JP2021032633A - Target presence likelihood calculation device and target presence likelihood calculation method - Google Patents

Abstract

To provide a target presence likelihood calculation device and a target presence likelihood calculation method, which can properly calculate a target presence likelihood even when the uncertainty of predictive receive level does not conform to a normal distribution.SOLUTION: Provided is a target presence likelihood calculation device for calculating a target presence likelihood that indicates the likelihood of existing position of a target, comprising: an input unit to which are inputted a plurality of pieces of environment information indicating an ambient environment state and corresponding to uncertainty, vessel information indicating the state of the host vessel, a measured receive level from a target, a sound source level and frequency information; a propagation loss calculator for calculating a plurality of propagation losses corresponding to uncertainty on the basis of the plurality of kinds of environment information, vessel information and frequency information; a predictive receive level calculator for calculating a plurality of predictive receive levels corresponding to uncertainty that indicates the predictive value of a signal radiated from a target, on the basis of the sound source level and the plurality of propagation losses; and a target likelihood calculator for calculating a target presence likelihood on the basis of the measured receive level and the plurality of predictive receive levels.SELECTED DRAWING: Figure 1

Description

The present invention relates to a target existence likelihood calculation device and a target existence likelihood calculation method for calculating the likelihood with respect to an approximate distance and an approximate depth of a target object.

Conventionally, as a sonar device mounted on a ship or the like, a passive sonar that detects the presence of a target by receiving sound waves radiated from the target is known. For example, Patent Document 1 discloses a target existence likelihood calculation device and a target existence likelihood calculation method for quickly determining the distance to the target and the depth of the target by using the information about the signal received from the target. Has been done.

The target existence likelihood calculation device described in Patent Document 1 is a sound wave based on environmental information indicating the environmental state around the own ship, own ship information indicating the state of the own ship, and frequency information of a signal radiated from the target. Using the propagation model, the propagation loss (distance characteristic) at a preset depth is calculated. Further, the target existence likelihood calculation device calculates a predicted reception level indicating a predicted value of the reception level of the signal radiated from the target based on the sound source level of the signal radiated from the target and the calculated propagation loss. To do. Then, the target existence likelihood calculation device calculates the target existence likelihood (distance characteristic) based on the actually measured reception level of the signal radiated from the target object and the calculated predicted reception level. The target existence likelihood in this case is the likelihood (likelihood) that the measured reception level will be the input measured reception level at an arbitrary distance, assuming that the uncertainty of the predicted reception level follows a normal distribution. Represents.

Japanese Unexamined Patent Publication No. 2018-36112

However, in general, the uncertainty of the predicted reception level does not always follow a normal distribution. In particular, when uncertainty occurs in the propagation loss calculated based on the environmental information, own ship information, and frequency information input to the target existence likelihood calculation device described in Patent Document 1, the calculated propagation loss is normal. It does not always follow the distribution. If the propagation loss does not follow the normal distribution, the uncertainty of the sound source level input to the target existence likelihood calculation device and the predicted reception level calculated based on the propagation loss does not follow the normal distribution. There is a problem that an error occurs in the target existence likelihood to be achieved.

Therefore, there is a demand for a target existence likelihood calculation device and a target existence likelihood calculation method that can appropriately calculate the target existence likelihood even when the uncertainty of the predicted reception level does not follow the normal distribution.

The target existence probability calculation device according to the present invention is mounted on a sonar device that estimates the distance to the target, and calculates the target existence probability that indicates the likelihood of the position where the target exists. A plurality of environmental information according to uncertainties indicating the environmental state around the sonar device, own ship information indicating the state of the ship on which the sonar device is mounted, and signals radiated from the target. Uncertainty based on the actual measurement reception level, the sound source level obtained from the target, the input unit into which the frequency information obtained from the target is input, the plurality of environmental information, the own ship information, and the frequency information. A propagation loss calculator that calculates a plurality of propagation losses according to the target, and a plurality of uncertainties that indicate predicted values of signals emitted from the target based on the sound source level and the calculated plurality of propagation losses. A target existence probability calculator that calculates the target existence probability based on the predicted reception level calculator that calculates the predicted reception level of the target, the actual measurement reception level from the target object, and the calculated plurality of predicted reception levels. It is equipped with.

