JP2019184360A - Sensor device, error detection method, and error detection program - Google Patents

Abstract

To obtain a sensor device with which it is possible to detect an error occurring among a plurality of sensors while suppressing an increase in scale of the device.SOLUTION: A sensor device comprises: a first arithmetic unit 20 for generating a partial locus on the basis of a first observation result and generating a first error covariance matrix; a transmission determination unit 40 for determining whether or not to transmit the partial locus; a second arithmetic unit 50 for generating locus information on the basis of the partial locus received from the transmission determination unit 40 and the other sensor devices 101, and generating a second error covariance matrix; a third arithmetic unit 60 for generating a third error covariance matrix on the basis of the first observation result; an error covariance determination unit 70 for comparing the third error covariance matrix with the second error covariance matrix, and determining whether or not the first observation result includes an error with respect to a second observation result obtained by the other sensor devices 101; and a system alignment error estimation unit 30 for estimating an error amount using the determination result of the error covariance determination unit 70.SELECTED DRAWING: Figure 2

Description

  The present invention relates to a sensor device for observing a target, an error detection method, and an error detection program.

  When the target object (hereinafter referred to as the target) is a flying object such as an aircraft, the detection distance becomes shorter for detection with a single sensor due to the stealth of the target and the target flying at a low altitude. There is a problem. Patent Document 1 discloses a technique for extending a detection distance by sharing a wake obtained by each sensor in a system including a plurality of sensors.

JP 2014-174108 A

  In a system having a plurality of sensors, each sensor obtained a system alignment error, which is an error caused by the installation position of each sensor, the orientation set at the time of installation of each sensor, the measurement time of each sensor, and the like. May be included in the data. As a method of detecting a system alignment error, there is a method of comparing data obtained by each sensor. However, when the number of sensors included in the system increases, there is a problem that the amount of processing increases because the comparison processing increases geometrically.

  The present invention has been made in view of the above, and an object of the present invention is to obtain a sensor device, an error detection method, and an error detection program capable of detecting an error occurring between a plurality of sensors while suppressing a processing amount. And

  In order to solve the above-described problems and achieve the object, the present invention is a sensor device that constitutes a sensor system together with other sensor devices. The sensor device generates a partial wake that is a wake of the target in the first period based on a first observation result of a sensor that observes the target, and generates a first error covariance matrix of the partial wake. And a transmission determination unit that determines whether or not to transmit a partial wake. Further, the sensor device generates wake information that is a target wake after the start of observation based on the partial wake received from the transmission determination unit and the partial wake received from another sensor device. A second arithmetic unit for generating a second error covariance matrix, and a third error covariance matrix for generating a third error covariance matrix of a target track after starting observation based on the first observation result And an arithmetic unit. In addition, the sensor device compares the third error covariance matrix and the second error covariance matrix, and the first observation result includes an error with respect to the second observation result obtained by another sensor device. An error that estimates an error amount included in the first observation result based on the first error covariance matrix using the determination result of the error covariance determination unit and the error covariance determination unit And an estimation unit.

  According to the present invention, there is an effect that the sensor device can detect an error occurring between a plurality of sensors while suppressing the processing amount.

The figure which shows the concept of the system alignment error in the system provided with a some sensor Block diagram showing a configuration example of a sensor system Flowchart showing operation for detecting system alignment error in sensor device The figure which shows the example of the error covariance matrix produced | generated with a sensor apparatus The flowchart which shows the process which determines whether a system alignment error is contained in an error covariance determination part The figure which shows the example in the case of comprising the processing circuit with which a sensor apparatus is equipped with a processor and memory The figure which shows the example in the case of comprising the processing circuit with which a sensor apparatus is equipped with exclusive hardware

  Hereinafter, a sensor device, an error detection method, and an error detection program according to an embodiment of the present invention will be described in detail with reference to the drawings. Note that the present invention is not limited to the embodiments.

Embodiment.
First, the concept of system alignment error in a system including a plurality of sensors will be described. FIG. 1 is a diagram illustrating a concept of system alignment error in a system including a plurality of sensors. Data obtained by the first sensor and the second sensor may include a system alignment error in each element such as position, direction, and time. The system alignment error is not a random error that is reduced by a smooth calculation such as a Kalman filter, but a bias error included in a sensor installation position, a set orientation, a deviation from a reference time at the time of observation, and the like. FIG. 1 shows an example in which the orientation set by each sensor is shifted and a system alignment error is included in the orientation element.

  The first sensor transmits the target track to the second sensor in order to share the target track, which is the track of the observed target, with the second sensor. At this time, as shown in the left diagram of FIG. 1, the first sensor considers that the direction of the second sensor coincides with the direction of the first sensor, and transmits the target track. For the target track transmitted from the first sensor, the position of the target track is specified based on the azimuth set in the first sensor. In practice, however, the orientation of the second sensor is different from the orientation of the first sensor, as shown in the right diagram of FIG. That is, there is a system alignment error between the orientation of the second sensor and the orientation of the first sensor. Thus, if there is a system alignment error, the target track cannot be recognized accurately. In the example of FIG. 1, the second sensor recognizes that the target track received from the first sensor is at the position of the solid arrow shown in the right diagram of FIG. Therefore, it is desirable to detect and correct system alignment errors.

