CN117949990B - Multisource information fusion measurement wild value detection inhibition method - Google Patents
Multisource information fusion measurement wild value detection inhibition method Download PDFInfo
- Publication number
- CN117949990B CN117949990B CN202410352997.0A CN202410352997A CN117949990B CN 117949990 B CN117949990 B CN 117949990B CN 202410352997 A CN202410352997 A CN 202410352997A CN 117949990 B CN117949990 B CN 117949990B
- Authority
- CN
- China
- Prior art keywords
- measurement
- information
- row
- array
- measurement information
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
- 238000005259 measurement Methods 0.000 title claims abstract description 213
- 238000000034 method Methods 0.000 title claims abstract description 51
- 238000001514 detection method Methods 0.000 title claims abstract description 47
- 230000004927 fusion Effects 0.000 title claims abstract description 47
- 230000005764 inhibitory process Effects 0.000 title claims abstract description 27
- 239000011159 matrix material Substances 0.000 claims abstract description 63
- 238000001914 filtration Methods 0.000 claims abstract description 56
- 238000004364 calculation method Methods 0.000 claims description 25
- 238000012821 model calculation Methods 0.000 claims description 13
- 238000013450 outlier detection Methods 0.000 claims description 11
- 238000010276 construction Methods 0.000 claims description 10
- 238000001279 elastic incoherent neutron scattering Methods 0.000 claims 1
- 230000002401 inhibitory effect Effects 0.000 claims 1
- 230000000875 corresponding effect Effects 0.000 description 26
- 230000008569 process Effects 0.000 description 21
- 238000003491 array Methods 0.000 description 6
- 230000001629 suppression Effects 0.000 description 6
- 238000010586 diagram Methods 0.000 description 5
- 238000012545 processing Methods 0.000 description 4
- 230000008859 change Effects 0.000 description 2
- 238000012937 correction Methods 0.000 description 2
- 230000002596 correlated effect Effects 0.000 description 2
- 238000013461 design Methods 0.000 description 2
- 230000000452 restraining effect Effects 0.000 description 2
- 230000005856 abnormality Effects 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 239000000969 carrier Substances 0.000 description 1
- 230000015556 catabolic process Effects 0.000 description 1
- 230000000295 complement effect Effects 0.000 description 1
- 238000005094 computer simulation Methods 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 238000006731 degradation reaction Methods 0.000 description 1
- 238000012938 design process Methods 0.000 description 1
- 230000008030 elimination Effects 0.000 description 1
- 238000003379 elimination reaction Methods 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 239000000463 material Substances 0.000 description 1
- 238000005457 optimization Methods 0.000 description 1
- 238000004088 simulation Methods 0.000 description 1
- 238000012360 testing method Methods 0.000 description 1
- 239000005436 troposphere Substances 0.000 description 1
- 239000013585 weight reducing agent Substances 0.000 description 1
Landscapes
- Position Fixing By Use Of Radio Waves (AREA)
- Navigation (AREA)
Abstract
The invention discloses a multisource information fusion measurement wild value detection inhibition method, which comprises the steps of firstly, carrying out satellite navigation system information fusion measurement wild value detection inhibition; then performing land-based radio navigation system information fusion measurement wild value detection inhibition; and finally, carrying out SAR radar matching positioning system information fusion measurement wild value detection inhibition. According to the invention, under the condition of not depending on a filtering covariance matrix, the wild value detection inhibition is realized only by measuring the information sequence, the quality of the measured information can be effectively improved, and the precision and stability of the integrated navigation system are ensured.
Description
Technical Field
The invention belongs to the technical field of navigation, and particularly relates to a multisource information fusion measurement wild value detection inhibition method.
Background
In the actual use process, the inertial navigation system has the characteristics of no external interference, high measurement frequency and comprehensive output information, but is influenced by the measurement error of the inertial sensor, and the positioning accuracy is obviously reduced during long-time working. In order to ensure the navigation precision in the use process, the inertial navigation system is generally used as a basis, and the measurement information of other auxiliary equipment is fused, so that the positioning precision of the navigation system in long-time working is improved. For carriers in air flight, auxiliary devices are typically satellite navigation systems, land-based radio navigation systems, SAR radar matched positioning systems, etc. The positioning accuracy of the satellite navigation system is not reduced along with the increase of working time, but satellite signals are weaker and are extremely easy to interfere; the land-based radio navigation system has the characteristics that the measurement accuracy does not decrease along with the increase of working time and the anti-interference capability is strong, but at a low elevation angle, the satellite-based radio navigation system is not only easily blocked by terrain to cause limited viewing distance, but also easily affected by the change of temperature and humidity of the ground of a cross region to cause degradation of correction accuracy of a troposphere; SAR radar matching positioning system is not easy to be interfered by the outside, but the flying height needs to meet certain conditions, and the SAR radar matching positioning system is usually used at the end of carrier flying. Therefore, based on an inertial navigation system, a satellite navigation system, a land-based radio navigation system and an SAR radar matching positioning system are used as auxiliary materials, so that the fusion of multi-source measurement information can be realized, the advantages of the systems are fully exerted, and the aim of complementary advantages is fulfilled.
The multisource information fusion generally takes extended Kalman filtering as an implementation framework, takes an error model of an inertial navigation system as a filtered state equation, takes a difference value between an inertial navigation resolving result and measurement results of auxiliary systems as measurement, and achieves real-time online estimation of errors of the inertial navigation system. The state quantity estimated on line represents the error value of the inertial navigation system, and the error value is fed back to the inertial navigation system, so that the correction of the inertial navigation system can be realized, the accumulated error is eliminated, and the navigation precision is improved. The process of estimating the inertial navigation system error by multi-source information fusion is essentially a process of solving the inertial navigation error based on known measurement information and the propagation rule of the inertial navigation system error. The quality of the measurement information directly influences the accuracy and the robustness of the inertial navigation system estimation, so that the improvement of the quality of the measurement information has important significance.
