as a constraint, so that the index result can comprehensively reflect the statistical deviation between frames and the local jump stability. The larger the index value is, the more outstanding the jump frequency of the frame is in the whole window, and the less active the jump behavior of its adjacent frames is, so it can be inferred that the frame is a concentrated carrier point of the jump event, which is suitable for identifying the key frame segment with the characteristics of sudden spectrum transition in the radar echo sequence, and is an important criterion for locking the core mutation area in the sliding mapping path.

[0087] The locking path construction submodule filters out the corresponding frame index interval according to the concentrated jump frame segment identification sequence, extracts the time path node sequence in the radar sliding record structure diagram, reads the concentrated frame segment path mapping information, establishes the mapping path set of the jump frame segment and the window period, and generates the data sliding mapping path diagram;

[0088] According to the centralized hopping frame segment identification sequence screening corresponding frame index, each frame segment in the radar sliding record structure diagram is located item by item, first, the frame index value marked in the centralized hopping frame segment sequence is extracted, such as in the previous stage, frame 1 is determined as a centralized frame segment, its channel number and frame index combination form node is obtained, for example, the node form is 1-1 (channel 1, 1st frame), if frame 3, frame 5 are also marked as centralized frame segments, nodes 1-3, 1-5 are added, at the same time, the upstream and downstream node path information connected by the above nodes in the sliding record structure diagram is read, the path connection relationship is structurally analyzed, the sub-path where the centralized frame segment is located in the path graph is extracted, and a window period path structure table is constructed in combination with the period index value to which it belongs, assuming that the periods to which frame 1, frame 3 and frame 5 belong are period 0, 1 and 2 respectively, the following mapping relationship structure is established.

[0089] Table 4 Period path structure table

[0090] Period number Centralized frame index Node number Forward connection Backward connection 0 1 1-1 0-6 1-3 1 3 1-3 1-1 1-5 2 5 1-5 1-3 1-7

[0091] As shown in Table 4, centralized frame segment nodes 1-1, 1-3 and 1-5 all have directed connection relationship with the front and rear frame segments in the sliding structure diagram, the rear connection node 1-7 indicates that the sliding path will be extended to the subsequent frame segment, at the same time, through the period number field in the structure, the time belonging of each frame segment can be classified at the period level, the channel number and index mapping value of each group of centralized node sequence is called to construct a channel-frame-period three-dimensional structure coordinate system, all nodes that meet the centralized judgment will constitute a significant hopping segment in the sliding path graph, a normalization numbering operation is performed on the node set of these paths, a unique path identification ID is assigned, and its connection relationship matrix is recorded, for example, the path ID is P1, the connection node sequence is [1-1, 1-3, 1-5], the connection direction is forward, the traceable trajectory of the path in the sliding record structure diagram is established, further, all path IDs and their corresponding frame segment node sequences are fused to form a sliding graph structure dictionary, the dictionary key is the path ID, and the value is the path sequence, finally, the mapping locking and trajectory expression of all centralized hopping frame segments in the radar structure diagram are completed, and the data sliding mapping path graph is established.

[0092] Please refer to Figure 5 , the branch binding module comprises:

[0093] The hopping amplitude extraction submodule obtains the first frame and second frame sample point values of each frame segment in the data sliding mapping path graph, extracts the starting sample amplitude pairs between the continuous frame segments of the same channel, calculates the amplitude difference value and uniformly converts it into dB value, establishes a channel-frame segment amplitude difference value table, and generates a frame segment hopping amplitude sequence;

[0094] In the process of obtaining the first frame sample point value and the second frame sample point value of each frame segment in the sliding mapping path diagram, first, each effective path segment in the sliding path diagram structure is extracted, and the starting sample point amplitude value pair of the two consecutive frames in each segment is obtained. Assuming that the starting sample point amplitude values of channel A in the 5th frame and the 6th frame in path P1 are 120 and 114 units respectively, the extracted values are converted into dB form, and the difference value between the two is calculated by the conversion formula The linear amplitude value is converted into dB value, 120 is converted into 20.79 dB, 114 is converted into 20.56 dB, and the difference value between the two is 0.23 dB. The same operation is continuously performed on all path segments to form a complete difference value sequence between frame segments. The difference value is calculated by subtracting the front frame from the rear frame, the positive and negative values are recorded to retain the directional information, and the difference value is recorded as a three-tuple structure of “frame segment number-channel number-difference value” and stored in the difference value sequence table. If there are multiple channel crossing frame segments in the path, the difference value of each frame segment needs to be recorded according to the channel to avoid data aliasing of different channels. In order to facilitate the calling of the data structure, the difference value sequence is aggregated according to the path number to form a path difference value set, and finally a frame segment jump amplitude sequence is generated.