  In the sensor system of the present embodiment, a system alignment error is detected in the sensor device. Hereinafter, the configuration and operation of the sensor system of the present embodiment will be described. In the following description, the system alignment error may be simply referred to as an error.

  FIG. 2 is a block diagram illustrating a configuration example of the sensor system 1 according to the present embodiment. The sensor system 1 includes sensor devices 100 and 101 and a command and control device 110. First, the configuration of the sensor device 100 will be described. The sensor device 100 includes a sensor 10, a first calculation unit 20, a system alignment error estimation unit 30, a transmission determination unit 40, a second calculation unit 50, and a system alignment error detection device 120. The system alignment error detection device 120 includes a third calculation unit 60 and an error covariance determination unit 70.

  The sensor 10 observes the target 80 and outputs an observation result including the observation position of the target 80 to the first calculation unit 20 and the third calculation unit 60. The sensor 10 is, for example, a radar that observes a flying object, but is not limited to this. An observation result obtained by the sensor 10 may be referred to as a first observation result.

  The first calculation unit 20 performs smoothing processing and prediction processing on the observation result output from the sensor 10 to generate a partial wake that is a wake of the target 80 in the first period. The first period is a period from the start of observation by the sensor 10 in the sensor device 100 to the present, or a period from the reset of the observation result held by the first arithmetic unit 20 to 0 to the present. In addition, the first arithmetic unit 20 generates a first error covariance matrix that is an error covariance matrix of the partial wake. Details of the operation of the first arithmetic unit 20 will be described later. The first calculation unit 20 outputs the partial wake and the first error covariance matrix to the system alignment error estimation unit 30.

  The system alignment error estimator 30 is an error estimator that acquires a determination result from the error covariance determiner 70 and estimates an error amount included in the observation result using the determination result and the first error covariance matrix. . The system alignment error estimation unit 30 corrects the partial track and the first error covariance matrix based on the estimated error amount. The system alignment error estimation unit 30 outputs the corrected partial track and the corrected first error covariance matrix to the transmission determination unit 40. Note that the system alignment error estimation unit 30 corrects only the partial track based on the estimated error amount and does not need to correct the first error covariance matrix.

  The transmission determination unit 40 acquires the partial track and the first error covariance matrix from the system alignment error estimation unit 30, acquires the track information and the second error covariance matrix from the second calculation unit 50, and acquires the partial track Is determined whether or not to transmit. The wake information is a wake generated by the second calculation unit 50, and is a wake of the target 80 in a period from when the sensor 10 starts observation in the sensor device 100 to the present. The transmission determination unit 40 uses the partial wake based on the partial wake, the wake information, the first error covariance matrix, and the second error covariance matrix, so that the accuracy of the wake information is increased in the second arithmetic unit 50. When it determines with improving, a partial track is transmitted to the 2nd calculating part 50. FIG. At this time, the transmission determination unit 40 also transmits the partial wake to the second calculation unit 51 included in the sensor device 101 and the second calculation unit 52 included in the command and control device 110.

  Based on the partial track received from the transmission determination unit 40 and the partial track received from the transmission determination unit 41 of the sensor device 101 described later, the second calculation unit 50 performs the same smoothing process as the first calculation unit 20 and A prediction process is performed, and wake information is generated, that is, updated. In addition, the second calculation unit 50 generates, that is, updates, a second error covariance matrix that is an error covariance matrix of wake information. The second calculation unit 50 notifies the first calculation unit 20 that the partial wake has been received from the transmission determination unit 40. In this case, the first calculation unit 20 receives the notification that the partial track has been received from the second calculation unit 50, and resets the observation result acquired and held from the sensor 10 to 0. In addition, the second calculation unit 50 outputs the updated track information and the second error covariance matrix to the transmission determination unit 40. In addition, the second calculation unit 50 outputs the updated second error covariance matrix to the error covariance determination unit 70.

  The third calculation unit 60 performs the same smoothing process and prediction process as the first calculation unit 20 based on the observation result output from the sensor 10, and after the sensor 10 starts observation in the sensor device 100. The wake of the target 80 in the period until now is generated. The third calculation unit 60 generates a third error covariance matrix that is an error covariance matrix of the generated wake. The third calculation unit 60 outputs the third error covariance matrix to the error covariance determination unit 70.