The measurement information is composed of inertial navigation solution results and measurement results of auxiliary systems. The measurement and output of the inertial navigation system are not interfered by the outside, and the information quality is stable in a short time; the satellite navigation system is influenced by satellite landing and high-power interference equipment, and the accuracy of the output positioning information changes; the land-based radio navigation system is influenced by the antenna direction, carrier attitude, ground station elevation angle and cross-region ground temperature and humidity change, and the output ranging information has a wild point; the SAR radar matching positioning system is influenced by reference map precision and matching precision, and the precision of the output positioning information fluctuates. Before information fusion, the detection and suppression of the wild value of the measurement information of a satellite navigation system, a land-based radio navigation system and an SAR radar matching positioning system are realized, and the method is an important way for improving the quality of the measurement information. Therefore, the detection and inhibition of the wild value of the multisource information fusion measurement have important significance for improving the accuracy and the robustness of the whole navigation system.
Currently, the method for detecting and suppressing the measurement wild value mainly uses chi-square detection. The Kalman detection method mainly comprises the steps of constructing normalized innovation variance through Kalman filtering innovation and innovation covariance matrix, and comparing the innovation variance value with a set detection threshold value to determine whether measurement information related to the innovation is used or not. Wherein, kalman filtering innovation is the difference value between the measurement information and the filtering one-step prediction information; the innovation covariance matrix is obtained by calculating a state one-step prediction covariance matrix, a measurement matrix and a preset measurement noise matrix; the set detection threshold is typically obtained by looking up a table based on the level of salience required.
The state one-step prediction covariance matrix is correlated with a predetermined system noise matrix, and thus the innovation covariance matrix is correlated with the predetermined system noise matrix and the measurement noise matrix. In the Kalman filter design, the system noise matrix and the measurement noise matrix are theoretically required to be in accordance with the noise characteristics of an actual system, but in the actual design process, due to the fact that the system modeling is different from the actual model and the system dynamic error exists, the system noise matrix and the measurement noise matrix are usually preset based on experience and are subjected to iterative optimization by combining mathematical simulation and flight test results. The situation can cause that the preset system noise matrix and the measurement noise matrix are inconsistent with the actual noise characteristics of the system, so that the normalized innovation variance is not accurate enough through the innovation covariance matrix, the detection threshold is set, the table lookup is difficult to determine, and the detection abnormality occurs in chi-square detection.
Disclosure of Invention
In order to overcome the defects of the prior art, the invention provides a multisource information fusion measurement wild value detection inhibition method, which comprises the steps of firstly carrying out satellite navigation system information fusion measurement wild value detection inhibition; then performing land-based radio navigation system information fusion measurement wild value detection inhibition; and finally, carrying out SAR radar matching positioning system information fusion measurement wild value detection inhibition. According to the invention, under the condition of not depending on a filtering covariance matrix, the wild value detection inhibition is realized only by measuring the information sequence, the quality of the measured information can be effectively improved, and the precision and stability of the integrated navigation system are ensured.
The technical scheme adopted for solving the technical problems is as follows:
Step 1: detecting and suppressing the information fusion measurement wild value of the satellite navigation system;
after the navigation system is started, the navigation system is shifted to a combined navigation working state through initial alignment, inertial navigation calculation and Kalman filtering time update are carried out every 5ms after the initialization of the extended Kalman filter, the navigation computer carries out wild value detection according to measurement information periodically provided by the satellite navigation system, and measurement update is carried out according to the detection result.
Step 2: detecting and suppressing land-based radio navigation system information fusion measurement wild value;
The navigation computer carries out outlier detection according to the measurement information periodically provided by the land-based radio navigation system, and carries out measurement updating according to the detection result;
Step 3: detecting and suppressing the SAR radar matching positioning system information fusion measurement wild value;
And the navigation computer performs outlier detection according to measurement information periodically provided by the SAR radar matching positioning system, and performs measurement updating according to the detection result.
Further, the step 1 specifically includes:
Step 1-1: after receiving the measurement information of the satellite navigation system, the navigation computer locates the position according to the satellite Inertial navigation solution position/>Calculating measurement information/>And store the measurement information in array/>In (1)/(w)For a 3 row 1 column matrix, the 3 elements respectively correspond to latitude information, longitude information and altitude information;
（1）
For a matrix of 3 rows and 10 columns,/> Is placed at/>Line 1, element 2 of (i) is placed at/>Line 2, element 3 placed at/>Line 3 of (2);
step 1-2: when the accumulated storage quantity of equivalent measurement information is 10, calculating an array Absolute median difference of elements of each rowAnd executing the subsequent steps, otherwise repeating the steps 1-1,/>The specific calculation formula is as follows:
（2）
In the method, in the process of the invention, Is the median of arrayFor/>Line 1 element,/>For/>Is a function of the element of line 2 of (c),For/>Line 3 element of (2);
From absolute median difference Calculating standard deviation/>, of measurement informationThe following are provided:
（3）
Step 1-3: constructing a weight array ，/>A matrix of 3 rows and 10 columns; judging arrays one by oneMiddle/>Line/>Column element，/>Whether or not in section/>In the section,/>, if in the sectionCorresponding element position 1, otherwise judge/>Whether or not it is in the intervalIf not in the interval,/>Corresponding element position 0, otherwise corresponding element position/>，/>The specific calculation mode is as follows:
（4）
In the method, in the process of the invention, For/>(1 /)Individual elements, if/>In the intervalIn the interior, there areOtherwise there is；
Step 1-4: searching an arrayThe 0 element in the row, all elements in the row are set to 0, and the array/>The measurement information in the system is introduced into measurement update after delay compensation, and the filtering innovation/>The construction is as follows:
（5）
In the method, in the process of the invention, For/>(1 /)Column element,/>Corresponding to the measurement matrix, obtaining by model calculation according to Kalman filtering theory,/>The model is obtained by calculation according to a Kalman filtering theory for one-step prediction state;
Wherein, 、/>、/>Respectively/>(1 /)Line/>Column elements;
step 1-5: filtering the new information Substituting the information fusion measurement wild value detection inhibition of the satellite navigation system by a conventional extended Kalman filtering model, and clearing the array/>, after measurement updating is finishedAnd (5) restarting the steps 1-1 to 1-5.