[0095] The target distinguishable judgment submodule judges whether each group of difference values is greater than the multi-target distinguishable threshold according to the frame segment jump amplitude sequence. Assuming that the threshold is set to 3 dB, the frame segments with a difference value greater than the multi-target distinguishable threshold are marked as distinguishable effective frame segments, and the index positions of all frame segments satisfying the condition are counted to obtain a distinguishable effective frame segment index set.

[0096] In the process of judging whether each group of difference values is greater than the multi-target distinguishable threshold according to the frame segment jump amplitude sequence, the threshold is set to 3 dB. According to the general standard of radar target detection, if the difference between two echo powers is less than 3 dB, it is considered to be indistinguishable. The judgment operation is performed on each pair of sample difference values in a traversal manner. If the amplitude difference value of a frame segment is 4.2 dB, the judgment result is distinguishable. If the difference value is 2.1 dB, the judgment result is indistinguishable. The judgment result is stored in the judgment vector in the form of a Boolean value. Then, the frame segment indexes marked with “True” are extracted to construct a frame segment index and difference value correspondence table, as shown in Table 5.

[0097] Table 5 distinguishable judgment result table

[0098] Frame segment number Channel number Difference value (dB) Whether distinguishable 1 A 4.2 Yes 2 A 2.1 No 3 B 3.7 Yes

[0099] As shown in Table 5, frame segments 1 and 3 are judged to be distinguishable segments, and frame segment 2 is excluded because it is below the threshold. The path ID, frame index, and channel information of all frame segments judged to be “Yes” are recorded, and the distinguishable effective frame segment index set is obtained.

[0100] The task node binding submodule constructs a task concurrent channel mapping and synchronously marks the current frame segment as a locked task input frame based on the resolved effective frame segment index set, taking each frame segment start index meeting the condition as an input node index number and associating the sliding path ID to which the current frame segment belongs, and establishes a branch task binding node table;

[0101] According to the resolved effective frame segment index set, the start frame index of each item is locked and mapped as a task input node. First, the path number and start frame index combination key of each effective frame segment is extracted, for example, the effective frame segment number in path P1 is 3, and the corresponding node is marked as P1-3. The node is then bound to the current task execution unit number and recorded as a branch task input lock point. Subsequently, the node structure is clustered and arranged by calling the channel number, forming a ternary relation table of "channel-task ID-frame segment index". To prevent conflicts caused by repeated node binding, a repeated index detection logic is set. If the same frame segment is shared by multiple paths in a channel, the first binding node is retained in ascending order of path number, and the remaining nodes are discarded. A unique binding index list is established. Finally, the binding results of all channels are executed for structure merging, and the output is a standardized task mapping structure table, which covers path number, input frame index, binding channel ID, task lock state identifier, and other content. A data structure that can be directly scheduled and called is constructed, and a branch task binding node table is established.

[0102] Please refer to Figure 6 , the path synchronization module includes:

[0103] The sampling difference monitoring submodule obtains the locked frame index in the branch task binding node table, extracts the first sampling point value of the current concurrent branch input locked frame and the first sampling point value of the corresponding initial binding frame, calculates the amplitude difference between the two groups of sampling point values in time axis order, forms the corresponding mapping of the frame pair sequence and the amplitude difference sequence, and obtains the locked frame difference amplitude sequence.

[0104] After obtaining the lock frame index in the branch task binding node table, the first sampling point value of the current lock frame and its corresponding initial binding frame under each concurrent path is extracted in turn. Assuming that the initial binding frame of path P1 in channel A is the 10th frame and the corresponding sampling point value is 124 units, the current lock frame is the 13th frame and the corresponding value is 127 units, the difference between the two is 3 units, and the difference is converted into the synchronization difference on the time axis, considering that the radar sampling rate is 5 MHz, 1 unit corresponds to 0.2 μs, and the difference time is 0.6 μs. After constructing the difference set of all frame pairs, the difference value is normalized according to the path number, and if there are multiple lock-binding frame combinations in a path, the difference value vector is generated in order. For example, in path P2, the lock frame index is [12, 13], the binding frame is [9, 10], the difference is 1.2 μs and 0.8 μs respectively, a path time difference mapping table is formed, input parameter structures are prepared for subsequent step loss tolerance judgment, and a lock frame difference amplitude sequence is finally obtained.