The target existence likelihood calculation method according to the present invention is mounted on a sonar device that estimates the distance to the target, and calculates the target existence likelihood that indicates the likelihood of the position where the target exists. It is a method, that is, a plurality of environmental information according to uncertainties indicating the environmental state around the sonar device, own ship information indicating the state of the ship on which the sonar device is mounted, and a frequency obtained from the target. Based on the information, a plurality of propagation losses according to the uncertainty are calculated, and the signal emitted from the target is predicted based on the sound source level obtained from the target and the calculated plurality of propagation losses. A plurality of predicted reception levels indicating values are calculated according to the uncertainty, and the target existence probability is based on the actually measured reception level of the signal radiated from the target and the calculated plurality of predicted reception levels. Is calculated.

As described above, according to the present invention, the target existence likelihood is calculated based on the plurality of predicted reception levels according to the uncertainty, which are calculated by using the plurality of environmental information according to the uncertainty. As a result, the target existence likelihood can be appropriately calculated even when the uncertainty of the predicted reception level does not follow the normal distribution.

It is a block diagram which shows an example of the structure of the target existence likelihood calculation apparatus which concerns on Embodiment 1. FIG. It is a hardware block diagram which shows an example of the structure of the target existence likelihood calculation apparatus of FIG. It is a hardware block diagram which shows another example of the structure of the control device of FIG. It is a block diagram which shows an example of the structure of the target existence likelihood calculation apparatus which concerns on Embodiment 2. FIG. It is a block diagram which shows an example of the structure of the target existence likelihood calculation apparatus which concerns on Embodiment 3. FIG.

Hereinafter, embodiments of the present invention will be described with reference to the drawings. The present invention is not limited to the following embodiments, and can be variously modified without departing from the gist of the present invention. In addition, the present invention includes all combinations of configurations that can be combined among the configurations shown in the following embodiments. Further, in each figure, those having the same reference numerals are the same or equivalent thereof, which are common in the entire text of the specification.

Embodiment 1.
The target existence likelihood calculation device according to the first embodiment will be described. The target existence likelihood calculation device according to the embodiment of the present invention is provided in, for example, a sonar device mounted on a ship for estimating the distance and depth to a target. The target existence probability calculation device is based on the surrounding environmental conditions, the state of the own ship on which the sonar device is mounted, and information on the signal received from the target, from the own ship to the target existing on the ocean or in the ocean. The target existence likelihood, which indicates the likelihood of the position where the target exists, such as the distance of the target and the depth of the target, is calculated.

In this target existence probability calculation device, as information about the signal received from the target, the actual measurement reception level of the signal radiated from the target, the sound source level such as the radiation noise obtained from the target, and the radiated from the target. The frequency information of the signal is used.

[Configuration of target existence likelihood calculation device]
FIG. 1 is a block diagram showing an example of the configuration of the target existence likelihood calculation device according to the first embodiment. As shown in FIG. 1, the target existence likelihood calculation device 10 includes a propagation loss calculator 11, a predicted reception level calculator 12, a target existence likelihood calculator 13, input terminals 1a to 1e as input units, and outputs. It has a terminal 2a. The target existence likelihood calculation device 10 is composed of an arithmetic unit such as a microcomputer that realizes various functions by executing software, or hardware such as a circuit device corresponding to various functions.

The input terminal 1a has a BT (BARHY-THERMOGRAPHY) profile, which is characteristic data showing the relationship between the water temperature and the water depth, and an environment showing the environmental conditions around the ship such as the seafloor topography, bottom sediment, wind speed, or sea state of the target direction with respect to the ship. Information is entered. In the first embodiment, N pieces of environmental information that fluctuate according to the uncertainty of the environmental information are input to the input terminal 1a.

Information indicating the state of the ship, such as sensor specifications such as the depth of the ship or the beam pattern, is input to the input terminal 1b. A sound source level SL such as radiation noise obtained from a target is input to the input terminal 1c. The actual measurement reception level RL of the signal radiated from the target is input to the input terminal 1d. The frequency information of the signal radiated from the target is input to the input terminal 1e.