Further, the step 2 specifically includes:
step 2-1: after receiving the measurement information of the land-based radio navigation system, the navigation computer carries out the distance measurement according to the land-based radio navigation Ground station geocentric geodetic coordinates/>Inertial navigation solution geocentric-geodetic coordinates/>Calculate measurement information/>And store the measurement information in array/>In (1)/(w)Is a scalar:
（6）
Is a1 row 10 column matrix,/> 、/>All are 3 rows and 1 column matrixes;
Step 2-2: when the accumulated storage quantity of equivalent measurement information is 10, calculating a matrix Absolute median difference of elements of each rowAnd executing the subsequent steps, otherwise repeating the steps 2-1,/>The specific calculation formula is as follows:
（7）
From absolute median difference Calculating standard deviation/>, of measurement informationThe following are provided:
（8）
Step 2-3: constructing a weight array ，/>A matrix of 1 row and 10 columns; judging arrays one by oneMiddle/>Column element/>Whether or not in section/>In the section,/>, if in the sectionCorresponding element position 1, otherwise judge/>Whether or not in section/>If not in the interval,/>Corresponding element position 0, otherwise corresponding element position/>，/>The specific calculation mode is as follows:
（9）
In the formula, if In section/>In the interior, there areOtherwise there is/>；
Step 2-4: array is arranged intoThe measurement information in the system is introduced into measurement update after delay compensation, and the filtering innovation/>The construction is as follows:
（10）
In the method, in the process of the invention, Corresponding to the measurement matrix, obtaining by model calculation according to Kalman filtering theory,/>Is a one-step prediction state, and is obtained by model calculation according to Kalman filtering theoryFor/>(1 /)Column elements;
step 2-5: filtering the new information Substituting the measurement field value into an extended Kalman filtering model to realize detection inhibition of the land-based radio navigation system information fusion measurement field value, and clearing an array/>, after measurement updating is finishedAnd (3) restarting the steps 2-1 to 2-5.
Further, the step 3 specifically includes:
Step 3-1: after receiving the measurement information of the SAR radar matching positioning system, the navigation computer positions the position according to the SAR radar Inertial navigation solution position/>Calculate measurement information/>And store the measurement information in array/>Wherein, among them,For a 3 row 1 column matrix, the 3 elements respectively correspond to latitude information, longitude information and altitude information;
（11）
For a3 row 4 column matrix,/> Is placed at/>Line 1, element 2 of (i) is placed at/>Line 2, element 3 placed at/>Line 3 of (2);
Step 3-2: when the accumulated storage quantity of equivalent measurement information is 4, calculating a matrix Absolute median difference of elements of each rowAnd executing the subsequent steps, otherwise repeating the steps 3-1,/>The specific calculation formula is as follows:
（12）
In the method, in the process of the invention, For/>Line 1 element,/>For/>Line 2 element,/>For/>Line 3 element of (2);
From absolute median difference Calculating standard deviation/>, of measurement informationThe following are provided:
（13）
Step 3-3: constructing a weight array ，/>A matrix of 3 rows and 4 columns; judging arrays one by oneMiddle/>Line/>Column element，/>Whether or not in section/>In the section,/>, if in the sectionCorresponding element position 1, otherwise judge/>Whether or not it is in the intervalIf not in the interval,/>Corresponding element position 0, otherwise corresponding element position/>，/>The specific calculation mode is as follows:
（14）
In the method, in the process of the invention, For/>(1 /)Individual elements, if/>In the intervalIn the interior, there areOtherwise there is；
Step 3-4: searching an arrayThe 0 element in the row, all elements in the row are set to 0, and the array/>The measurement information in the system is introduced into measurement update after delay compensation, and the filtering innovation/>The construction is as follows:
（15）
In the method, in the process of the invention, For/>(1 /)Column element,/>Corresponding to the measurement matrix, obtaining by model calculation according to Kalman filtering theory,/>For one-step prediction state, the model calculation is also performed according to the Kalman filtering theory to obtain:
、/>、/> Respectively/> (1 /)Line/>Column elements;
Step 3-5: filtering the new information Substituting the SAR radar matching positioning system information fusion measurement wild value detection inhibition into an extended Kalman filtering model, and clearing an array/>, after measurement updating is finishedAnd (3) restarting the steps from 3-1 to 3-5.
The beneficial effects of the invention are as follows:
According to the invention, under the condition of not depending on a filtering covariance matrix, the wild value detection inhibition is realized only by measuring the information sequence, the quality of the measured information can be effectively improved, and the precision and stability of the integrated navigation system are ensured. The method can effectively detect the suppression wild value, improve the quality of the measurement information and improve the precision and the output stability of the integrated navigation system.
Drawings
Fig. 1 (a) is a schematic diagram of information fusion speed error when no outlier detection is performed.