[0105] The tolerance step loss judgment submodule judges whether there is a frame pair exceeding the set tolerance threshold according to the lock frame difference amplitude sequence. Assuming that the pulse repetition frequency is 1 kHz and the corresponding tolerance threshold is 50 μs, the frame segment index greater than the tolerance threshold in the difference value sequence is extracted, and the state is marked as step loss. The path number and backtracking target of each group of step loss frame segments are recorded to generate a step loss path identification list.

[0106] According to the lock frame difference amplitude sequence to determine whether to lose step, the tolerance threshold setting method needs to be first clarified. In the embodiment, the pulse repetition frequency of the radar system is 1 kHz, that is, the radar transmits a pulse every 1 ms. This value is an important parameter for measuring the time domain synchronization control accuracy of the radar. Therefore, the tolerance threshold is set to 5% of this value, that is, ± 50 μs. In the judgment process, the difference value sequence of each path is traversed. If the absolute value of the difference value of a frame segment is greater than 50 μs, it can be judged that the path frame segment is a step loss frame segment. For example, in the aforementioned path P2, the difference values of two frame segments are 1.2 μs and 0.8 μs, both of which are less than 50 μs, so they are marked as a synchronization state. If the difference value of path P4 is 52 μs, it exceeds the threshold range and needs to be marked as a step loss state. All judgment actions need to use the absolute value function for operation, that is, the absolute value of each item of the difference value vector is converted and then compared with the threshold value 50 μs. At the same time, the path state identification code is generated for the judgment result, using binary identification, in which 1 represents step loss and 0 represents synchronization. The result structure body is generated for the judgment result, including the path number, the lock frame index, the binding frame index, the difference value, the judgment result, and the identification code field, forming a structured record table. For example, the structure body field corresponding to path P4 is: path number = 4, lock frame = 14, binding frame = 11, difference value = 52 μs, state = step loss, and identification = 1. All structure bodies are aggregated according to the path number to generate a complete step loss identification table of the system in the current period. In the table, it can be directly found whether the current frame segment of each path is out of step, which is convenient for subsequent path reconstruction operation and system state table update processing. Finally, the step loss path identification list is obtained.

[0107] Table 6: Small phased array radar path state determination example table

[0108] Path number Lock frame index Binding frame index Difference value (μs) State identification State code Judgment time (ms) Judgment reason P1 13 10 0.6 Synchronization 2 5.2 Difference value within threshold limit P2 13 10 0.8 Synchronization 2 5.2 Difference value within threshold limit P3 15 11 1.0 Synchronization 2 5.4 Difference value within threshold limit P4 14 11 52.0 Out of step 1 5.8 Difference value beyond limit

[0109] As shown in Table 6, the difference value of path P4 is 52 μs, which exceeds the tolerance threshold set by the pulse repetition frequency, triggering the step loss mark and entering the state update phase. The path will be updated to restart in the state code table in the subsequent.

[0110] The state backtracking update submodule sets the related path state to pause based on the step loss path identification list, and traces back to the original binding frame segment. The target path is reset to the initial anchor position path. The path state information record table in the update period is updated according to the time index, and the state field change flag is marked. The radar data processing graph structure table is established.

[0111] According to the path number, the lock frame index, and the binding frame index field contained in the step-out path identification list, the path state update operation is performed on the state management table in sequence, first, for each step-out path, the state mark field is adjusted from running (state code 2) to the pause state (state code 0), then the starting node of the path is repositioned according to the binding frame index field, the path backtracking operation is realized, for example, the binding frame index of path P4 is 11, then the frame 11 is re-enabled as the new starting point of the path, the path trajectory structure is restored, at the same time, the path restart flag is set to valid, and the path state code is set to 1 to indicate the restart state, after that, the centralized update operation of the state table is performed on all paths in the cycle, the current state field of each path is extracted and is given a unique state code: running is 2, pause is 0, and restart is 1, the path state mapping table is constructed, and is sorted according to the path number, forming a structured state array, which is used as the key frame segment state input reference in the subsequent radar map construction, in addition, the path state change timestamp and the change reason field need to be recorded, for example, the path P4 state change time is 5.8 ms, and the reason field is marked as “threshold difference step-out”, finally, the path state code, the lock frame index, the binding frame index, the state code value, the change time, and the change reason field are merged to construct the radar data processing map table, which is used as the dynamic reflection of the path state evolution trend, and the radar data processing map structure table is generated.