  The error covariance determination unit 70 compares the third error covariance matrix output from the third calculation unit 60 with the second error covariance matrix output from the second calculation unit 50, and detects the sensor 10. It is determined whether or not a system alignment error is included in the observation result obtained in step (b). That is, the error covariance determination unit 70 determines whether the observation result obtained by the sensor 10 includes an error with respect to the observation result obtained by the sensor 11 of the sensor device 101. The error covariance determination unit 70 outputs the determination result to the system alignment error estimation unit 30. Details of the operation of the error covariance determination unit 70 will be described later.

  Next, the configuration of the sensor device 101 will be described. The sensor device 101 includes a sensor 11, a first calculation unit 21, a system alignment error estimation unit 31, a transmission determination unit 41, and a second calculation unit 51. The sensor 11, the first calculation unit 21, the transmission determination unit 41, and the second calculation unit 51 are the sensor 10, the first calculation unit 20, the transmission determination unit 40, and the second calculation unit 50 of the sensor device 100. It is the same composition as. The system alignment error estimation unit 31 estimates the error amount included in the observation result using the first error covariance matrix. When the transmission determination unit 41 transmits the partial wake to the second calculation unit 51, the transmission determination unit 41 also transmits the partial wake to the second calculation unit 50 included in the sensor device 100 and the second calculation unit 52 included in the command control device 110. Send. An observation result obtained by the sensor 11 may be referred to as a second observation result.

  Next, the configuration of the command and control device 110 will be described. The command and control apparatus 110 includes a second calculation unit 52. The second calculation unit 52 has the same configuration as the second calculation unit 50 of the sensor device 100. The sensor system 1 may include other devices such as a fire control device instead of the command and control device 110.

  Next, the operation of the sensor device 100 will be described. FIG. 3 is a flowchart showing an operation of detecting a system alignment error in the sensor device 100 according to the present embodiment. The sensor 10 observes the target 80 (step S1). The sensor 10 outputs the observation result including the observation position of the target 80 to the first calculation unit 20 and the third calculation unit 60.

  The first computing unit 20 performs a smoothing process and a prediction process on the observation result output from the sensor 10, a partial track including a prediction result of the position, speed, and acceleration of the target 80 at a future time, and the first track An error covariance matrix is generated (step S2). The future time is a time that is a predetermined time after the current time. The smoothing process and the prediction process in the first calculation unit 20 are based on a general method. For example, the first arithmetic unit 20 may apply a process such as a Kalman filter process having low sensitivity to the motion of the target 80 or an M3 (Multiple Maneuver Model) filter process having high sensitivity to the motion of the target 80. it can. Here, the operation when the Kalman filter process is applied as an example of the smoothing process and the prediction process in the first arithmetic unit 20 will be described.

The observation value and the observation accuracy included in the observation result output from the sensor 10 are expressed as “z ^→ _k ” and “v ^→ _k ”, respectively, as vector quantities. The observation value includes the position, speed, acceleration, and the like of the target 80. In order to simplify the description, the position and speed of the target 80 are used here. Observed value vector z ^→ _k is a vector representing the observed value for the target 80 observed by the sensor 10 at the observation time t _k. Observation accuracy vector v ^→ _k is a vector representing the observation accuracy of the observation values with the target 80 observed by the sensor 10 at the observation time t _k. Here, the notation “z ^→ _k ” is an alternative notation of a character in which an arrow symbol “→” is arranged above “z”. Use the alternative notation in the text excluding the formula.

The observation vector z ^→ _k and the observation accuracy vector v ^→ _{k at the} observation time t _k can be expressed by the following equations (1) and (2), respectively. Here, the observed value vector z ^→ _k is a vector defined in a tracking reference coordinate system which is a reference orthogonal coordinate system. An example of the tracking reference coordinate system is the north reference orthogonal coordinate system. The observation accuracy vector v ^→ _k is a vector defined in a sensor reference polar coordinate system that is a polar coordinate system with the sensor 10 as a reference.

(1) In the formula, _{"x ok"} is x-axis component of the observed value at the measurement time _{t k. "Y ok"} is a y-axis component of the observed value at the observation time _{t k. "Z ok"} is a z-axis component of the observed value at the observation time _{t k.} The symbol “T” represents transposition. This notation is the same in mathematical formulas described later. In the equation (2), “σ _rk ” is a component in the distance direction of the observation accuracy at the observation time t _k . “Σ _ek ” is a component in the elevation direction of the observation accuracy at the observation time t _k . "Sigma _BYK" is a component of azimuthal measurement accuracy at the measurement time t _k. The component of the observation accuracy in the distance direction is called “distance accuracy”, the component of the observation accuracy in the elevation direction is called “elevation accuracy”, and the component of the observation accuracy in the azimuth direction is called “azimuth accuracy”. is there.

The observation error covariance matrix is represented by “R _k ”. The observation error covariance matrix R _k is a matrix that represents a range in which the true value of the target motion specification is considered to exist. The observation error covariance matrix R _k can be expressed by the following equation (3).