Fig. 1 (b) is a schematic diagram of an information fusion position error when no outlier detection is performed.
Fig. 2 (a) is a schematic diagram of information fusion speed error when the method of the present invention is used for detecting the outlier.
Fig. 2 (b) is a schematic diagram of an information fusion position error when the method of the present invention is used for detecting a outlier.
Fig. 3 is a flowchart of multi-source information fusion with measurement outlier detection suppression.
Detailed Description
The invention will be further described with reference to the drawings and examples.
How to realize the elimination of the measurement wild value without depending on the filtering covariance matrix has important significance for the design stability and precision of the integrated navigation system. The invention provides a multisource information fusion measurement outlier detection inhibition method which can realize outlier processing only through a measurement sequence under the condition of not depending on an innovation covariance matrix. The method can reject larger wild values, use smaller wild values by weight reduction, and realize wild value detection and suppression during multisource information fusion measurement.
The multisource information fusion measurement wild value detection inhibition method can be implemented in three steps:
Step one: detecting and restraining the information fusion measurement wild value of the satellite navigation system;
after the navigation system is started, the navigation system is shifted to a combined navigation working state through initial alignment according to a flow, inertial navigation calculation and Kalman filtering time update are carried out every 5ms after the initialization of the extended Kalman filter, the navigation computer carries out outlier detection according to measurement information provided by the satellite navigation system periodically (period is 0.1 s), and measurement update is carried out according to a detection result. The specific process is as follows:
1) After receiving the measurement information of the satellite navigation system, the navigation computer locates the position according to the satellite Inertial navigation solution position/>Calculating measurement information/>And store the measurement information in array/>In (1)/(w)For a 3 row 1 column matrix, the 3 elements correspond to latitude information, longitude information, and altitude information, respectively:
（1）
For a matrix of 3 rows and 10 columns,/> Is placed at/>Line 1, element 2 of (i) is placed at/>Line 2, element 3 placed at/>Line 3 of (2);
2) When the accumulated storage quantity of equivalent measurement information is 10, calculating a matrix Absolute median difference/>, of each row of elementsAnd executing the subsequent steps, otherwise repeating step 1),/>The specific calculation formula is as follows:
（2）
In the method, in the process of the invention, Is the median of arrayFor/>Line 1 element,/>For/>Is a function of the element of line 2 of (c),For/>Line 3 element of (2);
From absolute median difference Calculating standard deviation/>, of measurement informationThe following are provided:
（3）
3) Constructing a weight array ，/>A matrix of 3 rows and 10 columns; judging arrays one by oneMiddle/>Line/>Column element/>，Whether or not in section/>In the section,/>, if in the sectionCorresponding element position 1, otherwise judge/>Whether or not it is in the intervalIf not in the interval,/>Corresponding element position 0, otherwise corresponding element position/>，/>The specific calculation mode is as follows:
（4）
In the method, in the process of the invention, For/>(1 /)Individual elements, if/>In the intervalIn the interior, there areOtherwise there is。
4) Searching an arrayThe 0 element in the row, all elements in the row are set to 0, and the array/>The measurement information in the system is introduced into measurement update after delay compensation, and the filtering innovation/>The construction is as follows:
（5）
In the method, in the process of the invention, For/>(1 /)Column element,/>Corresponding measurement matrix can be obtained by model calculation according to Kalman filtering theoryThe prediction state is a one-step prediction state, and can be obtained by model calculation according to a Kalman filtering theory;
Wherein, 、/>、/>Respectively/>(1 /)Line/>Column elements;
5) Filtering the new information Substituting the information fusion measurement wild value detection inhibition of the satellite navigation system into a conventional extended Kalman filtering model, and emptying the array after measurement updating is finishedAnd (5) rerun the steps 1) to 5).
Stage two: detecting and restraining the land-based radio navigation system information fusion measurement wild value;
The navigation computer detects the wild value according to the measurement information periodically provided by the land-based radio navigation system, and carries out measurement updating according to the detection result. The specific process is as follows:
1) After receiving the measurement information of the land-based radio navigation system, the navigation computer carries out the distance measurement according to the land-based radio navigation Ground station geocentric geodetic coordinates/>Inertial navigation solution geocentric-geodetic coordinates/>Calculate measurement information/>And store the measurement information in array/>In (1)/(w)Is a scalar:
（6）
Is a1 row 10 column matrix,/> 、/>All are 3 rows and 1 column matrixes;
2) When the accumulated storage quantity of equivalent measurement information is 10, calculating a matrix Absolute median difference/>, of each row of elementsAnd executing the subsequent steps, otherwise repeating step 1),/>The specific calculation formula is as follows:
（7）
From absolute median difference Calculating standard deviation/>, of measurement informationThe following are provided:
（8）
3) Constructing a weight array ，/>A matrix of 1 row and 10 columns; judging arrays one by oneMiddle/>Column element/>Whether or not in section/>In the section,/>, if in the sectionCorresponding element position 1, otherwise judgeWhether or not in section/>If not in the interval,/>Corresponding element position 0, otherwise corresponding element position/>，/>The specific calculation mode is as follows:
（9）
In the formula, if In section/>In the interior, there areOtherwise there is/>；
4) Array is arranged intoThe measurement information in the system is introduced into measurement update, delay compensation and the like are needed before the update, and filtering is carried out on the information/>The construction is as follows:
（10）
In the method, in the process of the invention, Corresponding measurement matrix can be obtained by model calculation according to Kalman filtering theoryIs a one-step prediction state, and can be obtained by model calculation according to Kalman filtering theoryFor/>(1 /)Column elements.