In the propagation loss calculator 11, the input terminals 1a, 1b and 1e are connected to the input side, and the predicted reception level calculator 12 is connected to the output side. The propagation loss calculator 11 is based on the environmental information input to the input terminal 1a, the own ship information input to the input terminal 1b, and the frequency information of the signal radiated from the target input to the input terminal 1e. ^{The distance characteristics TL (n)} (r) (n = 1, 2, ..., N) of N propagation losses are calculated using the sound wave propagation model. Here, r indicates the distance r from the own ship on which the target existence likelihood calculation device 10 is mounted. The propagation loss calculator 11 outputs the calculated propagation loss TL ⁽ⁿ⁾ (r) to the predicted reception level calculator 12.

In the predicted reception level calculator 12, the input terminal 1c and the propagation loss calculator 11 are connected to the input side, and the target presence likelihood calculator 13 is connected to the output side. The predicted reception level calculator 12 is radiated from the target based on the sound source level SL input to the input terminal 1c and the N propagation losses TL ^{(n) (r) input from the propagation loss calculator 11.} _{N predictive reception levels RL p} ⁽ⁿ⁾ (r) (n = 1, 2, ..., N) indicating the predicted values of the signals are calculated. The predicted reception level calculator 12 outputs the calculated predicted reception levels RL _p ⁽ⁿ⁾ (r) to the target existence likelihood calculator 13.

The target existence likelihood calculator 13 has an input terminal 1d and a predicted reception level calculator 12 connected to the input side, and an output terminal 2a connected to the output side. The target existence likelihood calculator 13 includes an actually measured reception level RL of the signal radiated from the target object input to the input terminal 1d and N predicted reception level RL _p ^{(n) input from the predicted reception level calculator 12. )} Based on (r), the target existence likelihood p _RL (r | RL) is calculated. The target existence likelihood calculator 13 outputs the calculated target existence likelihood p _RL (r | RL) from the output terminal 2a.

_{The target existence likelihood p RL} (r | RL) output from the output terminal 2a is displayed by, for example, a display unit (not shown) provided as a part of the output unit. _{The display unit visually displays the target existence likelihood pRL} (r | RL) calculated by the target existence likelihood calculator 13 by a graph, a map, or the like. The display of the target existence likelihood p _RL (r | RL) is not limited to this example, and may be performed by, for example, a display unit provided in the sonar device equipped with the target existence likelihood calculation device 10. ..

FIG. 2 is a hardware configuration diagram showing an example of the configuration of the target existence likelihood calculation device of FIG. When various functions of the target existence likelihood calculation device 10 are executed by hardware, the target existence likelihood calculation device 10 of FIG. 1 is composed of a processing circuit 41 as shown in FIG. In the target existence likelihood calculation device 10 of FIG. 1, each function of the propagation loss calculator 11, the predicted reception level calculator 12, and the target existence likelihood calculator 13 is realized by the processing circuit 41.

When each function is executed by hardware, the processing circuit 41 may be, for example, a single circuit, a composite circuit, a programmed processor, a parallel programmed processor, an ASIC (Application Specific Integrated Circuit), or an FPGA (Field-Programmable Gate). Array), or a combination of these. The target existence likelihood calculation device 10 may realize the functions of each part of the propagation loss calculator 11, the predicted reception level calculator 12, and the target existence likelihood calculation device 13 by the processing circuit 41, or the functions of each part may be realized. It may be realized by one processing circuit 41.

FIG. 3 is a hardware configuration diagram showing another example of the configuration of the control device of FIG. When various functions of the target existence likelihood calculation device 10 are executed by software, the target existence likelihood calculation device 10 of FIG. 1 includes a processor 51 and a memory 52 as shown in FIG. In the target existence likelihood calculation device 10, each function of the propagation loss calculator 11, the predicted reception level calculator 12, and the target existence likelihood calculator 13 is realized by the processor 51 and the memory 52.