[0112] The above is only a preferred embodiment of the present application, and does not limit the present application in other forms. Any skilled person in the art can modify or change the above disclosed technical content to equivalent embodiments applied to other fields, but any simple modification, equivalent change, and modification made according to the technical essence of the present application to the above embodiments without departing from the technical solution content of the present application still belongs to the protection scope of the present application.

Claims

1. A small phased array radar data real-time processing system, characterized by, The system comprises: The jump positioning module obtains the starting sampling point sequence of each channel receiving frame in the phased array radar antenna array, sequentially performs difference operation, extracts the corresponding starting frame site position parameter when three consecutive groups exceed the receiving jump identification threshold, confirms the corresponding channel-time intersection point, marks the sliding positioning starting point, and obtains the channel jump starting point set; The frame sequence reconstruction module collects the continuous starting sampling point value sequence of the corresponding channel based on the array channel and frame index information in the channel jump starting point set, constructs the backtracking sliding frame record index sequence, and outputs the radar sliding record structure diagram; The window mapping module performs in-window mapping positioning on each receiving frame number range according to the frame sequence sequence in the radar sliding record structure diagram, judges the sliding locking area, and generates the data sliding mapping path diagram; The branch binding module obtains the first frame sampling point value and the second frame sampling point value in the data sliding mapping path diagram and calculates the difference amplitude, and if the continuous frame segment is greater than the multi-target resolution threshold, the corresponding index is bound as the current concurrent branch input locking frame, and the branch task binding node table is generated; The path synchronization module obtains the real-time difference sequence of the concurrent branch input locking frame based on the branch task binding node table, and if it exceeds the set tolerance threshold, triggers the path state pause instruction, re-enables the initial anchor positioning path, and generates the small phased array radar data real-time processing diagram.

2. The small phased array radar data real-time processing system of claim 1, wherein, The channel jump starting point set comprises jump frequency distribution points, starting frame site position parameters, and channel-time intersection points, the radar sliding record structure diagram comprises frame number index chain, out-of-order frame adjustment markers, and time base alignment labels, the data sliding mapping path diagram comprises sliding locking area identifiers, mapping window sequences, and period step indexes, the branch task binding node table comprises concurrent branch input locking frame indexes, frame segment difference amplitude records, and multi-target resolution judgment labels, and the small phased array radar data real-time processing diagram comprises path state update records, initial anchor positioning path indexes, and real-time difference tolerance judgment labels.

3. The small phased array radar data real-time processing system of claim 1, wherein, The jump positioning module comprises: The starting point extraction submodule obtains the starting sampling point sequence of each channel receiving frame in the phased array radar antenna array, time-sequentially sorts each channel frame header sampling point according to the frame time base index, groups the channel sample set according to the array channel number, constructs the channel-time two-dimensional coordinate relationship, maps the position according to the sorted starting sampling point sequence and the frame time base index, and generates the channel frame header starting position sequence; The difference identification submodule sequentially performs first-order difference operation on adjacent frame header sampling values based on the channel frame header starting position sequence, compares the difference value sequence with the receiving jump identification threshold point by point, marks the sampling point position exceeding the receiving jump identification threshold, and generates the jump mutation candidate position set; The jump point confirmation submodule judges whether the three groups of continuous sampling points exceed the receiving jump identification threshold value according to the jump mutation candidate position set, extracts the starting frame position parameter corresponding to the continuous sampling point section meeting the condition, combines the array channel number and the frame time base index to position the channel-time cross point, constructs the channel jump starting point set and marks it as the sliding positioning starting point, and generates the channel jump starting point set.

4. The small phased array radar data real-time processing system of claim 1, wherein, The frame sequence reconstruction module comprises: The jump frame acquisition submodule obtains the array channel number and frame index parameter marked in the channel jump starting point set, extracts the starting sampling point value sequence of the corresponding channel at the continuous multi-frame time base position according to each group of channel numbers, arranges the extracted starting sampling value in time sequence to construct the channel frame sampling flow data set, generates the channel frame sampling value sequence set in the structure of the continuous frame sampling point value grouped by the channel; The frame time base alignment submodule calculates the starting sampling point time index difference value between the continuous frames of each channel based on the channel frame sampling value sequence set, judges whether the adjacent frame index difference value meets the time base distance consistency reference value, constructs the frame disorder marking index list, and rearranges the original frame sequence data in the time base sequence according to the frame disorder marking index list, adjusts the frame index relationship of all channel frame data after the rearrangement, and obtains the reconstructed time base sequence matrix; The sliding record construction submodule constructs a two-dimensional sliding index graph structure according to all the channel and frame index cross points in the reconstructed time base sequence matrix, maps the channel frame sliding window distribution on the time axis in combination with the starting point sampling value sequence, establishes a sliding frame traceable record path structure in combination with the time distribution sequence, collects the sliding index path information of all channels, outputs the inter-node connection relationship and the channel frame source identification, and generates a radar sliding record structure diagram.

5. The small phased array radar data real-time processing system of claim 1, wherein, The window mapping module comprises: The frame window positioning submodule extracts the sliding step index value of each channel node on the time axis based on the frame sequence in the radar sliding record structure diagram, segments each channel frame index sequence according to the window step length, identifies the period section of each frame through the period index value, constructs a frame number-window period two-dimensional mapping table for each group of mapping results, and outputs the frame index distribution table in the window; The frequency set judgment submodule counts the number of jump frames in each window according to the frame index distribution table in the window, extracts the distribution of the jump frequency in each frame index interval, calculates the frame section concentration index, judges whether it exceeds the concentration threshold value, and performs the concentrated frame section judgment to obtain the concentrated jump frame section identification sequence; The locking path construction submodule filters out the corresponding frame index interval according to the concentrated jump frame section identification sequence, extracts the time path node sequence in the radar sliding record structure diagram, reads the concentrated frame section path mapping information, establishes the mapping path set of the jump frame section and the window period, and generates a data sliding mapping path diagram.

6. The small phased array radar data real-time processing system of claim 1, wherein, The branch binding module comprises: The jump amplitude extraction submodule extracts the sampling point values of the first frame and the second frame of each frame section in the data sliding mapping path diagram, extracts the starting sampling amplitude pairs between the continuous frame sections of the same channel, calculates the amplitude difference value and uniformly converts it into the dB value, establishes the channel-frame section amplitude difference table, and generates the frame section jump amplitude sequence; The target distinguishable judgment submodule judges whether each group of difference values is greater than a multi-target distinguishable threshold according to the frame segment jump amplitude sequence, marks a frame segment as a distinguishable effective frame segment if the difference value is greater than the multi-target distinguishable threshold, and obtains a distinguishable effective frame segment index set by counting all index positions satisfying the condition; The task node binding submodule takes each frame segment starting index satisfying the condition as an input node index number based on the distinguishable effective frame segment index set, associates a current frame segment belonging to a sliding path ID, constructs a task concurrent channel mapping, and synchronously marks a locked task input frame, and establishes a branch task binding node table.

7. The small phased array radar data real-time processing system of claim 1, wherein, The path synchronization module comprises: The sampling difference monitoring submodule obtains a locked frame index in the branch task binding node table, extracts a first sampling point value of a current concurrent branch input locked frame and a first sampling point value of a corresponding initial binding frame, calculates an amplitude difference between the two groups of sampling point values in a time axis order, forms a corresponding mapping of a frame pair sequence and an amplitude difference sequence, and obtains a locked frame difference amplitude sequence; The tolerance step loss judgment submodule judges whether there is a frame pair exceeding a set tolerance threshold limit according to the locked frame difference amplitude sequence, extracts a frame segment index greater than the tolerance threshold limit in the difference value sequence, marks a state as step loss, records a path number and a backtracking target for each group of step loss frame segments, and generates a step loss path identification list; The state backtracking updating submodule sets a related path state as pause based on the step loss path identification list, backtracks to an original binding frame segment, redefines a target path as an initial anchor position path, updates a path state information record table in a time index update period, marks a state field change flag, and establishes a radar data processing graph structure table.

Images