5) Filtering the new informationSubstituting the information fusion measurement wild value detection inhibition of the land-based radio navigation system by a conventional extended Kalman filtering model, and clearing an array/>, after measurement updating is finishedAnd (5) rerun the steps 1) to 5).
Stage three: SAR radar matching positioning system information fusion measurement wild value detection inhibition section;
And the navigation computer performs outlier detection according to measurement information periodically provided by the SAR radar matching positioning system, and performs measurement updating according to the detection result. The specific process is as follows:
1) After receiving the measurement information of the SAR radar matching positioning system, the navigation computer positions the position according to the SAR radar Inertial navigation solution position/>Calculate measurement information/>And store the measurement information in array/>In (1)/(w)For a 3 row 1 column matrix, the 3 elements respectively correspond to latitude information, longitude information and altitude information;
（11）
For a3 row 4 column matrix,/> Is placed at/>Line 1, element 2 of (i) is placed at/>Line 2, element 3 placed at/>Line 3 of (2);
2) When the accumulated storage quantity of equivalent measurement information is 4, calculating a matrix Absolute median difference/>, of each row of elementsAnd executing the subsequent steps, otherwise repeating step 1),/>The specific calculation formula is as follows:
（12）
In the method, in the process of the invention, For/>Line 1 element,/>For/>Line 2 element,/>For/>Line 3 element of (2);
From absolute median difference Calculating standard deviation/>, of measurement informationThe following are provided:
（13）
3) Constructing a weight array ，/>A matrix of 3 rows and 4 columns; judging arrays one by oneMiddle/>Line/>Column element/>，Whether or not in section/>In the section, if the section is in the section,Corresponding element position 1, otherwise judge/>Whether or not it is in the intervalIf not in the interval,/>Corresponding element position 0, otherwise corresponding element position/>，/>The specific calculation mode is as follows:
（14）
In the method, in the process of the invention, For/>(1 /)Individual elements, if/>In the intervalIn the interior, there areOtherwise there is；
4) Searching an arrayThe 0 element in the row, all elements in the row are set to 0, and the array/>The measurement information in the system is introduced into measurement update, delay compensation is needed before the update, and filtering is performed on the information/>The construction is as follows:
（15）
In the method, in the process of the invention, For/>(1 /)Column element,/>Corresponding measurement matrix can be obtained by model calculation according to Kalman filtering theoryFor the one-step prediction state, the model calculation can also be used for obtaining according to the Kalman filtering theory:
、/>、/> Respectively/> (1 /)Line/>Column elements;
5) Filtering the new information Substituting the SAR radar matching positioning system information fusion measurement wild value detection inhibition into a conventional extended Kalman filtering model, and clearing an array/>, after measurement updating is finishedAnd (5) rerun the steps 1) to 5).
Examples:
taking a navigation system of a certain air flight carrier as an example, a specific implementation mode of a multi-source information fusion navigation system with a measurement wild value detection and suppression function is described, and a simple flow chart is shown in fig. 3. The navigation system is based on an inertial navigation system, and adopts a satellite navigation system, a land-based radio navigation system and an SAR radar matching positioning system as auxiliary equipment to carry out multi-source information fusion. The inertial sensor adopted by the inertial navigation system is a navigation level sensor.
Step 1: after the navigation system is electrified, the navigation computer executes initial alignment according to the flow, and after the alignment is finished, the navigation is started, and the integrated navigation working state is entered, and inertial navigation calculation and time update are carried out every 5ms period. After the carrier takes off, the inertial navigation system fuses the measurement information of the satellite navigation system, the land-based radio navigation system and the SAR radar matching positioning system to navigate.
Step 2: after receiving the measurement information of the satellite navigation system, the navigation computer locates the position according to the satelliteInertial navigation solution position/>Calculate measurement information/>And store the measurement information in array/>Is a kind of medium. When the accumulated storage quantity of the measurement information of the satellite navigation system is smaller than 10, only the measurement information is stored, and measurement update is not carried out. When the accumulated storage quantity of the measurement information of the satellite navigation system is equal to 10, starting measurement update, and calculating the standard deviation/>, of the measurement informationThen building a weight arrayJudge/>Whether or not in section/>In the section, if the section is in the section,Corresponding element position 1, otherwise judge/>Whether or not it is in the intervalIf not in the interval,/>Corresponding element position 0, otherwise corresponding element position/>Find array/>In the column where the 0 element is located, all elements of the column are set to 0.
Step 3: after receiving the measurement information of the land-based radio navigation system, the navigation computer measures the distance according to the land-based radio navigationGround station geocentric geodetic coordinates/>Inertial navigation solution geocentric-geodetic coordinates/>Calculating metrology informationAnd store the measurement information in array/>Is a kind of medium. When the accumulated storage quantity of the measurement information of the land-based radio navigation system is smaller than 10, only the measurement information is stored, and no measurement update is performed. When the accumulated storage quantity of the measurement information of the land-based radio navigation system is equal to 10, starting measurement update, and calculating the standard deviation/>, of the measurement informationWeight array/>, is then constructedJudge/>Whether or not in section/>In the section,/>, if in the sectionCorresponding element position 1, otherwise judge/>Whether or not in section/>If not in the interval,/>Corresponding element position 0, otherwise corresponding element position/>。
Step 4: after receiving the measurement information of the SAR radar matching positioning system, the navigation computer positions the position according to the SAR radarInertial navigation solution position/>Calculate measurement information/>And store the measurement information in array/>Is a kind of medium. When the accumulated storage quantity of the measurement information of the SAR radar matching positioning system is smaller than 4, only the measurement information is stored, and the measurement information of the SAR radar matching positioning system does not meet the measurement updating condition. When the accumulated storage quantity of the measurement information of the SAR radar matching positioning system is equal to 4, the measurement information of the SAR radar matching positioning system meets the measurement updating condition, and the standard deviation/> -of the measurement information is calculatedWeight array/>, is then constructedJudge/>Whether or not in section/>In the section,/>, if in the sectionCorresponding element position 1, otherwise judge/>Whether or not it is in the intervalIf not in the interval,/>Corresponding element position 0, otherwise corresponding element position/>Find array/>In the column where the 0 element is located, all elements of the column are set to 0.