When each function is executed by software, in the target existence likelihood calculation device 10, the functions of the propagation loss calculator 11, the predicted reception level calculator 12, and the target existence likelihood calculator 13 are software, firmware, or software. It is realized by combining with firmware. The software and firmware are written as a program and stored in the memory 52. The processor 51 realizes the functions of each part by reading and executing the program stored in the memory 52.

Examples of the memory 52 include non-volatile or volatile semiconductor memories such as RAM (Random Access Memory), ROM (Read Only Memory), flash memory, EPROM (Erasable and Programmable ROM) and EEPROM (Electrically Erasable and Programmable ROM). Is used. Further, as the memory 52, for example, a removable recording medium such as a magnetic disk, a flexible disk, an optical disk, a CD (Compact Disc), an MD (Mini Disc), or a DVD (Digital Versatile Disc) may be used.

As described above, the target existence likelihood calculation device 10 can realize each of the above-mentioned functions by hardware, software, firmware, or a combination thereof.

[Calculation method of target existence likelihood]
Next, a method of calculating the target existence likelihood in the target existence likelihood calculation device 10 according to the first embodiment will be described. When environmental information, own ship information, sound source level SL, and actual measurement reception level RL from the target object are input to the target existence likelihood calculation device 10, the target existence likelihood calculation device 10 inputs the input information. It is used to calculate the target existence likelihood p _RL (r | RL).

(Calculation of propagation loss)
First, the target existence likelihood calculation device 10 calculates the distance characteristics TL ⁽ⁿ⁾ (r) of the propagation loss by the propagation loss calculator 11. The propagation loss calculator 11 is based on the N environmental information input to the input terminal 1a, the own ship information input to the input terminal 1b, and the frequency information of the signal from the target input to the input terminal 1e. , using a wave propagation model, the distance properties of the propagation loss in the preset target depth z _t (hereinafter, appropriately referred to as a "transmission loss") is calculated TL ^{(n) (r).}

The sound wave propagation model used at this time may be any one that can reflect the difference of N environmental information in the propagation loss. Specifically, as a sound wave propagation model, for example, the document "Michael D. Collins," A split-step Pade solution for the parabolic equation method "J. Acoust. Soc. Am. 93 (4), Pt. 1, 1993. A sound wave propagation model based on the wave theory as described in "." May be used. Not limited to this, as a sound wave propagation model, for example, the literature "Henry Weinberg and Ruth Eta Keenan," Gaussian ray bundles for modeling high-frequency propagation loss under shallow-water conditions, "J. Acoust. Soc. Am. 100 (3), 1996. ”, a sound wave propagation model based on sound ray theory may be used.

(Calculation of predicted reception level)
Next, the target existence likelihood calculation device 10 calculates the predicted reception level RL _p ⁽ⁿ⁾ (r) by the predicted reception level calculator 12. The predicted reception level calculator 12 uses the equation (1) based on the sound source level SL input to the input terminal 1c and the propagation loss TL ^{(n) (r) input from the propagation loss calculator 11 to N.} Calculate the predicted reception levels RL _p ⁽ⁿ⁾ (r).

(Calculation of target existence likelihood)
Next, the target existence likelihood calculation device 10 calculates the distance characteristic p _RL (r | RL) of the target existence likelihood by the target existence likelihood calculator 13. First, the target existence likelihood calculator 13 extracts the minimum value RL _min and the maximum value RL _{max from the N predicted reception levels RL p} ⁽ⁿ⁾ (r), and the extracted minimum value RL _min and the maximum value. The _{section including RL max} is divided into M ranges [RL _m-1 , RL _m ] (m = 1, 2, ..., M). In this case, the value RL _m is calculated based on the equation (2). Further, in the equation (2), the value ΔRL is calculated based on the equation (3).

Then, the target existence likelihood calculator 13 determines the number of _{predicted reception levels RL p} ⁽ⁿ⁾ (r) that satisfy the equation (4) for each of m (= 1, 2, ..., M). Count and let this number be the value P _m (m = 1, 2, ..., M). The value P _{m at} this time indicates the frequency distribution of the predicted reception level RL _p ^{(n) (r).}

Next, the target existence likelihood calculator 13 searches for m such that the actually measured reception level RL satisfies the equation (5). Then, the target existence likelihood calculator 13 uses the number P _m for m obtained by the search as the value P (RL). If m that satisfies the equation (5) does not exist, the target existence likelihood calculator 13 sets “P (RL) = 0”.

