CN104808197A - Multi-surveillance-source flying target parallel track processing method - Google Patents

Abstract

The invention discloses a multi-surveillance-source flying target parallel track processing method. The method includes the steps: multi-surveillance-source data receiving; multi-surveillance-source data analysis; radar data processing; ADS-B (automatic dependent surveillance-broadcast) data processing; multi-surveillance-source data fusion. Surveillance of quality of data accessing to radar is realized by monitoring and analyzing quality of radar signals. In addition, real-time receive processing of the radar data is realized by means of multithreading, high safety, high reliability and high usability of a data processing system can be further guaranteed, and accuracy and quickness in track processing of flying targets in different data types from different surveillance sources can be realized.

Description

A Parallel Tracking and Processing Method for Flying Targets with Multiple Surveillance Sources

technical field

The invention relates to a multi-monitoring source flying target parallel tracking processing method.

Background technique

At present, air traffic control basically relies on radar surveillance and very high frequency (VHF) communication, both of which are limited by line-of-sight propagation, and the coverage range is relatively small. Due to the limitations of various factors, radar and VHF coverage cannot be achieved in the vast ocean airspace and remote mountains, deserts or jungle areas, resulting in blind spots in the flight airspace and bringing hidden dangers to flight safety. In order to overcome these shortcomings and achieve effective air traffic surveillance, the International Civil Aviation Organization (ICAO: International Civil Aviation Organization): proposed a new concept of surveillance, based on the application of satellite technology in the future air traffic surveillance, which is automatic correlation Monitoring (ADS: Automatic Dependent Surveillance). The 1992 General Assembly of ICAO formally passed the new air navigation system plan based on satellite navigation, satellite communication and data link communication, thus starting the transition of air traffic control from the existing ground-based system to the new navigation system. The new navigation system consists of four parts: communication, navigation, surveillance, and air traffic management. Surveillance is automatic dependent surveillance. It measures the four-dimensional position data of the aircraft by the navigation and positioning system on the aircraft, and automatically sends them to the aircraft through the ground-air communication data link. The ground air traffic control center conducts air traffic management and flow management. Therefore, after the introduction of the new navigation system, various means of monitoring aircraft have appeared, such as radar, ADS-B, etc. If the monitoring data from different monitoring sources, different times, and different coordinate systems can be integrated, multiple Coverage, expanding the scope of monitoring and control, improving the accuracy of target positioning, and increasing the credibility of predictions, so as to ensure the safety of flight and the efficiency of space area utilization.

In short, data fusion technology is: comprehensive processing of information from multiple sensors or sources to obtain more accurate and reliable conclusions. A stricter definition is: the use of computer technology to automatically analyze and synthesize the observation information of several sensors acquired in time series under certain criteria to complete the required decision-making and estimation tasks.

The shortcoming that prior art exists is as follows:

1. The quality of each monitoring source signal connected to the multi-monitoring source data processing system is directly related to the quality of the fusion signal. The existing ATC system has certain limitations in monitoring and analyzing the quality of the monitoring source signal.

2. For the radar data processing module, the parameters are mainly the type and rate of each radar signal format, the geographic coordinates and declination angle of the radar antenna position, the antenna rotation speed, the track start value and track end value of track quality management, and related The threshold parameters of multi-radar data fusion such as distance, relative altitude, relative speed, and relative heading, etc., and the static and dynamic weighting values of all radars participating in the fusion during multi-radar fusion, etc. The existing system adopts the method of setting the parameters in the program to configure the radar parameters. It is impossible to modify the parameters such as the radar data fusion threshold, the static weighting value and the dynamic weighting value of the radars participating in the fusion. Any change of these parameters requires the programmer Technical support, lack of flexibility, poor operability.

3. In the process of radar data tracking and processing, data receiving, decoding, and processing have high requirements for real-time performance. In the radar processing system, single-threaded programming technology is adopted, and radar data receiving and tracking processing are carried out cyclically, which may easily cause data loss. Radar Whether the terminal can receive radar data in real time and accurately is particularly important.

Contents of the invention

In order to overcome the shortcoming of prior art, the present invention provides a kind of parallel tracking processing method of multi-monitoring source flight target, by monitoring and analyzing the quality of radar signal, monitor the data quality of access radar; Real-time receiving and processing further ensures the high security, high reliability and high availability of the data processing system, and can accurately and quickly track and process flight targets of different data types from different monitoring sources.

The technical scheme adopted in the present invention is: a kind of multi-monitoring source flight target parallel tracking processing method, comprises the following steps:

Step 1. Multi-monitoring source data reception;

Step 2, Multi-monitoring source data analysis: Discriminate the message type information on the original data received in step 1, then call the data parser to analyze the data, and normalize the parsed message data, and then stored in an internal data table;

Step 3. Radar data processing:

(1) Real-time quality control;

(2) Radar data preprocessing;

(3) Radar data processing: including single radar tracking processing and multi-radar fusion processing;

Step 4, ADS-B data processing: generate single ADS-B track and ADS-B fusion track, and store the track in the ADS-B track table;

Step five, multi-monitoring source data fusion:

(1) System track association;

(2) System track tracking;

(3) Fusion of ADS-B track and radar track.

Compared with prior art, positive effect of the present invention is:

1. In the process of radar data processing, a multi-threaded control structure is adopted, and the combination of concurrent and serial methods is used to realize the processing of different data types of each radar, which effectively controls the number of radar data processing threads in the system and reduces the burden on the system. The demand for memory and CPU during operation also meets the real-time requirements of radar data processing.

2. The present invention introduces radar signal quality monitoring, which will monitor the frame loss of radar data in real time, and make it easy to check the track (track) data of each radar signal. By setting the system fault tolerance parameters, the system can automatically select the radar whose signal quality meets the requirements for subsequent fusion processing.

3. The present invention uses a visual interface to configure radar parameters. The radar parameter value and the data fusion threshold parameter value, mostly the static weighted value and dynamic weighted value of the radars participating in the fusion when the radar is fused, are configured with a visual interface, and the system users can modify the corresponding parameters according to the configuration requirements on the seat. Parameter value, configuration is simple, flexible, convenient, and highly operable.

4. The monitoring source of the monitoring information provided by the present invention is composed of multiple monitoring sources such as primary radar, secondary radar, and ADS-B, which is a typical multi-sensor function synthesis scheme. The combination of ADS-B data and multi-radar can achieve multiple coverage, expand the scope of monitoring and control, improve the accuracy of target positioning, and increase the reliability of prediction, so as to ensure the safety of flight and the efficiency of space area utilization.

5. The data fusion of multi-radar and ADS-B adopts a distributed fusion structure, which has good tracking performance, low requirements for communication broadband, fast calculation speed and good reliability.

Description of drawings

The invention will be illustrated by way of example with reference to the accompanying drawings, in which:

Fig. 1 is the flow chart of parallel tracking processing of multi-monitoring source flying targets of the present invention;

Fig. 2 is the flow chart of multi-monitoring source data receiving of the present invention;

Fig. 3 is a schematic diagram of an earth-centered fixed coordinate system;

Fig. 4 is the flowchart of single radar data processing of the present invention;

Fig. 5 is the flowchart of multi-radar fusion data processing of the present invention;

Fig. 6 is a flow chart of radar track correlation;

Fig. 7 is a schematic diagram of multi-radar and ADS-B track distributed fusion model;

Detailed ways

The multi-monitoring source flight target parallel tracking processing system designed by the present invention includes five parts: data receiving, data preprocessing, radar data processing, ADS-B data processing and fusion data processing, and its flow is shown in Figure 1.

In the parallel tracking process of flying targets with multiple monitoring sources, the data of each radar head and ADS-B data must be obtained first and stored in the database. Then these data need to be verified, classified and some erroneous data should be eliminated. The target track obtained in one scanning cycle of the single radar is tracked and processed by the radar processor, and the track data of the single radar is generated, and the state estimation of the target is given. At the same time, radar local tracks are generated through multi-radar data processing. According to the same processing flow, ADS-B point track is processed to generate single ADS-B track and ADS-B local track. The radar local track and ADS-B local track are sent to the fusion processing module for system track correlation and fusion, and finally realize the tracking and monitoring of large-scale flying targets.

The specific implementation steps are as follows:

Step 1. Multi-monitoring source data reception (reception of original data)

The underlying interface of the original data receiving module is a network interface, and through the interface provided by the development kit, the original data is received, and then the received original data is submitted to the preprocessing module, which analyzes the original data according to different protocols, and CRC checks , Calculate, process, and store in the database after processing.

The upper layer of data receiving is the database, here is the Oracle database. The original information and configuration information of the equipment are configured through the visual interface and stored in the database. After the data receiving module reads the configuration from the database, it starts data classification, and then performs Data reception.

Considering that there are many monitoring sources of flight targets, in order to ensure the real-time performance of software processing data, the software will use multi-threading technology to ensure that data can be received quickly, prevent buffer overflow and data loss, and meet the requirements of real-time processing. After the software starts, it will load the protocol according to the radar parameter settings in the database, and initialize the service, including the receiving port and protocol type, etc. Start the thread, which is responsible for listening to the specified port.

As shown in Figure 2, the flow of multi-monitoring source data reception is as follows:

1. Establish a database connection, read the data source information from the data source parameter configuration table, if the read data gives an abnormal value in the log record, and exit the program.

2. Create corresponding data receivers according to the IP address, port, data type, and data subtype of the data source, and each data receiver is a thread.

3. Start the thread and receive the raw data of the corresponding port.

4. The receiver thread stores the received original data into the original database and waits for subsequent processing.

Step 2. Multi-monitoring source data analysis

Data analysis Discriminates the message type information of the original data, calls the ADS-B data parser or SSR data parser to analyze the data, and converts the parsed message data into a standardized monitoring data format and stores it in the in the internal datasheet.

The data analysis module is realized by multi-thread technology, and the analysis process is divided into three steps: data reading, data processing, and data saving.

1. Start the data reading thread to read the original data from the original database.

2. Start the corresponding type of processing thread according to the data type of the original data, and normalize the original data. Normalization processing can be divided into data preprocessing, classifiers, data parsers, and format converters.

The classifier judges according to the type information of the received message, and invokes the corresponding parser to parse the data. In the process of data decoding, a version processing strategy is adopted: when a message is received, according to the SAC and SIC in the message, look up whether the corresponding SAC|SIC in the version cache table has a version record, and if so, use this version to parse the message text; if there is no version or if there is a version but the parsing is unsuccessful, try other versions from high to low for parsing. Returns a parsing error on success. When using a certain version for data decoding, complete the following steps:

The first verification of the message: verify whether the message length is at least 3 bytes; whether the message type (the first byte of the message) matches the parser; whether the message length is consistent with the value of the message length field, etc. If the verification is passed, go to the next step, otherwise stop parsing and return a parsing error;

Analyze the SAC and SIC information, and find the version number used when the last analysis was successful according to the value of the message SAC|SIC;

Try parsing: If the version that was successfully parsed last time is found, parse the message according to this version, if not found, try to parse with each version in turn, if a version parses successfully, record the message of the SAC|SIC station The version of is used for the next parsing; if none succeeds, a parsing error will be returned.

Each data item of the successfully parsed message is stored in the internal variable of the parser, and the default value is set for the data item not provided in the message. When the parse is successful, an SQL statement can be generated as needed for updating to the database.

Data preprocessing is to perform abnormal data processing, data quality monitoring and statistics in the process of data identification and normalization. When the slant distance difference of multiple consecutive points of recorded data is zero, it is regarded as tracking loss; for the jump point. For the statistics of abnormal values, a counter is added to the program to record the number of abnormal data. When the ratio of abnormal data to total data is greater than the fault tolerance value of the system, the data is considered invalid. The fault tolerance value of the system can be set in the software of radar signal quality monitoring, analysis and recording. Raw data preprocessing directly eliminates large pieces of missing data. Apply the corresponding algorithm to the jump point data and restore it.

Data processing According to the type of data read from the database, when a message is received, the main thread starts a secondary thread to classify the original data and convert the data format, and the main thread continues to monitor. After the secondary thread receives the original data from the main thread, it checks and initially analyzes the data according to the protocol to obtain the data type of the original data, including monitoring data, aftn message data (flight plan, flight information, weather information), flight flow Surveillance, airspace operations, air traffic control infrastructure. After the secondary thread finishes processing the data, it is stored in the database according to the agreement, and the life of the secondary thread ends after confirmation. If an error occurs during the period, enter the fault-tolerant processing module. After receiving the termination command from the user, the software terminates the main thread after processing the data.

3. Start the data storage thread, and store the parsed data in the corresponding data preprocessing table for subsequent use.

Step 3. Radar data processing

Considering that there are many monitoring sources of flying targets, in order to ensure the real-time performance of software processing data, the design of the software adopts a two-dimensional multi-thread control structure, so the system can be easily expanded and can quickly process multiple radar sites.

Since the speed of radar data processing depends on the reading of radar data, it is necessary to monitor whether the transmission of radar data has been completed in advance. Radar data arrives every 4s, so it must be within 4s, that is, before the arrival of the next radar data. The radar data processing of the station is completed. The data processing of each radar station is implemented by threads, and the program controls the processing of different data types of each radar through a multi-threaded control structure. Since the threads of different data types of each radar read data serially in their own way, data access conflicts between radars and different data types of radars are avoided, and there is no need to use other mechanisms to control data access .

The present invention designs a multi-threaded control structure, which can be regarded as a two-dimensional array of m rows and n columns, wherein the rows define M radar data types to be processed, and the columns define N radar sites to be monitored . Here Control[M][N] is used to represent the two-digit array, and Control[i][j] represents whether the i-th data type processing process of the j-th radar site can be started. Control[i][j]=true, then the processing of data type i of radar site j can be started currently. If Control[i][j]=false, it indicates that the i-th data type of the radar site has been processed or is waiting to be processed. Through the multi-threaded control structure, it is easy to implement and process the radar data of various radar sites and different data types. According to the actual operation situation, instead of processing different data types of all radar sites concurrently, the above-mentioned combination of concurrent and serial methods is used to effectively control the number of system radar data processing threads and reduce the system runtime. The demand for memory and CPU does not affect the real-time performance of radar data processing.

1. Start the data reading thread and read the normalized radar data from the data preprocessing table.

2. According to the radar site and the data type of the normalized radar, start the corresponding thread to complete the processing of the radar data.

Radar data processing completes the field analysis of various radar normalized data, and generates single-radar track and multi-radar track. The processing of this module can be divided into real-time quality control, radar data preprocessing, and radar data processing.

As shown in Figure 4, the radar data processing flow is as follows:

1) Overload processing, site status judgment

Monitor the data source of the radar station in real time, evaluate the data quality of the station and make a correct response in time; monitor the traffic of the radar source data, generate acceptable overflow when overloaded, and do corresponding alarm processing; monitor the dual input The source signal is quality-gated.

Real-time quality control of radar data processing includes real-time quality monitoring of radar station data sources, overload processing and comparison and selection of radar data channels:

a) Real-time quality monitoring of radar station data sources. By processing the status report of the radar ground station, the status value of the station is obtained, the data quality of the station is evaluated, and the status information of the station is stored in the database.

b) Radar data overload processing. The monitoring data processing software accepts the track/point track data of multiple ground stations at the same time, and its single-station track processing and multi-station data fusion processing capabilities are based on the budget flow of each station and radar data, and are based on 50% of the maximum processing capacity of the system used Just can complete the processing task design. In order to prevent the data overload of the radar site, the total amount of various messages of a certain monitoring source can be counted in real time, and it can be evaluated in real time whether it exceeds the allowable flow. Once the set adaptability threshold is exceeded, the software will immediately cut off the data output of the site to the data processing module and the data usage part until it detects that the data flow of the site returns to normal.

c) Radar data channel comparison and selection. Radar data channel comparison and selection is to perform quality gating on dual-input monitoring source signals. According to the set time unit, the quality quantitative indicators of each radar signal are dynamically monitored in real time. When the signal quality is lower than the set indicator threshold, an alarm is generated.

2) Radar data preprocessing

Radar data preprocessing includes data validity check, coordinate transformation and space-time alignment, etc.

a) Validity check

Data validity check includes altitude data check and speed data check. In the altitude speed check, the air pressure altitude can be corrected according to the QNH value, and the QNH data update, QNH partition processing, altitude layer and altitude conversion processing can be performed through automatic processing of meteorological information and manual input.

b) Coordinate transformation

Coordinate conversion is to convert the position, altitude, speed and other data provided in the message into a unified ECEF coordinate system. Coordinate transformation includes coordinate transformation of position information and speed information. The earth-centered earth-fixed coordinate system (ECEF) represents the meridian section of the earth, as shown in Figure 3: C is the center of the earth, N is the north pole, S is the south pole, EF is the equator, HK is the ground plane of the observation point O, and OP is vertical to HK . OM is parallel to SN, and the included angle between it and OH is φ, φ is the geographic latitude of point O, and the angle OPF is equal to φ.

c) Spatiotemporal Calibration

Since the local radars have different scanning periods and synchronous clocks, in addition to a unified coordinate space, a unified time reference is also required to move the kinematic parameters in all track data to a periodicity on the time axis. This is necessary for track association and fusion processing. After space-time alignment, the data of multiple radars are unified into the same coordinate system, and the data are aligned in time.

①Data space alignment technology

The so-called data space alignment is to select a reference coordinate system, and unify data from different radars into this coordinate system. The internal processing of the present invention adopts the ECEF coordinate system for unification.

②Time alignment technology

The so-called time alignment is to interpolate and extrapolate the target observation data collected by each sensor within the same time slice, and calculate the data at the high-precision observation time to the low-precision observation time point. The system uses a simple linear recursion method to interpolate or extrapolate the data to perform time alignment.

3) Radar data processing

Radar data processing includes single radar data tracking processing and multi-radar data fusion processing. The radar data processor can receive the detection data of multiple different radars, adopt the structural model of the distributed multi-sensor information fusion system, and perform track correlation and fusion after the track tracking processing of each radar to form and maintain the system track. The multi-radar data fusion processing technology can improve the probability of target inspection, expand the radar coverage area, increase the reliability of the system, improve the accuracy of target measurement and speed up the effect of system track display refresh.

a) Single radar tracking processing

①Target tracking

For the radar data transformed by coordinates, the system will follow the track tracking rules to track the point tracks into track. The process includes: tracking start, tracking maintenance, tracking end.

The specific algorithm of single radar track tracking is as follows:

●Track start

After the first scan of the radar system, many traces are generated. After the radar system conducts the second scan, the traces of this scan are correlated according to the expected range of the trace correlation gate in the first stage, and the initial track is generated after the correlation is successful.

●Confirm new track

In the subsequent scanning process of the system, according to the secondary code, height and distance of the scanning point track, it is judged whether it is related to the existing track. If there is no relevant track, a new track will be generated. If there is a relevant track, update the track status according to the point track information.

●Track maintenance and state estimation

In the subsequent radar scanning process, use the association rules in the previous step to confirm the track maintenance, and perform track filtering (Kalman filtering technology) to generate a stable single radar track, and at the same time, estimate the next occurrence position. Kalman filtering is a kind of linear filtering, and it needs to be corrected when the target is maneuvering; the purpose of track state filtering is to reduce the error caused by various disturbances as much as possible, and obtain the track with the smallest error.

●Track Extrapolation

When a certain track loses relevant point track data, its status will be changed to extrapolated track, and within the extrapolation range set by the system, if a new related point track appears, the track status will return to normal track.

●Track termination

When no relevant track is found in the N scan cycles, the track is terminated, where the value of N is a preset system parameter (in this embodiment, the value of N is 5).

②Bomen

The wave gate is used to determine whether the dot track in a certain scan period comes from the track established by itself, or belongs to the beginning of a new track. The wave gate is a spatial region centered on the predicted value of a certain radar scan. It is a preliminary verification to determine whether an observation is related to a known track or a new target. Track observation pairing. If only one observation is within a track gate and no other track gate is within that observation, then that observation is associated with the track, and the track is then updated with that observation. If multiple observations are within track gates, or one observation is within multiple gates, further correlation logic processing is required. It is used to initialize a new track when the observation does not satisfy the gate conditions of any known track.

b) Multi-radar fusion processing

Multi-radar data fusion technology performs format transformation, coordinate transformation, and data-related space-time calibration on multiple radar data from different platforms, and realizes fusion on this basis. According to the radar detection accuracy and track tracking quality of each radar's local track, the weighted parameters required for track fusion are calculated to realize the update and maintenance of the system track.

The multi-radar data fusion processing flow is shown in Figure 5, and the specific processing methods are as follows:

① Track association

The track association function uses the weighted statistical distance test method to associate the input single radar track of multiple radars corresponding to the same target with an existing system track or a newly generated system track. The correlation of multi-radar tracks is based on the track identifier/address or call sign and the track number of the local system track, and the real-time statistical judgment is made by using the horizontal distance, secondary code, altitude, speed and heading and other state indexes. The radar track correlation processing method is shown in Figure 6.

Firstly, it is judged whether the track number is related, if the track number is not related, then the task is not the same track, and the subsequent related processing is directly carried out. If the track number is relevant, check the horizontal distance, secondary code (SSR code), altitude, speed and heading, and if these values pass the test (for example, the distance is within the distance deviation range set by the system), then the track can be confirmed. trace association. Otherwise, it enters the correlation judgment. The correlation judgment processing method of the present invention is to first judge the heading, speed and distance, and if it is not within the deviation range preset by the system, it is confirmed that it is not relevant. If the course, speed and distance all pass the inspection, then judge the secondary code and the altitude, and if one of them passes the inspection, then the correlation can be confirmed.

②Radar system track establishment

The establishment method of the system track is similar to that of a single radar track. A new system track start is initiated when a single radar track for the current primary radar within the associated gate cannot be associated with any system track that already exists. When the initial system track is associated continuously, the system track is converted into a confirmed system track. When the system track cannot be associated with any single radar track for multiple consecutive times (the number is a system parameter), the system track is terminated. Before terminating the system track, the track can be extrapolated (the number of extrapolation is 5 times).

③ Data fusion

The present invention adopts the dynamic weighted average method to carry out track fusion, so as to reduce the random position error. When calculating the weighting factor, it is divided into two parts: static factor and dynamic factor. The static part of the weighting factor depends on the radar detection accuracy. The higher the accuracy of the target detected by the radar with high precision, the greater the weighting factor. The dynamic part is related to the quality of single radar track tracking, and the higher the quality, the greater the weighting factor. Since the weighting factor not only takes into account the factors of projection accuracy, but also the accuracy of the target feature value, it can quickly overcome the phenomenon of target position jumping, feature parameter mutation, etc. suppression.

The dynamic weighting algorithm can be expressed by the following formula:

In the formula, Si is the eigenvalue of the i-th single radar track participating in the fusion, a _{i static} is the static weighting factor of the i-th single radar track, and a _{i dynamic} is the dynamic weighting factor of the i-th single radar track , S is the eigenvalue of the integrated track after fusion, and N is the number of simultaneous detection radars.

3. Start the data storage thread, and store the processed single-radar track data and multi-radar track data into the track data table.

Step 4. ADS-B data processing

The flow of ADS-B data processing is similar to the processing of radar data, please refer to the flow of radar data processing. Radar data processing is mainly to complete the processing of CAT001 message. And ADS-B completes the field analysis of the CAT021 message, ADS-B finally generates a single ADS-B track and an ADS-B fusion track, and stores the track in the ADS-B track table.

Step 5. Multi-monitoring source data fusion

1. Distributed system structure

Multi-surveillance source data fusion adopts distributed track fusion architecture, as shown in Figure 7, it is proposed to fuse multi-radar track and ADS-B track, and the quality of the generated track is not lower than that of any monitoring data involved in the fusion process track quality.

In the distributed structure system, the original data is first tracked, predicted, and track correlated to form their own local tracks, and the processing results are sent to the fusion center, where the accurate system track is generated through fusion . Since the distributed structure effectively compresses the amount of data, it reduces the communication load between monitoring and data fusion centers, and the central processor only performs track data fusion, which greatly reduces the burden on the fusion center. Since the distributed structure has the ability of local independent tracking, it also has good global monitoring and evaluation characteristics, and the network transmission and calculation amount are low, which increases the stability and reliability of the system.

In the distributed processing system, the system association should be carried out in the scanning cycle set by the system on the basis of the continuous local track coordinates. Since the selected scanning period may be different from the detection time of the monitoring source, the weighted statistical distance criterion needs to be specially considered according to the difference between the scanning time and the detection time when performing interpolation and extrapolation. Information provided by tracked tracks may also be used to improve correlation performance when tracked tracks are available.

2. System track association

The system track correlation method is similar to the radar track correlation method. It associates the local track data corresponding to the same target obtained from the radar data processing and ADS-B data processing modules with the existing or newly generated multi-radar/ADS -B integrated track. In the process of local track tracking, a unique track number is assigned and maintained to the formed track, therefore, the local track number already associated with it in any generated system track data is later An important basis for cruising trajectory correlation. After the track number association is established, five items of data must be checked and confirmed (horizontal distance, SSR code, altitude, speed, heading) to determine whether to maintain the original relationship or cancel the association relationship.

3. System track tracking

The track tracking method of the system track is similar to that of the radar track. For details, see Radar Track Tracking.

4. Fusion of ADS-B track and radar track

In the multi-monitoring source flight target parallel tracking processing system, the system track is correlated with multiple local tracks, and the parameters of the system track are determined by the associated local radar track or the local ADS-B parameters according to certain parameters. Fusion algorithm Fusion formation. By fusing motion parameters such as position and velocity obtained by observing the same target with multiple monitoring sources, not only the reliability of the result data can be improved, but also the error can be further suppressed.

Multi-monitoring source data fusion uses the same weighted average algorithm as multi-radar data fusion. It gives a weight coefficient to each local radar track or ADS-B track participating in the fusion, and each local track participates in the updating process of the system track according to the weight coefficient. The weight coefficient of the weighted average represents the data quality of each local track participating in the fusion. The higher the quality, the larger the weight coefficient, and the lower the quality, the smaller the weight coefficient.

After each sensor performs data processing on the measured target locally, the estimated value of the target state at time t and the covariance matrix are obtained after time alignment. Track fusion is carried out using the covariance matrix as the system input. The filter covariance matrix given by each radar or ADS-B is different, which represents the difference in the accuracy of track data of different surveillance sources. The weighted fusion algorithm uses the covariance matrix to fuse the tracks.

Claims (8)

1. a kind of parallel tracking processing method of multi-surveillance source flight target, it is characterized in that: comprise the steps: Step 1. Multi-monitoring source data reception; Step 2, Multi-monitoring source data analysis: Discriminate the message type information on the original data received in step 1, then call the data parser to analyze the data, and normalize the parsed message data, and then stored in an internal data table; Step 3. Radar data processing: (1) Real-time quality control; (2) Radar data preprocessing; (3) Radar data processing: including single radar tracking processing and multi-radar fusion processing; Step 4, ADS-B data processing: generate single ADS-B track and ADS-B fusion track, and store the track in the ADS-B track table; Step five, multi-monitoring source data fusion: (1) System track association; (2) System track tracking; (3) Fusion of ADS-B track and radar track. 2. a kind of multi-monitoring source flight target parallel tracking processing method according to claim 1 is characterized in that: the multi-monitoring source data reception described in step 1 comprises the following process: (1) Establish a database connection and read data source information from the data source parameter configuration table; (2) Judging whether the read data gives an abnormal value in the log record: if yes, exit the program; if not, enter the next step; (3) Create corresponding data receivers according to the IP address, port, data type, and data subtype of the data source, and each data receiver is a thread; (4) Start the thread and receive the original data of the corresponding port; (5) The receiver thread stores the received raw data into the original database, waiting for subsequent processing. 3. a kind of multi-surveillance source flight target parallel tracking processing method according to claim 1 is characterized in that: the original data is carried out preprocessing earlier before the message type information discrimination is carried out to the original data described in step 2: when When the slant distance difference of multiple consecutive points of the recorded data is zero, it is regarded as a tracking loss; when the difference between a certain point or several points of the recorded data and the previous and subsequent data is significantly greater than the normal difference, it is regarded as a jump. point; make statistics on the abnormal data, and when the ratio of the abnormal data to the total data is greater than the set system fault tolerance value, the data will be regarded as invalid. 4. a kind of multi-monitoring source flight target parallel tracking processing method according to claim 1 is characterized in that: the data parser described in step 2 adopts version processing strategy when parsing data: when receiving message, According to the SAC and SIC in the message, look up whether the corresponding SAC and SIC in the version cache table have a version record, if there is, use this version to parse the message; if there is no such version or if there is this version but the parsing is unsuccessful, other versions are used Parse from high to low. If one parsing is successful, it will be recorded in the cache corresponding to SAC and SIC for the next parsing. If none of the parsing is successful, a parsing error will be returned. 5. a kind of multi-monitoring source flight target parallel tracking processing method according to claim 1 is characterized in that: the real-time quality control described in step 3 comprises: a) Real-time quality monitoring of radar station data sources: through the radar ground station status report, the station status value is obtained, the data quality of the station is evaluated, and the status information of the station is stored in the database; b) Radar data overload processing: make real-time statistics on the total amount of various messages from the monitoring source and judge whether it exceeds the set allowable flow in real time. If it exceeds, immediately cut off the data output of the station until the data of the station is detected until the flow returns to normal; c) Radar data channel comparison and selection: According to the set time unit, the quantitative indicators of the quality of radar signals of each channel are dynamically monitored in real time, and an alarm will be generated if the signal quality is lower than the set indicator threshold. 6. a kind of multi-surveillance source flight target parallel tracking processing method according to claim 1 is characterized in that: the radar data preprocessing described in step 3 comprises: a) Data validity check: including altitude data check and speed data check; b) Coordinate conversion: convert the position, altitude, speed and other data provided in the message into a unified ECEF coordinate system; c) Spatio-temporal calibration: move the kinematic parameters in all track data to a periodic time point on the time axis. 7. A kind of multi-surveillance source flight target parallel tracking processing method according to claim 1, is characterized in that: the single radar tracking processing described in step 3 comprises: ① Track start: after the first scan of the radar system, the track is generated; after the second scan of the radar system, according to the expected range of the relevant gate of the track generated in the first scan, the track of this scan is carried out. Correlation, the initial track will be generated after the correlation is successful; ②Confirm the new track: In the subsequent scanning process, judge whether it is related to the existing track according to the secondary code, height and distance of the scanned point track: if not, a new track will be generated; if it is related, it will be based on the point track information Update track status; ③ Track maintenance and state estimation: In the subsequent scanning process, use the association rules in the previous step to confirm the track maintenance, and perform track filtering to generate a stable single radar track. estimate; ④Track extrapolation: When a certain track loses relevant track data, its status will be changed to extrapolated track. Within the extrapolation range set by the system, if a new related track appears, the track status will return to normal track; ⑤ Termination of the track: When no relevant point is found in the preset N scanning cycles, the track will be terminated. 8. a kind of multi-surveillance source flight target parallel tracking processing method according to claim 1 is characterized in that: the multi-radar fusion processing described in step 3 comprises: ①Track correlation: firstly determine whether the track numbers are related: if not, then the task is not the same track, and directly perform subsequent related processing; Check, if all pass the test, confirm the track association; otherwise, first judge whether the course, speed and distance are within the deviation range preset by the system, if not, confirm that they are not relevant, if so, continue to check the secondary code or altitude , if one of them passes the test, the track association is confirmed; ② Radar system track establishment: when the single radar track of a current main radar in the relevant wave gate cannot be associated with any existing system track, start to create a new system track; when the initial system track is continuous When a correlation occurs, the system track is converted into a confirmed system track; when the system track cannot be associated with any single radar track within the set consecutive times, the system track is terminated; ③ Data Fusion: Use dynamic weighted average method for track fusion: In the formula, Si is the eigenvalue of the i-th single radar track participating in the fusion, a _{i static} is the static weighting factor of the i-th single radar track, and a _{i dynamic} is the dynamic weighting factor of the i-th single radar track , S is the eigenvalue of the integrated track after fusion, and N is the number of simultaneous detection radars.