Step 5: if the satellite navigation system measurement information meets the measurement updating condition, calculating a value based on Kalman filteringAnd/>Measurement/>Weight array/>Construction of New information/>Based on the sequential processing mode, substituting 10 groups of measurement information into measurement update, and clearing the array/>, after the measurement update is finished。
Step 6: if the land-based radio navigation system measurement information meets the measurement update condition, calculating a value based on Kalman filteringAnd/>Measurement/>Weight array/>Construction of New information/>Based on the sequential processing mode, substituting 10 groups of measurement information into measurement update, and clearing the array/>, after the measurement update is finished。
Step 7: if the SAR radar matching positioning system measurement information meets the measurement updating condition, calculating a value based on Kalman filteringAnd/>Measurement/>Weight array/>Construction of New information/>Based on the sequential processing mode, substituting 4 groups of measurement information into measurement update, and clearing the array/>, after the measurement update is finished。
As shown in fig. 1 (a) and 1 (b), the information fusion speed and the position error when the outlier detection is not performed are schematic diagrams; as shown in fig. 2 (a) and 2 (b), the information fusion speed and the position error when the method of the present invention is used for detecting the outlier are shown. The result shows that the method can effectively detect the suppression wild value, improve the quality of the measurement information and improve the precision and the output stability of the integrated navigation system.
Claims (1)
1. The method for detecting and inhibiting the multisource information fusion measurement wild value is characterized by comprising the following steps of:
Step 1: detecting and suppressing the information fusion measurement wild value of the satellite navigation system;
After the navigation system is started, the navigation system is shifted to a combined navigation working state through initial alignment, inertial navigation calculation and Kalman filtering time update are carried out every 5ms after the initialization of the extended Kalman filter, the navigation computer carries out wild value detection according to measurement information periodically provided by the satellite navigation system, and measurement update is carried out according to the detection result;
Step 1-1: after receiving the measurement information of the satellite navigation system, the navigation computer calculates measurement information Z P according to the satellite positioning position P BD, the inertial navigation resolving position P INS and stores the measurement information in an array A, wherein Z P is a3 row 1 column matrix, and 3 elements respectively correspond to latitude information, longitude information and altitude information;
ZP＝PINS-PBD (1)
A is a matrix of 3 rows and 10 columns, the 1 st element of Z P is placed on the 1 st row of A, the 2 nd element is placed on the 2 nd row of A, and the 3 rd element is placed on the 3 rd row of A;
Step 1-2: when the accumulated storage quantity of equivalent measurement information is 10, calculating an absolute middle bit difference A MAD of each row of elements in the array A, and executing the subsequent steps, otherwise, repeating the steps 1-1 and A MAD, wherein the specific calculation formula is as follows:
wherein medium is the median of the groups, a 1,： is the line 1 element of a, a 2,： is the line 2 element of a, a 3,： is the line 3 element of a;
Standard deviation a σ of the measurement information is calculated from the absolute middle potential difference a MAD as follows:
Aσ＝1.4826AMAD (3)
Step 1-3: constructing a weight array AK, wherein AK is a matrix with 3 rows and 10 columns; judging elements A i,j of the j th row and the j th column in the array A one by one, wherein i=1, 2 and 3; j=1, 2, …,10 is within the interval [ mean (a i,：)-Aσ(i),median(Ai,：)+Aσ (i) ], if so, AK corresponds to element position 1, otherwise, it is determined whether a i,j is within the interval [ mean (a i,：)-2Aσ(i),median(Ai,：)+2Aσ (i) ], if not, AK corresponds to element position 0, otherwise, it corresponds to element position λ A,λA as follows:
Where a σ (i) is the i-th element of a σ, if a i,j is within the interval [ medium (a i,：)+Aσ(i),median(Ai,：)+2Aσ (i) ] then there is Δκ A＝|A1,j-(median(A1,：)+Aσ (1)) ] or else there is Δκ A＝|A1,j-(median(A1,：)-Aσ (1));
Step 1-4: searching the row of 0 element in the array AK, setting all elements in the row to 0, delay compensating the measurement information in the array A, introducing measurement update, and filtering new information The construction is as follows:
Wherein A ：,j is the j-th element of A, H j corresponds to the measurement matrix, and is obtained by calculation of a model according to a Kalman filtering theory, The model is obtained by calculation according to a Kalman filtering theory for one-step prediction state;
Wherein AK 1,j、AK2,j、AK3,j is the i-th row and j-th column element of AK respectively;
Step 1-5: substituting the filter innovation (XI j) into a conventional extended Kalman filter model to realize detection inhibition of information fusion measurement wild values of a satellite navigation system, and clearing an array A to re-operate the steps 1-1 to 1-5 after measurement updating is finished;
step 2: detecting and suppressing land-based radio navigation system information fusion measurement wild value;
The navigation computer carries out outlier detection according to the measurement information periodically provided by the land-based radio navigation system, and carries out measurement updating according to the detection result;