As shown in the equation (3), the observation error covariance matrix R _k is the distance accuracy σ _rk , elevation accuracy σ _ek, and azimuth accuracy σ _byk of the observation accuracy vector v ^→ _k shown in equation (2). a matrix with the diagonal elements component multiplication can be represented by using the coordinate transformation matrix G _k. The coordinate transformation matrix _Gk is a transformation matrix from the sensor reference polar coordinate system to the reference orthogonal coordinate system.

A vector of predicted values at the observation time t _k is represented by “X ^→ _{k | k−1} ”. The predicted value vector X ^→ _{k | k−1 at the} observation time t _k can be expressed by the following equation (4).

In the equation (4), “x _{k | k−1} ” is the x-axis component of the position predicted value at the observation time t _k . “Y _{k | k−1} ” is the y-axis component of the predicted position value at the observation time t _k . “Z _{k | k−1} ” is a z-axis component of the predicted position value at the observation time t _k . ^"X _{· k | k-1"} is the x-axis component of the velocity predicted value at the observation time _{t k.} ^"Y _{· k | k-1"} is a y-axis component of the velocity estimated value at the measurement time _{t k.} "Z ^· _{k | k-1"} is the z-axis component of the velocity estimated value at the measurement time t _k. Here, the notation “x ^· _{k | k−1} ” is an alternative notation of a character in which a dot symbol “·” is arranged above “x”. The same applies to “y” and “z”. Use the alternative notation in the text excluding the formula.

A vector of smooth values at the observation time t _k−1 is represented by “X ^→ _{k−1 | k−1} ”. Observation time _{t k-1,} when viewed from the observation time _{t k,} is the last of the observation time. In addition, the observation time t _k may be referred to as "this time of observation time". The smooth value vector X ^→ _{k−1 | k−1 at the} previous observation time t _k−1 can be expressed by the following equation (5).

In the equation (5), “x _{k−1 | k−1} ” is an x-axis component of the position smooth value at the previous observation time t _k−1 . “Y _{k−1 | k−1} ” is the y-axis component of the position smooth value at the previous observation time t _k−1 . “Z _{k−1 | k−1} ” is the z-axis component of the position smooth value at the previous observation time t _k−1 . “X ^· _{k−1 | k−1} ” is an x-axis component of the velocity smooth value at the previous observation time t _k−1 . “Y ^· _{k−1 | k−1} ” is the y-axis component of the velocity smooth value at the previous observation time t _k−1 . “Z ^· _{k−1 | k−1} ” is the z-axis component of the velocity smooth value at the previous observation time t _k−1 .

The relationship between the smoothed value vector X ^→ _{k−1 | k−1 at the} previous observation time t _k−1 and the predicted value vector X ^→ _{k | k−1} at the current observation time t _k is expressed by the following (6). It can be expressed by a formula.

In the equation (6), Φ (t _k , t _k−1 ) is a matrix representing a change in the target specification according to the elapsed time from the previous observation time t _k−1 to the current observation time t _k . Hereinafter, this matrix is referred to as a “state transition matrix”.

When the state transition matrix Φ (t _k , t _k−1 ) shown in the equation (6) is used, the prediction error covariance matrix P _{k | k−1} at the observation time t _k calculated by the first calculator 20 is used. Can be expressed using the following equation (7).

In equation (7), “P _{k−1 | k−1} ” is the smoothing error covariance matrix at the previous observation time t _k−1 , and “Q _k−1 ” is the previous observation time t _{k−. 2} is a system noise covariance matrix representing system noise at ₁ . Further, the Kalman gain matrix at the observation time t _k is represented by “K _k ”. The Kalman gain matrix K _k can be expressed by the following equation (8).

In the equation (8), “P _{k | k−1} ” is a prediction error covariance matrix at the observation time t _k represented by the equation (7), and “R _k ” is represented by the equation (3). it is the observation error covariance matrix in that observation time t _k. “H _k ” is a sensor observation matrix of positions at the observation time t _k . The sensor observation matrix H _k can be expressed by the following equation (9).

In the equation (9), “O _{3 × 3} ” is a zero matrix of 3 rows and 3 columns. The position sensor observation matrix H _k is a matrix for extracting the position components of the x axis, the y axis, and the z axis from the vector composed of the position and velocity components.

(8) Using the formulas and (9), the observation time _t smoothed value of _k vector X ^→ _{k | k,} and the observation time _t error covariance matrix _{P k} in _{k | k,} respectively, the following (10 ) And (11).