Then, the target existence likelihood calculator 13 sets the number P _m (m = 1, 2, ..., M) _{of the obtained predicted reception levels RL p} ^{(n) (r) and the value P (RL).} Based on this, the distance characteristic of the target existence likelihood (hereinafter, appropriately referred to as “target existence likelihood”) p _RL (r | RL) is calculated using the equation (6). Here, the formula (7) showing the value of the denominator in the formula (6) shows the maximum value in the _{number P m.}

In the formula (6), the value of the denominator represented by the formula (7) indicates the frequency with respect to the level most likely to be received at the distance r from the own ship equipped with the target existence likelihood calculation device 10. Further, the molecular value P (RL) in the formula (6) indicates the frequency with respect to the level actually received.

As described above, in the first embodiment, the _{normal distribution is not assumed as the uncertainty distribution of the predicted reception level RL p} ⁽ⁿ⁾ (r), and the uncertainty based on the result when the propagation loss is calculated is not assumed. It is based on the distribution of.
That is, when the target existence likelihood calculation method according to the first embodiment is used, the target existence likelihood can be appropriately calculated even when the propagation loss does not follow the normal distribution.

As described above, in the first embodiment, the target existence likelihood is calculated based _{on the N predicted reception levels RL p} ^{(n) (r) for the N environmental information according to the uncertainty.} Further, at this time, the frequency distribution P _{m of} the plurality of predicted reception levels RL _p ⁽ⁿ⁾ (r) is calculated, and the target existence likelihood is calculated from the calculated _{frequency distribution P m.} As a result, the target existence likelihood can be appropriately calculated even when the uncertainty of the predicted reception level does not follow the normal distribution.

Embodiment 2.
Next, the second embodiment will be described. The second embodiment is different from the first embodiment in that the target existence likelihood is calculated based on the probability density function of the predicted reception level represented by the mixed normal distribution. In the following description, the parts common to the first embodiment are designated by the same reference numerals, and detailed description thereof will be omitted.

[Configuration of target existence likelihood calculation device]
FIG. 4 is a block diagram showing an example of the configuration of the target existence likelihood calculation device according to the second embodiment. As shown in FIG. 4, the target existence likelihood calculation device 20 includes a propagation loss calculator 11, a predicted reception level calculator 12, a target existence likelihood calculator 23, input terminals 1a to 1e as input units, and outputs. It has a terminal 2a. The target existence likelihood calculation device 20 is composed of an arithmetic unit such as a microcomputer that realizes various functions by executing software, or hardware such as a circuit device corresponding to various functions.

The target existence likelihood calculator 23 has an input terminal 1d and a predicted reception level calculator 12 connected to the input side, and an output terminal 2a connected to the output side. The target existence likelihood calculator 23 includes an actually measured reception level RL of the signal radiated from the target object input to the input terminal 1d and N predicted reception level RL _p ^{(n) input from the predicted reception level calculator 12. )} Based on (r), the target existence likelihood p _RL (r | RL) is calculated. The target existence likelihood calculator 13 outputs the calculated target existence likelihood p _RL (r | RL) from the output terminal 2a.

_{The target existence likelihood p RL} (r | RL) output from the output terminal 2b is visually displayed by a graph, a map, or the like on a display unit (not shown) provided as a part of the output unit, for example. The display of the target existence likelihood p _RL (r | RL) is not limited to this example, and may be performed by, for example, a display unit provided in the sonar device equipped with the target existence likelihood calculation device 20. ..

[Calculation method of target existence likelihood]
Next, a method of calculating the target existence likelihood in the target existence likelihood calculation device 20 according to the second embodiment will be described. When environmental information, own ship information, sound source level SL, and actual measurement reception level RL from the target object are input to the target existence likelihood calculation device 20, the target existence likelihood calculation device 20 inputs the input information. It is used to calculate the target existence likelihood p _RL (r | RL).

In the second embodiment, the ^{method of calculating the propagation loss TL (n)} (r) by the propagation loss calculator 11 and the calculation of the predicted reception level RL _p ⁽ⁿ⁾ (r) by the predicted reception level calculator 12 The method is the same as in the first embodiment. Therefore, here, a method of calculating the target existence likelihood by the target existence likelihood calculator 23 will be described.

(Calculation of target existence likelihood)
In the second embodiment, the target existence likelihood calculator 23 expresses the probability density function f _RL ⁽ⁿ⁾ _{(x) representing the distribution of N predicted reception levels RL p} ⁽ⁿ⁾ (r) by the equation (8). ) Is represented by the mixed normal distribution. The mixed normal distribution shows the sum of multiple normal distributions, and the probability density function of a mixed normal distribution shows the sum of the probability density functions of multiple normal distributions. The function f (x, μ, σ) in the equation (8) is a probability density function of the mean μ and the standard deviation σ represented by the equation (9). _{Further, the value w k} in the equation (8) represents a weight that mixes K normal distributions, and satisfies the equation (10).

_{The mean μ k} , standard deviation σ _k, and weight w _k in equation (8) are generally described in the literature “CM Bishop,“ 9.2 Mixed Gaussian Distribution ”, Pattern Recognition and Machine Learning, Maruzen Publishing, 2015”. It can be calculated by a method called Expectation-Maximization (EM) algorithm as described above.

At this time, the mixing number K may be determined in advance or may be determined by performing a convergence calculation. In the convergence calculation, for example, f _RL ⁽¹⁾ (x), f _RL ⁽²⁾ (x), ... _Are ^{calculated in order from K = 1, and f RL (K-1)} (x) and f _RL ^{( K) K} when the difference from (x) becomes sufficiently small is defined as the final mixed number K.

Next, the target existence likelihood calculator 23 derives a _{value x max} , which is a value x at which _{the probability density function f RL} ^{(K) (x) of the mixed normal distribution is maximized.} The target existence likelihood calculator 23 derives and derives a value x at which the derivative of _{the probability density function f RL} ^(K) (x) is “0” based on the equation (11) by a general numerical calculation method. Of the values, the value x at which the probability density function f _RL ^(K) (x) is maximized is defined as the value x _max .

Alternatively, the target existence probability calculator 23 sets a value between the minimum value RL _min and the maximum value RL _{max of N predicted reception levels RL p} ⁽ⁿ⁾ (r) as shown in equation (12). Create J sample values of x, and of the probability density functions f _RL ^(K) (x _j ) (j = 1, 2, ..., J), the probability density functions f _RL ^(K) (x _j ) _{May be set to "x max} = x _j " for the value j at which is the maximum.

Finally, the target existence likelihood calculator 23 calculates the target existence likelihood p _RL (r | RL) based on the equation (13).

As described above, in the second embodiment, the probability density function representing the distribution of a plurality of predicted reception levels is estimated, and the target existence likelihood is calculated based on the probability density function. The probability density function at this time is represented by a mixed normal distribution which is the sum of the probability density functions of a plurality of normal distributions. As a result, in the first embodiment, the accuracy of the target existence likelihood is determined according to the number of divisions M when calculating the _{frequency distribution P m, but in the second embodiment, it depends on the number of divisions M.} The target existence likelihood can be calculated appropriately without any problem.

Embodiment 3.
Next, the third embodiment will be described. The third embodiment is different from the first and second embodiments in that the target existence likelihood is calculated by using the standard deviation of the sound source level. In the following description, the parts common to the first and second embodiments are designated by the same reference numerals, and detailed description thereof will be omitted.

[Configuration of target existence likelihood calculation device]
FIG. 5 is a block diagram showing an example of the configuration of the target existence likelihood calculation device according to the third embodiment. As shown in FIG. 5, the target existence likelihood calculation device 30 includes a propagation loss calculator 11, a predicted reception level calculator 12, a target existence likelihood calculator 33, input terminals 1a to 1f as input units, and outputs. It has a terminal 2a. The target existence likelihood calculation device 30 is composed of an arithmetic unit such as a microcomputer that realizes various functions by executing software, or hardware such as a circuit device corresponding to various functions.

The target existence likelihood calculator 33 is connected to the input terminals 1d and 1f and the predicted reception level calculator 12 on the input side, and the output terminal 2a is connected to the output side. _{The standard deviation σ SL} of the sound source level SL is input to the input terminal 1f.

The target existence likelihood calculator 33 calculates the actual measurement reception level RL of the signal radiated from the target object input to the input terminal 1d, the standard deviation σ _SL of the sound source level input to the input terminal 1f, and the predicted reception level. The target existence likelihood p _RL (r | RL) is calculated based on the N predicted reception levels RL _p ^{(n) (r) input from the device 12.} The target existence likelihood calculator 13 outputs the calculated target existence likelihood p _RL (r | RL) from the output terminal 2a.

_{The target existence likelihood p RL} (r | RL) output from the output terminal 2b is visually displayed by a graph, a map, or the like on a display unit (not shown) provided as a part of the output unit, for example. The display of the target existence likelihood p _RL (r | RL) is not limited to this example, and may be performed by, for example, a display unit provided in the sonar device equipped with the target existence likelihood calculation device 30. ..

[Calculation method of target existence likelihood]
Next, a method of calculating the target existence likelihood in the target existence likelihood calculation device 30 according to the third embodiment will be described. When environmental information, own ship information, sound source level SL, and actual measurement reception level RL from the target object are input to the target existence likelihood calculation device 30, the target existence likelihood calculation device 30 inputs the input information. It is used to calculate the target existence likelihood p _RL (r | RL).

In the third embodiment, the ^{method of calculating the propagation loss TL (n)} (r) by the propagation loss calculator 11 and the calculation of the predicted reception level RL _p ⁽ⁿ⁾ (r) by the predicted reception level calculator 12 The method is the same as in the first and second embodiments. Therefore, here, a method of calculating the target existence likelihood by the target existence likelihood calculator 33 will be described.

(Calculation of target existence likelihood)
In the third embodiment, the target existence likelihood calculator 33 has a probability density function f _RL ^(K) _{representing the distribution of N predicted reception levels RL p} ^{(n) (r), as in the second embodiment.} (X) is represented by the mixed normal distribution shown in the equation (8).

Here, the probability density function f _RL ^(K) (x) of the predicted reception level includes the uncertainty of the propagation loss due to the uncertainty of the environmental information, but another one for calculating the predicted reception level. It does not include the uncertainty of the sound source level, which is an element. Therefore, assuming that the uncertainty of the sound source level follows a normal distribution, the probability density function f _RL ^(K) (x) is expressed by Eq. (14) _{using the standard deviation σ SL of the sound source level input from the input terminal 1f.} Converted to the probability density function g _RL ^(K) (x) to be performed.

In this probability density function g _RL ^(K) (x), the standard deviation of each of the K normal distributions to be mixed is the standard deviation of the sound source level as compared with the _{standard deviation σ k in the probability density function f RL} ^(K) (x). It shows that the deviation is increased by the amount of _{σ SL.}

Next, the target existence likelihood calculator 33 derives a _{value x max} , which is a value x at which _{the probability density function g RL} ^{(K) (x) is maximized.} The target existence likelihood calculator 33 is derived by deriving a value x at which the derivative of _{the probability density function g RL} ^{(K) (x) is “0” based on the equation (15) by a general numerical calculation method.} Of the values, the value x at which the probability density function g _RL ^(K) (x) is maximized is defined as the value x _max .

Alternatively, the target existence probability calculator 33 sets a value between the minimum value RL _min and the maximum value RL _{max of N predicted reception levels RL p} ⁽ⁿ⁾ (r) as shown in equation (16). Create J sample values of x, and of the probability density functions g _RL ^(K) (x _j ) (j = 1, 2, ..., J), the probability density functions g _RL ^(K) (x _j ) _{May be set to "x max} = x _j " for the value j at which is the maximum.

Finally, the target existence likelihood calculator 33 calculates the target existence likelihood p _RL (r | RL) based on the equation (17).

As described above, in the third embodiment, the target existence likelihood is calculated by further using the standard deviation of the input sound source level. As a result, not only the uncertainty of the propagation loss due to the uncertainty of the environmental information but also the uncertainty of the sound source level can be included as the cause of the uncertainty of the predicted reception level, so that the target existence likelihood is calculated more accurately. be able to.

In the above-described first to third embodiments, it has been described that only the environmental information input to the input terminal 1a fluctuates, but this is not limited to this example, and for example, the own ship information input to the input terminal 1b. Alternatively, the frequency information input to the input terminal 1e may also fluctuate. That is, the propagation loss calculator 11 may calculate a plurality of propagation losses using the fluctuating own ship information or frequency information.

1a, 1b, 1c, 1d, 1e, 1f input terminal, 2a output terminal, 10, 20, 30 Target existence likelihood calculator, 11 Propagation loss calculator, 12 Prediction reception level calculator, 13, 23, 33 Target existence Likelihood calculator, 41 processing circuits, 51 processors, 52 memories.

Claims (8)

It is a target existence likelihood calculation device that is mounted on a sonar device that estimates the distance to a target and calculates the target existence likelihood that indicates the likelihood of the position where the target exists.
A plurality of environmental information according to uncertainties indicating the environmental state around the sonar device, own ship information indicating the state of the ship on which the sonar device is mounted, an actual measurement reception level of a signal radiated from the target, and the above. An input unit into which the sound source level obtained from the target and the frequency information obtained from the target are input, and
A propagation loss calculator that calculates a plurality of propagation losses according to uncertainty based on the plurality of environmental information, the own ship information, and the frequency information.
A predicted reception level calculator that calculates a plurality of predicted reception levels according to uncertainty, which indicates a predicted value of a signal radiated from the target based on the sound source level and the calculated plurality of propagation losses.
A target existence likelihood calculation device including a target existence likelihood calculator that calculates the target existence likelihood based on the actually measured reception level from the target and the calculated predicted reception levels. .. The target existence likelihood calculator is
Based on the plurality of predicted reception levels, the frequency distribution of the predicted reception levels is calculated.
The target existence likelihood calculation device according to claim 1, wherein the target existence likelihood is calculated based on the calculated frequency distribution. The target existence likelihood calculator is
Estimate the probability density function that represents the distribution of the plurality of predicted reception levels.
The target existence likelihood calculation device according to claim 1, wherein the target existence likelihood is calculated based on the estimated probability density function. The target existence likelihood calculator is
The target existence likelihood calculation device according to claim 3, wherein the probability density function is estimated as being represented by a mixed normal distribution which is the sum of the probability density functions of a plurality of normal distributions. Further, the standard deviation of the sound source level is input to the input unit.
The target existence likelihood calculator is
The target existence likelihood calculation device according to claim 1, wherein the target existence likelihood is calculated based on the standard deviation of the sound source level, the actually measured reception level, and the calculated predicted reception level. The target existence likelihood calculator is
Using the standard deviation of the sound source level, a probability density function representing the distribution of the plurality of predicted reception levels is estimated.
The target existence likelihood calculation device according to claim 5, wherein the target existence likelihood is calculated based on the estimated probability density function. The target existence likelihood calculator is
The target existence likelihood calculation device according to claim 6, wherein the probability density function is estimated as being represented by a mixed normal distribution which is the sum of the probability density functions of a plurality of normal distributions. It is a target existence likelihood calculation method that is mounted on a sonar device that estimates the distance to a target and calculates the target existence likelihood that indicates the likelihood of the position where the target exists.
Based on a plurality of environmental information according to uncertainty indicating the environmental state around the sonar device, own ship information indicating the state of the ship on which the sonar device is mounted, and frequency information obtained from the target. Calculate multiple propagation losses according to uncertainty,
Based on the sound source level obtained from the target and the calculated plurality of propagation losses, a plurality of predicted reception levels according to uncertainty, which indicate the predicted values of the signals radiated from the target, are calculated.
A target existence likelihood calculation method, characterized in that the target existence likelihood is calculated based on an actually measured reception level of a signal radiated from the target and the calculated predicted reception levels.

Images