Step 2-1: after receiving the measurement information of the land-based radio navigation system, the navigation computer calculates measurement information Z L according to the land-based radio navigation ranging value L LD, the ground station geocentric and geocentric fixed system coordinate E DM and the inertial navigation resolving geocentric and geocentric fixed system coordinate E INS, and stores the measurement information in an array B, wherein Z L is a scalar:
B is a1 row and 10 column matrix, E DM、EINS is a3 row and 1 column matrix;
step 2-2: when the accumulated storage quantity of the equivalent measurement information is 10, calculating an absolute middle bit difference B MAD of each row of elements in the matrix B, and executing the subsequent steps, otherwise, repeating the specific calculation formulas of the steps 2-1 and B MAD as follows:
BMAD＝median(|B-median(B)|) (7)
The standard deviation B σ of the measurement information is calculated from the absolute middle bit difference B MAD as follows:
Bσ＝1.4826BMAD (8)
Step 2-3: constructing a weight array BK, wherein BK is a1 row and 10 column matrix; judging whether the j-th column element B j in the array B is in the interval [ medium (B) -B σ,median(B)+Bσ ] one by one, if so, judging whether B j is in the interval [ medium (B) -2B σ,median(B)+2Bσ ], if not, corresponding the BK to the element position 0, otherwise, corresponding the element position lambda B,λB, wherein the specific calculation mode is as follows:
Wherein if B j is within the interval [ medium (B) -B σ,median(B)+2Bσ ], then there is Δκ B＝|Bj-(median(B)+Bσ) |, otherwise there is Δκ B＝|Bj-(median(B)-Bσ) |;
Step 2-4: the measurement information in the array B is introduced into measurement update after delay compensation, and the filtering innovation theta j is constructed as follows:
Wherein H j corresponds to the measurement matrix, is obtained by model calculation according to the Kalman filtering theory, The model is obtained by calculation according to a Kalman filtering theory, wherein BK j is the j-th element of BK;
Step 2-5: substituting the filtered innovation Θ j into an extended Kalman filtering model to realize detection and inhibition of information fusion measurement wild values of the land-based radio navigation system, and clearing an array B to re-operate the steps 2-1 to 2-5 after measurement updating is finished;
Step 3: detecting and suppressing the SAR radar matching positioning system information fusion measurement wild value;
The navigation computer carries out outlier detection according to the measurement information periodically provided by the SAR radar matching positioning system, and carries out measurement updating according to the detection result;
Step 3-1: after receiving the measurement information of the SAR radar matching positioning system, the navigation computer calculates measurement information Z SAR according to the SAR radar positioning position P SAR and the inertial navigation resolving position P INS, and stores the measurement information in an array C, wherein Z SAR is a 3 row and 1 column matrix, and 3 elements respectively correspond to latitude information, longitude information and altitude information;
ZSAR＝PSAR-PBD (11)
c is a 3-row 4-column matrix, the 1 st element of Z SAR is placed on the 1 st row of C, the 2 nd element is placed on the 2 nd row of C, and the 3 rd element is placed on the 3 rd row of C;
Step 3-2: when the accumulated storage quantity of equivalent measurement information is 4, calculating the absolute middle bit difference C MAD of each row of elements in the matrix C, and executing the subsequent steps, otherwise, repeating the specific calculation formulas of the steps 3-1 and C MAD as follows:
Wherein C 1,： is a1 st line element of C, C 2,： is a2 nd line element of C, and C 3,： is a3 rd line element of C;
The standard deviation C σ of the measurement information is calculated from the absolute middle potential difference C MAD as follows:
Cσ＝1.4826CMAD (13)
Step 3-3: constructing a weight array CK, wherein the CK is a matrix of 3 rows and 4 columns; judging whether an ith row and an mth column of elements C i,m in the array C, wherein m=1, 2,3 and 4 are in a section [ medium (C i,：)-Cσ(i),median(Ci,：)+Cσ (i) ] one by one, if so, judging whether C i,m is in the section [ medium (C i,：)-2Cσ(i),median(Ci,：)+2Cσ (i) ] or not, if not, judging that the CK corresponds to the element position 0, otherwise, calculating the corresponding element position lambda C,λC specifically as follows:
Where C σ (i) is the i-th element of C σ, if C i,m is within the interval [ medium (C i,：)+Cσ(i),median(Ci,：)+2Cσ (i) ] then there is Δk C＝|C1,j-(median(C1,：)+Cσ (1))|, otherwise there is Δk C＝|C1,j-(median(C1,：)-Cσ (1))|;
Step 3-4: searching a row in which 0 element in the array CK is located, setting all elements in the row to 0, performing delay compensation on measurement information in the array C, introducing measurement update, and constructing a filtering innovation omega m as follows:
Wherein, C ：,m is the j-th element of C, H m corresponds to the measurement matrix, and is obtained by calculation of a model according to the Kalman filtering theory, For one-step prediction state, the model calculation is also performed according to the Kalman filtering theory to obtain:
CK 1,m、CK2,m、CK3,m is the ith row m column element of CK, respectively;
Step 3-5: substituting the filtering innovation omega m into an extended Kalman filtering model to realize detection inhibition of the SAR radar matching positioning system information fusion measurement wild value, and clearing the array C to re-operate the steps 3-1 to 3-5 after measurement updating is finished.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410352997.0A CN117949990B (en) | 2024-03-26 | 2024-03-26 | Multisource information fusion measurement wild value detection inhibition method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410352997.0A CN117949990B (en) | 2024-03-26 | 2024-03-26 | Multisource information fusion measurement wild value detection inhibition method |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN117949990A CN117949990A (en) | 2024-04-30 |