  Returning to the flowchart of FIG. The system alignment error estimator 30 obtains the determination result obtained by the error covariance determiner 70 by the process described later, specifically, the position, azimuth and time elements including the system alignment error, and the axis estimation result. As an initial value, an error amount of the system alignment error is estimated with respect to the partial track and the first error covariance matrix output from the first calculation unit 20 (step S3). Here, the observation result when the sensor 10 observes the target 80 is defined by a polar coordinate system based on the sensor 10, that is, a coordinate system having components of relative distance, azimuth angle, and elevation angle. However, the sensor 10 performs coordinate conversion from the polar coordinate system to the orthogonal coordinate system and outputs the observation result until it is output to each calculation unit. The system alignment error estimation unit 30 estimates how much system alignment error has occurred in each component of relative distance, azimuth angle, and elevation angle in the polar coordinate system based on the sensor 10. Then, the system alignment error estimation unit 30 converts the partial wake from the orthogonal coordinate system to the polar coordinate system, corrects the partial wake based on the error amount of the system alignment error estimated in the polar coordinate system, and the corrected partial wake is converted into the polar coordinate system. To the Cartesian coordinate system. The system alignment error estimating unit 30 converts the corrected partial track from the polar coordinate system to the orthogonal coordinate system as a corrected partial track to be output to the transmission determining unit 40. The system alignment error estimator 30 can correct the first error covariance matrix in the same manner as the partial wake. Further, the system alignment error estimation unit 30 uses the determination result obtained by the error covariance determination unit 70 as an initial value, thereby limiting the range of error candidates when estimating the error amount of the system alignment error. Thus, the processing load can be reduced, and the error convergence can be improved. The system alignment error estimation unit 30 outputs the corrected partial track and the corrected first error covariance matrix to the transmission determination unit 40.

  The transmission determination unit 40 acquires the partial wake and the first error covariance matrix from the system alignment error estimation unit 30, and acquires the wake information and the second error covariance matrix from the second calculation unit 50. The transmission determination unit 40 determines whether the accuracy of the track information generated by the second calculation unit 50 is improved by using the partial track (step S4). The transmission determination unit 40 uses the second calculation unit 50 when the second calculation unit 50 expects that the accuracy of the wake information is deteriorated to be equal to or less than the threshold value defined by the second calculation unit 50 or transmits the partial wake. If it is expected that the accuracy of will improve by more than a prescribed threshold, it is determined to transmit a partial track. The case where the accuracy of the track information is expected to deteriorate below a prescribed threshold in the second calculation unit 50 is, for example, the velocity component of the second error covariance matrix calculated by the second calculation unit 50 This is a case where the maximum eigenvalue of becomes greater than or equal to the threshold. In addition, the case where it is expected that the accuracy of the wake information is improved by more than a prescribed threshold in the second calculation unit 50 by transmitting the partial wake is the second calculation unit when the partial wake is not transmitted. 50 track accuracy, that is, the maximum eigenvalue of the velocity component of the second error covariance matrix calculated by the second calculation unit 50, and the accuracy of the track information of the second calculation unit 50 when the partial track is transmitted, That is, it is defined by comparing the first eigenvalue covariance matrix of the partial wake and the maximum eigenvalue of the velocity component of the smooth error covariance matrix calculated from the second error covariance matrix of the second calculation unit 50. This is a case where a difference equal to or greater than the threshold value has occurred. If the second calculation unit 50 determines that the accuracy of the wake information is improved by using the partial wake (step S4: Yes), the transmission determination unit 40 transmits the partial wake to the second calculation units 50 to 52. (Step S5). If the second calculation unit 50 determines that the accuracy of the wake information is not improved even if the partial wake is used (step S4: No), the transmission determination unit 40 sends the partial wake to the second calculation units 50 to 52. The process ends without outputting.

  When the second arithmetic unit 50 receives a partial wake from at least one of the transmission determination units 40 and 41, the second arithmetic unit 50 performs smoothing processing and prediction processing on the partial wake to generate or update the wake information. In addition, the second calculation unit 50 generates, that is, updates, the second error covariance matrix of the wake information (step S6). The smoothing process and the prediction process in the second calculation unit 50 are the same as the smoothing process and the prediction process in the first calculation unit 20. Thereby, in the sensor system 1, in each 2nd calculating parts 50-52, a partial wake can be shared and the equivalency of wake information can be ensured. When the second calculation unit 50 acquires the partial wake from the transmission determination unit 40, the second calculation unit 50 notifies the first calculation unit 20 that the partial wake has been acquired from the transmission determination unit 40. When the first calculation unit 20 receives a notification from the second calculation unit 50 that the partial wake has been acquired, the first calculation unit 20 performs a reset process. In this case, the first calculation unit 20 generates a partial track and a first error covariance matrix from the time after the reset process. In addition, the second calculation unit 50 outputs the updated track information and the second error covariance matrix to the transmission determination unit 40. In addition, the second calculation unit 50 outputs the updated second error covariance matrix to the error covariance determination unit 70.

  Similar to the first calculation unit 20, the third calculation unit 60 performs smoothing processing and prediction processing on the observation result output from the sensor 10, and predicts the position, speed, and acceleration of the target 80 at the future time. Then, a third error covariance matrix is generated (step S7). The process of generating the third error covariance matrix by the third calculation unit 60 is the same as the process of the first calculation unit 20 described above. In addition, unlike the 1st calculating part 20, the 3rd calculating part 60 is while the 2nd calculating part 50 is maintaining track information, even when the 2nd calculating part 50 acquires a partial wake. Reset processing is not performed. About the process of step S7, you may perform in parallel with the process of step S2 to step S6. When the first error covariance matrix, the second error covariance matrix, and the third error covariance matrix are not distinguished, they may be referred to as an error covariance matrix.

  The error covariance determination unit 70 compares the third error covariance matrix output from the third calculation unit 60 with the second error covariance matrix output from the second calculation unit 50, and the observation result It is determined whether or not a system alignment error is included (step S8). The error covariance determination unit 70 outputs the determination result to the system alignment error estimation unit 30 (step S9).

  Here, the error covariance matrix will be described. FIG. 4 is a diagram illustrating an example of an error covariance matrix generated by the sensor device 100 according to the present embodiment. The error covariance matrix shown in FIG. 4 is a matrix that shows the correlation between the error amounts related to position and velocity. In FIG. 4, an error component at a location indicated by a frame A is a component indicating an error amount of a component related to the position. Further, the error components at the locations indicated by the frames B and D are components indicating the error amount of the correlation component between the position and the speed. Further, the error component at the location indicated by the frame C is a component indicating the error amount of the component relating to the speed. As shown in FIG. 4, the components of the position and velocity error components are represented by an x component, a y component, and a z component. In the error covariance matrix, the element in which the error is generated and the magnitude of the error can be indicated by the size of each component shown in FIG. In the error covariance matrix, each component is represented for each axis of the coordinate system that specifies the position of the target 80, specifically, an x component, a y component, and a z component.

  FIG. 5 is a flowchart showing processing for determining whether or not a system alignment error is included in the error covariance determination unit 70 according to the present exemplary embodiment. The error covariance determination unit 70 includes each component of the third error covariance matrix generated by the third calculation unit 60 and each component of the second error covariance matrix generated by the second calculation unit 50. Are compared to determine whether or not there is a difference between specific components of the error covariance matrix (step S21). If the error covariance determination unit 70 determines that there is no difference in specific components of the error covariance matrix (step S21: No), it determines that there is no system alignment error (step S22), and ends the process. .

  When the error covariance determination unit 70 determines that a difference occurs in a specific component of the error covariance matrix (step S21: Yes), the error covariance matrix characteristic amount is correlated with a plurality of targets. analyse. The error covariance determination unit 70 generates a difference between the error covariance matrix for the target 80 that is the detection target of the sensor device 100 and the error covariance matrix for a different target that is different from the target 80. It is determined whether or not the component is the same. That is, the error covariance determination unit 70 determines whether or not a different target has the same component as the target 80 (step S23).

  The error covariance determination unit 70 determines that a system alignment error related to the position has occurred (step S24) and determines that a difference has occurred in the same component even in different targets (step S23: Yes). Exit. When the error covariance determination unit 70 determines that there is no difference in the same component in different targets (step S23: No), the component in which the difference is generated in the error covariance matrix from the positional relationship between the sensor 10 and each target. Calculate the orientation specified by. The error covariance determination unit 70 determines whether or not the component having the difference in the different target has a correlation in the azimuth (step S25). In the example of FIG. 4, the error covariance determination unit 70 uses a component related to the position indicated by the frame A in the error covariance matrix.

  The error covariance determination unit 70 determines that a system alignment error relating to the azimuth has occurred (step S26) when it is determined that the component causing the difference in the different target is correlated with the azimuth (step S25: Yes). ), The process is terminated. When the error covariance determination unit 70 determines that the component having the difference in the different target has no correlation in the azimuth (step S25: No), The feature quantity of the error covariance matrix for the same target 80 generated in the past is compared. The error covariance determination unit 70 determines whether or not there is a correlation between the time-series change of the component having a difference in the error covariance matrix and the target traveling direction in the same target 80 (step S27). In the example of FIG. 4, the error covariance determination unit 70 uses a component related to the speed indicated by the frame C in the error covariance matrix.

  When the error covariance determination unit 70 determines that there is a correlation between the time-series change of the component causing the difference in the error covariance matrix and the target traveling direction (step S27: Yes), the system alignment error regarding the time Is determined (step S28), and the process is terminated. When the error covariance determination unit 70 determines that there is no correlation between the time series change of the component causing the difference in the error covariance matrix and the target traveling direction (step S27: No), a system alignment error occurs. However, it is difficult to specify the generation factor in a state where a plurality of factors are generated in combination, and it is determined that the generation factor is unknown (step S29), and the process ends.

  The error covariance determination unit 70 determines which element of the position, orientation, or time is the cause of occurrence of the system alignment error. The axis where the system alignment error has occurred in the variance matrix is output to the system alignment error estimation unit 30 as a determination result.

  As described above, the error covariance determination unit 70 determines the position, orientation, and time in the observation result based on the feature amounts obtained from the corresponding components of the third error covariance matrix and the second error covariance matrix. It is determined whether or not each element includes an error. The error covariance determination unit 70 estimates the component axis of the element including the error for the element determined to include the error. The error covariance determination unit 70 outputs the element determined to contain an error and the estimated axis to the system alignment error estimation unit 30 as a determination result. The system alignment error estimation unit 30 uses the determination result acquired from the error covariance determination unit 70 as an initial value when estimating the system alignment error, and estimates the error amount of the system alignment error. Thereby, the system alignment error estimation part 30 can reduce the processing load at the time of system alignment error estimation, and can improve the convergence of an error.

  Next, the hardware configuration of the sensor device 100 will be described. In the sensor device 100, the sensor 10 is a measuring instrument. The function of outputting the partial wake to the sensor device 101 and the command and control device 110 in the transmission determination unit 40 and the function of acquiring the partial wake from the sensor device 101 in the second calculation unit 50 are realized by an interface circuit. The first calculation unit 20, the system alignment error estimation unit 30, other functions of the transmission determination unit 40, the second calculation unit 50, the third calculation unit 60, and the error covariance determination unit 70 are realized by a processing circuit. Is done. The processing circuit may be a processor and a memory that execute a program stored in the memory, or may be dedicated hardware.

  FIG. 6 is a diagram illustrating an example in which the processing circuit included in the sensor device 100 according to the present embodiment is configured with a processor and a memory. When the processing circuit includes the processor 91 and the memory 92, each function of the processing circuit of the sensor device 100 is realized by software, firmware, or a combination of software and firmware. Software or firmware is described as a program and stored in the memory 92. In the processing circuit, each function is realized by the processor 91 reading and executing the program stored in the memory 92. That is, the processing circuit includes a memory 92 for storing a program in which processing of each component of the sensor device 100 is executed as a result. These programs can also be said to cause a computer to execute the procedures and methods of each component of the sensor device 100. In FIG. 6, a measuring instrument 94 is the aforementioned measuring instrument. The interface circuit 95 is the above-described interface circuit.

  Here, the processor 91 may be a CPU (Central Processing Unit), a processing device, an arithmetic device, a microprocessor, a microcomputer, a DSP (Digital Signal Processor), or the like. The memory 92 is nonvolatile or volatile, such as RAM (Random Access Memory), ROM (Read Only Memory), flash memory, EPROM (Erasable Programmable ROM), EEPROM (registered trademark) (Electrically EPROM), and the like. Such semiconductor memory, magnetic disk, flexible disk, optical disk, compact disk, mini disk, DVD (Digital Versatile Disc), and the like are applicable.

  FIG. 7 is a diagram illustrating an example of a case where the processing circuit included in the sensor device 100 according to the first embodiment is configured with dedicated hardware. When the processing circuit is configured by dedicated hardware, the processing circuit 93 shown in FIG. 7 includes, for example, a single circuit, a composite circuit, a programmed processor, a parallel programmed processor, an ASIC (Application Specific Integrated Circuit), An FPGA (Field Programmable Gate Array) or a combination thereof is applicable. Each function of each component of the sensor device 100 may be realized by the processing circuit 93 according to function, or each function may be realized by the processing circuit 93 collectively.

  In addition, about each function of each component of the sensor apparatus 100, a part may be implement | achieved by exclusive hardware and a part may be implement | achieved by software or firmware. As described above, the processing circuit can realize the above-described functions by dedicated hardware, software, firmware, or a combination thereof.

  As described above, according to the present embodiment, in the sensor system 1, the sensor device 100 generates the partial wake and the first error covariance matrix based on the observation result obtained by the sensor 10, and the partial A wake information and a second error covariance matrix are generated based on the wake and the partial wake received from another sensor device 101, and a third error covariance matrix is generated after the observation is started based on the observation result. To do. The sensor device 100 compares the third error covariance matrix and the second error covariance matrix to determine whether or not a system alignment error is included in the observation result, and uses the determination result to perform system alignment. The error was estimated. Thereby, the sensor device 100 can detect an error occurring between the plurality of sensor devices while suppressing an increase in the scale of the device.

  In the present embodiment, the configuration in which the sensor system 1 includes two sensor devices and the sensor device 100 includes the system alignment error detection device 120 out of the two sensor devices has been described. However, the present invention is not limited to this. The sensor system 1 may include three or more sensor devices. In the sensor system 1, among the plurality of sensor devices, the number of sensor devices including the system alignment error detection device 120 may be one or plural.

  The configuration described in the above embodiment shows an example of the contents of the present invention, and can be combined with another known technique, and can be combined with other configurations without departing from the gist of the present invention. It is also possible to omit or change the part.

  DESCRIPTION OF SYMBOLS 1 Sensor system 10,11 Sensor, 20,21 1st calculating part, 30, 31 System alignment error estimation part, 40, 41 Transmission determination part, 50, 51, 52 2nd calculating part, 60 3rd calculation Unit, 70 error covariance determination unit, 80 target, 100, 101 sensor device, 110 command and control device, 120 system alignment error detection device.

Claims (9)

A sensor device that constitutes a sensor system together with other sensor devices,
A first calculation that generates a partial wake that is a wake of the target in a first period based on a first observation result of a sensor that observes the target, and generates a first error covariance matrix of the partial wake And
A transmission determination unit for determining whether to transmit the partial wake;
Based on the partial wake received from the transmission determination unit and the partial wake received from the other sensor device, wake information that is the target wake after the start of observation is generated, and the wake information A second arithmetic unit for generating an error covariance matrix of 2;
A third arithmetic unit that generates a third error covariance matrix of the target track from the start of observation based on the first observation result;
The third error covariance matrix is compared with the second error covariance matrix, and whether or not the first observation result includes an error with respect to the second observation result obtained by the other sensor device. An error covariance determination unit for determining whether or not
An error estimation unit that estimates an error amount included in the first observation result based on the first error covariance matrix using the determination result of the error covariance determination unit;
A sensor device comprising: The components representing the third error covariance matrix and the second error covariance matrix include position, azimuth, and time elements, and each component is represented for each axis of the coordinate system that specifies the target position. If
The error covariance determination unit, based on feature amounts obtained from corresponding components of the third error covariance matrix and the second error covariance matrix, in the first observation result, And whether each element of time and time includes an error, estimates the component axis of the element including the error for the element determined to include the error, and the element determined to include the error, and the estimation Output the determined axis to the error estimation unit as the determination result,
The error estimation unit estimates the error amount using an element and an axis acquired as the determination result from the error covariance determination unit as initial values.
The sensor device according to claim 1. In the first period, the transmission determination unit transmits the partial wake to the second calculation unit, and the first calculation unit notifies that the partial wake has been received from the second calculation unit. It is a period after receiving and resetting, or a period after the sensor starts observation,
The sensor device according to claim 1, wherein: The transmission determination unit includes the partial wake and the first error covariance matrix acquired from the first calculation unit via the error estimation unit, the wake information acquired from the second calculation unit, and the first wake information. When it is determined that the accuracy of the wake information is improved by using the partial wake based on the error covariance matrix of 2, the partial wake is transmitted to the second arithmetic unit and the other sensor device.
The sensor device according to any one of claims 1 to 3, wherein An error detection method for a sensor device that constitutes a sensor system together with another sensor device,
A first calculation unit generates a partial wake that is a wake of the target in a first period based on a first observation result of a sensor that observes the target, and a first error covariance matrix of the partial wake A first step of generating
A second step of determining whether or not the transmission determination unit transmits the partial wake;
Based on the partial wake received from the transmission determination unit and the partial wake received from the other sensor device, a second calculation unit generates wake information that is the target wake after the start of observation. A third step of generating a second error covariance matrix of the wake information;
A fourth step in which a third calculation unit generates a third error covariance matrix of the target track after starting observation based on the first observation result;
An error covariance determination unit compares the third error covariance matrix with the second error covariance matrix, and performs the first observation on the second observation result obtained by the other sensor device. A fifth step of determining whether the result includes an error;
A sixth step in which an error estimation unit estimates an error amount included in the first observation result based on the first error covariance matrix using the determination result of the error covariance determination unit;
An error detection method comprising: The components representing the third error covariance matrix and the second error covariance matrix include position, azimuth, and time elements, and each component is represented for each axis of the coordinate system that specifies the target position. If
In the fifth step, the error covariance determination unit determines the first coherence based on the feature amount obtained from the corresponding components of the third error covariance matrix and the second error covariance matrix. In the observation results, determine whether each element of position, orientation, and time includes an error, estimate the component axis of the element that includes the error for the element determined to include the error, and include the error Output the determined element and the estimated axis to the error estimation unit as the determination result,
In the sixth step, the error estimation unit estimates the error amount using an element and an axis acquired as the determination result from the error covariance determination unit as initial values.
The error detection method according to claim 5. In the first period, the transmission determination unit transmits the partial wake to the second calculation unit, and the first calculation unit notifies that the partial wake has been received from the second calculation unit. It is a period after receiving and resetting, or a period after the sensor starts observation,
The error detection method according to claim 5 or 6, characterized in that: In the second step, the transmission determination unit includes the partial wake and the first error covariance matrix acquired from the first calculation unit via the error estimation unit, and the second calculation unit. When it is determined that the accuracy of the wake information is improved by using the partial wake based on the acquired wake information and the second error covariance matrix, the partial wake is converted to the second arithmetic unit and the other To the sensor device
The error detection method according to claim 5, wherein the error detection method is an error detection method.   An error detection program for causing a computer to execute the error detection method according to any one of claims 5 to 8.

Images