| CN117949990B true CN117949990B (en) | 2024-06-14 |
Family
ID=90792462
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202410352997.0A Active CN117949990B (en) | 2024-03-26 | 2024-03-26 | Multisource information fusion measurement wild value detection inhibition method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN117949990B (en) |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105628024A (en) * | 2015-12-29 | 2016-06-01 | 中国电子科技集团公司第二十六研究所 | Single person positioning navigator based on multi-sensor fusion and positioning and navigating method |
| CN108226980A (en) * | 2017-12-23 | 2018-06-29 | 北京卫星信息工程研究所 | Difference GNSS and the adaptive close coupling air navigation aids of INS based on Inertial Measurement Unit |
Family Cites Families (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US7400246B2 (en) * | 2006-04-11 | 2008-07-15 | Russell Mark Breeding | Inertial Sensor Tracking System |
| CN100575877C (en) * | 2007-12-12 | 2009-12-30 | 南京航空航天大学 | Spacecraft shading device combined navigation methods based on many information fusion |
| CN102914308B (en) * | 2012-10-24 | 2014-12-10 | 南京航空航天大学 | Anti-outlier federated filtering method based on innovation orthogonality |
| CN103791916B (en) * | 2014-01-28 | 2016-05-18 | 北京融智利达科技有限公司 | A kind of combination onboard navigation system based on MEMS inertial navigation |
| CN103983996A (en) * | 2014-05-09 | 2014-08-13 | 北京航空航天大学 | Tight-integration adaptive filtering method of resisting to outliers of global positioning system, |
| CN104833352B (en) * | 2015-01-29 | 2016-06-29 | 西北工业大学 | High-precision vision/inertia integrated navigation method under multimedium complex environment |
| RU2584368C1 (en) * | 2015-02-13 | 2016-05-20 | Открытое акционерное общество "Лётно-исследовательский институт имени М.М. Громова" | Method of determining control values of parameters of spatial-angular orientation of aircraft on routes and pre-aerodrome zones in flight tests of pilot-navigation equipment and system therefor |
| CN111366151A (en) * | 2018-12-26 | 2020-07-03 | 北京信息科技大学 | Information fusion method for ship navigation in polar region |
| CN109724599B (en) * | 2019-03-12 | 2023-08-01 | 哈尔滨工程大学 | Wild value resistant robust Kalman filtering SINS/DVL integrated navigation method |
| CN110823215B (en) * | 2019-10-25 | 2021-08-10 | 南京航空航天大学 | Unmanned aerial vehicle relative navigation information fusion method |
| CN115143954B (en) * | 2022-09-05 | 2022-11-29 | 中国电子科技集团公司第二十八研究所 | Unmanned vehicle navigation method based on multi-source information fusion |
-
2024
- 2024-03-26 CN CN202410352997.0A patent/CN117949990B/en active Active
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105628024A (en) * | 2015-12-29 | 2016-06-01 | 中国电子科技集团公司第二十六研究所 | Single person positioning navigator based on multi-sensor fusion and positioning and navigating method |
| CN108226980A (en) * | 2017-12-23 | 2018-06-29 | 北京卫星信息工程研究所 | Difference GNSS and the adaptive close coupling air navigation aids of INS based on Inertial Measurement Unit |
Also Published As
| Publication number | Publication date |
|---|---|
| CN117949990A (en) | 2024-04-30 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108226980B (en) | Differential GNSS and INS self-adaptive tightly-coupled navigation method based on inertial measurement unit | |
| CN109781099B (en) | Navigation method and system of self-adaptive UKF algorithm | |
| CN109738917B (en) | Multipath error weakening method and device in Beidou deformation monitoring | |
| CN110823217B (en) | Combined navigation fault tolerance method based on self-adaptive federal strong tracking filtering | |
| US9035826B2 (en) | Satellite differential positioning receiver using multiple base-rover antennas | |
| CN110779521A (en) | Multi-source fusion high-precision positioning method and device | |
| CN107193028B (en) | Kalman relative positioning method based on GNSS | |
| CN109471146B (en) | Self-adaptive fault-tolerant GPS/INS integrated navigation method based on LS-SVM | |
| CN114166221B (en) | Auxiliary transportation robot positioning method and system in dynamic complex mine environment | |
| US20230296786A1 (en) | System and method for validating gnss ambiguities | |
| CN115390096A (en) | Low-orbit satellite real-time relative orbit determination method based on full-view satellite-borne GNSS (Global navigation satellite System) receiving system | |
| CN115480279A (en) | GNSS navigation method and terminal, integrated navigation system and storage medium | |
| CN110208843B (en) | Fault-tolerant navigation method based on augmented pseudo-range information assistance | |
| CN106371092B (en) | It is a kind of that the deformation monitoring method adaptively combined is observed with strong-motion instrument based on GPS | |
| CN112697154B (en) | Self-adaptive multi-source fusion navigation method based on vector distribution | |
| CN117949990B (en) | Multisource information fusion measurement wild value detection inhibition method | |
| CN108205151B (en) | Low-cost GPS single-antenna attitude measurement method | |
| US20220244407A1 (en) | Method for Generating a Three-Dimensional Environment Model Using GNSS Measurements | |
| CN110082797B (en) | Multi-system GNSS data static post-processing high-dimensional ambiguity fixing method | |
| CN115792979A (en) | Satellite step-by-step satellite selection method based on PDOP contribution degree | |
| CN112904382B (en) | Laser odometer-assisted rapid optimization satellite selection method under urban canyon environment | |
| CN115014321A (en) | Bionic polarization multi-source fusion orientation method based on adaptive robust filtering | |
| CN113970772A (en) | Multi-frequency BDS-2/BDS-3/INS vehicle-mounted combined positioning method for urban environment | |
| CN110941002A (en) | Self-adaptive anti-difference sequential least square precise point positioning method | |
| CN117031521B (en) | Elastic fusion positioning method and system in indoor and outdoor seamless environment |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant |