JP6154540B2 - Ship characteristic estimation device - Google Patents

Description

  The present invention relates to a ship characteristic estimation apparatus that estimates the characteristics of a ship.

  Conventionally, various attempts have been made to accurately grasp the state of a ship (the characteristics of a ship that changes over time). For example, Patent Document 1 discloses a system capable of diagnosing the state of a ship based on operation information (temperature, pressure, rotation speed, etc. of the main engine of the ship) detected by various sensors attached to the ship. Yes.

JP 2002-316692 A

  By the way, in grasping the state of the ship, grasping the relationship between the value of the parameter that affects the water speed of the ship (the speed of the ship relative to water) and the actual water speed of the ship corresponding to each value. Is very important. By grasping this relationship, for example, it becomes possible to make a ship navigation plan more accurately as an example. However, the above-mentioned patent document does not describe anything about this.

  For example, the parameters affecting the water velocity of the ship include the rotation speed of the ship's propeller and the wind force with respect to the ship. Then, in a certain ship, when trying to grasp the relationship between the rotation speed of the propeller and the water speed of the ship caused by the rotation speed, other parameters (wind power, etc.) that affect the water speed of the ship are set. It is necessary to conduct experiments while maintaining a constant state, which is very time-consuming. The same applies to the case of grasping the relationship between the wind force for the ship and the water speed of the ship caused by the wind force.

  The present invention is for solving the above-mentioned problems, and its purpose is to facilitate the relationship between the value of a parameter that affects the water speed of a ship and the water speed of the ship caused by the parameter. To figure out.

  (1) In order to solve the above-described problem, a ship characteristic estimation device according to an aspect of the present invention includes a surface flow velocity vector that is a surface flow velocity vector at a target point where a navigating ship is located, and the target point. In response to the input of the rotation speed of the propeller of the ship at, the value corresponding to each condition specified by the combination of the value of the surface flow velocity vector received and the value of the rotation speed is caused by the rotation of the propeller A target speed vector obtained by synthesizing a propeller propulsion speed vector that is a speed vector of the ship and the surface flow speed vector, and an estimator that outputs the estimated speed vector, and the target speed vector estimated by the estimator A propeller propulsion speed calculation unit that calculates the propeller propulsion speed vector based on an output value and the surface layer flow velocity vector , And a.

  (2) Preferably, the propeller propulsion speed calculation unit calculates the propeller propulsion speed vector by subtracting the surface layer flow speed vector from the output value output from the estimator.

  (3) In order to solve the above-described problem, a ship characteristic estimation device according to an aspect of the present invention includes a ground speed calculation unit that calculates a ground speed vector of a ship at a target point where a navigating ship is located, and the target A surface flow velocity vector that is a surface flow velocity vector at the point, and an input of the rotation speed of the propeller of the ship at the target point, and the values of the received surface flow velocity vector and the rotation speed A value corresponding to each condition specified by the combination is estimated as a target speed vector in which a propeller propulsion speed vector that is a speed vector of the ship due to the rotation of the propeller and the surface flow speed vector are combined. The output estimator, the output value as the target speed vector estimated by the estimator, and the ground speed calculation unit Based on said ground speed vectors, and a, a wind propulsion speed calculation unit for calculating a wind propulsion velocity vector is a velocity vector of the ship caused by wind against the ship in the target point.

  The ground speed is a speed relative to the ground.

  (4) Preferably, the wind power propulsion speed calculation unit calculates the wind power propulsion speed vector by subtracting the output value output from the estimator from the ground speed vector.

  (5) Preferably, the estimator is configured using a neural network.

  (6) Preferably, the ship characteristic estimation device further includes a ground speed calculation unit that calculates a ground speed vector of the ship, and the estimator is an average value of the plurality of ground speed vectors corresponding to the respective conditions. Is output as the output value.

  (7) Preferably, the ship characteristic estimation device includes a ground speed calculation unit that calculates a ground speed vector of the ship, the output value from the estimator, and the ground speed vector calculated from the ground speed vector. And an updating unit that updates the estimator so that an error between the output value and the ground speed vector is reduced.

  (8) More preferably, the estimator is configured using a neural network, each of which includes a plurality of input units to which any one of the information on the surface flow velocity vector and the rotation speed is input, and the target velocity. An output unit that outputs the output value as a vector, and a value output from the input side unit in the neural network is multiplied by a coupling coefficient and then transmitted to the output side unit, and the update The unit compares the output value with the ground speed vector as a teacher signal, and updates the coupling coefficient so that an error between the output value and the teacher signal is reduced.

  (9) Preferably, the ship characteristic estimation device further includes a ground speed calculation unit that calculates a ground speed vector of the ship, and a GNSS signal reception unit that is mounted on the ship and receives a GNSS signal, The ground speed calculator calculates the ground speed vector based on the GNSS signal received by the GNSS signal receiver and the time when the GNSS signal was received.

  (10) Preferably, the ship characteristic estimation device further includes a propeller rotational speed detection unit that detects the rotational speed, and a tide meter that measures the surface layer flow velocity vector, and the estimator includes the propeller The rotation speed detected by the rotation speed detector and the surface flow velocity vector measured by the tide meter are input.

(11) In order to solve the above-described problem, a ship characteristic estimation device according to an aspect of the present invention receives the rotation speed of the propeller of the ship at a target point where the navigating ship is located, and the received rotation speed And an estimator for estimating and outputting a propeller propulsion speed vector that is a speed vector of the ship resulting from the rotation of the propeller.

  (12) Preferably, the estimator is configured using a neural network.

  (13) Preferably, the ship characteristic estimation device further includes a water speed calculation unit that calculates a water speed vector of the ship, and the estimator includes a plurality of the water speeds when the rotational speed is detected. An average value of a plurality of the water speed vectors corresponding to each value of the rotation speed is output as the propeller propulsion speed vector.

  (14) Preferably, the ship characteristic estimation device is configured to calculate a water speed vector of the ship, a water speed calculation unit that calculates an output value from the estimator, and the water speed calculation unit. An update unit that compares the water velocity vector and updates the estimator so as to reduce an error between the output value and the water velocity vector is further provided.

  (15) More preferably, the estimator is configured using a neural network, and includes an input unit to which the rotational speed is input and an output unit to output the output value, and the input side in the neural network The value output from the unit is multiplied by a coupling coefficient and then transmitted to the output unit, and the updating unit compares the output value with the water velocity vector as a teacher signal. The coupling coefficient is updated so that an error between the output value and the teacher signal is reduced.

  (16) Preferably, the ship characteristic estimation device is based on an output value output from the estimator and a water velocity vector of the ship at the target point, and is caused by wind force with respect to the ship at the target point. A wind power propulsion speed calculating unit that calculates a wind power propulsion speed vector that is a speed vector of the ship;

  (17) More preferably, the wind power propulsion speed calculation unit calculates the wind power propulsion speed vector by subtracting the output value output from the estimator from the water speed vector.

  ADVANTAGE OF THE INVENTION According to this invention, the relationship between the value of the parameter which influences the water speed of a ship, and the water speed of the ship resulting from this parameter can be grasped | ascertained easily.

It is a block diagram which shows the structure of the ship characteristic estimation apparatus which concerns on embodiment of this invention. It is a schematic diagram which shows an example of a structure of the estimator shown in FIG. It is a vector diagram which shows the relationship between object speed, propeller propulsion speed, and surface layer flow speed. It is a figure which shows an example of the main machine characteristic displayed on a display part. It is a figure for demonstrating the reason the output value from an estimator converges on object speed. It is a block diagram which shows the structure of the ship characteristic estimation apparatus which concerns on a modification. It is a block diagram which shows the structure of the ship characteristic estimation apparatus which concerns on a modification. It is a figure for demonstrating in detail about the estimator shown in FIG. It is a block diagram which shows the structure of the ship characteristic estimation apparatus which concerns on a modification. It is a block diagram which shows the structure of the ship characteristic estimation apparatus which concerns on a modification. It is a block diagram which shows the structure of the learning coefficient setting process part shown in FIG. It is a figure which shows typically the table memorize | stored in a memory | storage part, and the self-organization map SOM memorize | stored corresponding to each cell of the table. It is a block diagram which shows the structure of the learning coefficient setting process part of the surface layer flow estimation apparatus which concerns on a modification. It is a figure which shows the table memorize | stored in a memory | storage part, and the learning data memorize | stored corresponding to each cell of the table. It is a block diagram which shows the structure of the ship characteristic estimation apparatus which concerns on a modification. It is a schematic diagram for demonstrating complementation of learning data. It is a figure which shows an example of the rotational speed meter displayed on a display part, Comprising: (A) is a figure of the rotational speed meter in the state in which the main machine has not deteriorated, (B) is the rotation in the state in which the main equipment has deteriorated somewhat. The speedometer diagram, (C) is a diagram of the tachometer when the main machine has deteriorated and needs repair. It is a block diagram which shows the structure of the ship characteristic estimation apparatus which concerns on a modification. It is a figure which shows an example of the wind force characteristic displayed on a display part. It is a block diagram which shows the structure of the ship characteristic estimation apparatus which concerns on a modification. It is a figure which shows an example of the example of a display of the display part of FIG. It is a block diagram which shows the structure of the ship characteristic estimation apparatus which concerns on a modification. It is a figure which shows typically an example of a structure of the estimator shown in FIG. It is a figure for demonstrating the reason the output value from an estimator converges on a propeller propulsion speed. It is a block diagram which shows the structure of the ship characteristic estimation apparatus which concerns on a modification. It is a block diagram which shows the structure of the ship characteristic estimation apparatus which concerns on a modification. It is a figure for demonstrating in detail about the estimator shown in FIG. It is a block diagram which shows the structure of the ship characteristic estimation apparatus which concerns on a modification. It is a block diagram which shows the structure of the ship characteristic estimation apparatus which concerns on a modification. It is a vector diagram showing the relationship between water speed, propeller propulsion speed, and wind power propulsion speed V. It is a block diagram which shows the structure of the water speed calculation part for calculating a water speed.

  A ship characteristic estimation device 1 according to an embodiment of the present invention will be described with reference to the drawings. The ship characteristic estimation apparatus 1 which concerns on embodiment of this invention is mounted in the own ship (ship). The ship characteristic estimation device 1 estimates a speed resulting from rotation of a propeller of a ship among water speeds of the ship. The ship characteristic estimation device 1 automatically obtains predetermined parameters (in this embodiment, the rotation speed of the propeller of the ship, the bow direction surface flow velocity, and the starboard direction surface flow velocity) at each predetermined timing. For each condition specified by the combination of parameter values, the speed resulting from the rotation of the propeller of the ship is calculated.

  In the following, the speed resulting from the rotation of the propeller of the ship out of the water speed of the ship is referred to as the propeller propulsion speed.

[overall structure]
FIG. 1 is a block diagram showing a configuration of a ship characteristic estimation apparatus 1 according to an embodiment of the present invention. As shown in FIG. 1, the ship characteristic estimation device 1 includes a GPS signal reception unit 2, a propeller rotation speed detection unit 3, a tide meter 4, a calculation unit 10, and a display unit 5.

  The GPS signal receiving unit 2 is provided as a GNSS signal receiving unit for receiving a GPS signal as a navigation signal (GNSS signal) transmitted from a navigation satellite (not shown). The GPS signal receiving unit 2 is configured by, for example, a GPS antenna. The GPS signal received by the GPS signal receiving unit 2 (that is, the position information of the ship) is notified to the calculation unit 10 together with the time when the GPS signal is received.

  In this embodiment, the GPS signal receiving unit 2 is used as the GNSS signal receiving unit. However, the present invention is not limited to this, and a receiving unit used in other GNSS systems may be used. Here, GNSS is an abbreviation for Global Navigation Satellite Systems (GNSS). This GNSS is a collective term for “GPS” operated by the United States, “GALILEO” operated by Europe, “GLONASS” operated by Russia, and the like.

  The propeller rotational speed detection unit 3 is for detecting the rotational speed per unit time of the propeller for generating the thrust of the ship, and is constituted by a sensor capable of detecting the rotational speed, for example. The rotation speed detected by the propeller rotation speed detection unit 3 is notified to the calculation unit 10.

  The tide meter 4 is for measuring the flow velocity of the surface layer of seawater at the ship position, specifically, the surface flow velocity in the bow direction and the surface flow velocity in the starboard direction. Equipped to be exposed to. The tide meter 4 measures a surface current in a relatively shallow portion at the ship position. The bow direction surface flow velocity and starboard direction surface flow velocity measured by the tide meter 4 are notified to the arithmetic unit 10.

  The calculation unit 10 estimates the propeller propulsion speed at the ship position (target point) at each predetermined timing based on various information notified from the GPS signal reception unit 2, the propeller rotation number detection unit 3, and the tide meter 4. To do. The calculation unit 10 includes a ground speed calculation unit 11, an estimator 12, a propeller propulsion speed calculation unit 13, a coupling coefficient update unit 14, and a main engine characteristic calculation unit 15.

  The ground speed calculator 11 calculates the ground speed (ground speed vector) of the ship based on the position information of the ship notified from the GPS signal receiver 2 and the time when the ship position information is acquired. Specifically, the ground speed calculation unit 11 calculates the ground speed of the ship based on the ship position at at least two timings and the time when the position information of each ship position is acquired. The ground speed calculation unit 11 notifies the coupling coefficient update unit 14 of the ground speed calculated in this way.

  The estimator 12 is configured to estimate a speed (hereinafter referred to as a target speed) obtained by combining the propeller propulsion speed and the surface layer flow speed at the ship position. In the present embodiment, the estimator 12 includes the propeller rotation speed detected by the propeller rotation speed detection unit 3, the bow direction surface flow velocity and the starboard surface flow at the ship position measured by the tide meter. Speed is entered. The estimator 12 sets a value corresponding to a condition specified by a combination of these input values (a condition specified by a combination of a certain rotation speed, a certain bow direction surface flow velocity, and a certain starboard direction surface flow velocity), The target speed (target speed vector) is output to the propeller propulsion speed calculation unit 13.

FIG. 2 is a diagram schematically illustrating an example of the configuration of the estimator 12. In this embodiment, the estimator 12 is configured using a generally known neural network. Specifically, the estimator 12, a plurality of input units _U IN_1 constituting the input _layer, U _{IN_2,} a _{U IN_3,} a plurality of intermediate units _U MID_1 constituting the hidden _layer, U _{MID_2,} a _{U MID_3,} output layer Output units U _{OUT — 1} and U _{OUT — 2} are included. The configuration of the estimator 12 shown in FIG. 2 is merely an example, and the number of units in each layer and the number of hidden layers are not limited to those shown in FIG.

The estimator 12, each of the input units _{U IN_1,} U _{IN_2,} when the input value _{U IN_3} (rotational speed of the propeller, etc.) is input, the coupling coefficient _{W I} for those input _{values, M} is multiplied , _And output to the hidden layer intermediate units U _{MID — 1} , U _{MID — 2} and U _{MID — 3} . Each of the hidden layer intermediate units U _{MID — 1} , U _{MID — 2} and U _{MID — 3 sums} the input values and multiplies the values based on the total values by the coupling coefficients W _{M, O} to the output units U _{OUT — 1} and U _{OUT — 2} . Output. The output units U _{OUT — 1} and U _{OUT — 2} sum each input value and output a value based on the total value to the coupling coefficient update unit 14 as an output value. Note that the value input to the estimator 12 does not necessarily have to be a parameter value itself such as a propeller rotational speed, but is a numerical value having a one-to-one relationship with those parameters (for example, proportional to the rotational speed). Or a voltage value that changes).

  In the estimator 12, appropriate initial values are set for the respective coupling coefficients W in the initial state. Each coupling coefficient W is updated by the coupling coefficient updating unit 14 as needed. Specifically, each coupling coefficient W is coupled to the coupling coefficient updating unit 14 so that an error between the output value output from the estimator 12 and the surface layer flow velocity (teacher signal) measured by the tide meter 4 is reduced. Updated by. As a result, the output value output from the estimator 12 converges to the target speed every time the coupling coefficient W is updated, as will be described in detail later.

  The propeller propulsion speed calculation unit 13 calculates the propeller propulsion speed based on the output value as the target speed output from the estimator 12 and the surface layer flow speed measured by the tide meter 4. Specifically, the propeller propulsion speed calculation unit 13 calculates the propeller propulsion speed by subtracting the surface layer flow speed from the target speed.

FIG. 3 is a vector diagram showing a relationship among the target velocity vector V _S , the propeller propulsion velocity vector _VP, and the surface layer flow velocity vector V _T. As described above, the target velocity vector V _S is a vector obtained by synthesizing the propeller propulsion velocity vector _VP and the surface flow velocity vector V _T. Thus, the propelling speed calculating section 13, as described above, by subtracting the surface flow velocity vector V _T from the target speed vector V _S, the propeller propulsion velocity vector V _P is calculated.

  The coupling coefficient update unit 14 sets the coupling coefficient W of the estimator 12 so that an error between the output value output from the estimator 12 and the ground speed (teacher signal) calculated by the ground speed calculation unit 11 is reduced. Update. For example, the coupling coefficient updating unit 14 updates the coupling coefficient W by using back propagation (error back propagation method), for example.

  The main engine characteristic calculation unit 15 is for calculating characteristics of the main engine of the ship (a mechanism part for generating thrust in the ship). In the present embodiment, the main engine characteristic calculator 15 calculates the propeller propulsion speed with respect to the rotation speed of the propeller. Based on the propeller propulsion speed for each propeller rotation speed calculated by the propeller propulsion speed calculation section 13, the main engine characteristic calculation section 15 obtains the main engine characteristics by using, for example, regression analysis.

  FIG. 4 is a diagram illustrating an example of main machine characteristics displayed on the display unit 5. The main unit characteristics as shown by the solid line in FIG. 4 are displayed on the display unit 5.

  In general, the main engine characteristics are ideal characteristics when a new boat is trialed or immediately after repair, and more specifically, characteristics such that the propeller propulsion speed with respect to the rotation speed of the propeller is increased (see the broken line in FIG. 4). . Therefore, for example, by comparing the main engine characteristics at the time of trial of a new boat with the main engine characteristics of the current ship, it is possible to infer the deterioration state of the engine, the fouling state of the hull surface, and the like. By knowing these states, the user can determine a more accurate operation plan, necessity of abnormality inspection, and the like.

[Output value output from the estimator]
5, the output value output from the estimator 12, for each coupling coefficient W of the estimator 12 is updated, the target speed (the speed obtained by combining the propeller propulsion velocity V _P and the surface velocity V _T) It is a figure for demonstrating the reason converged to. As described above, each coupling coefficient W stored in the estimator 12 has a coupling coefficient W so that an error between a value output from the estimator 12 and the ground speed as a teacher signal calculated from time to time becomes small. It is updated by the update unit 14.

  The wind speed varies in size and direction due to the sea area, time, weather conditions, and the like. Therefore, it is considered that the ground speed in the case where the rotation speed of the propeller and the surface flow velocity are the same includes wind speed components of all magnitudes and directions. Therefore, when these are averaged, the wind speed components cancel each other, and the speed component obtained by synthesizing the propeller propulsion speed and the surface layer flow speed, that is, the target speed component remains. That is, when the coupling coefficient W of the estimator 12 is updated so that the error between the output value of the estimator 12 and the ground speed becomes small as described above, the influence of the wind speed component included in the ground speed gradually increases. Therefore, the output value of the estimator 12 converges to the target velocity component described above. Therefore, when the learning is sufficiently advanced (that is, when the coupling coefficient is updated a sufficient number of times), the output value from the estimator 12 can be estimated as the target speed.

[effect]
As described above, in the ship characteristic estimation device 1 according to this embodiment, the propeller propulsion speed is based on the rotation speed of the propeller, the surface flow velocity measured by the tidal meter 4, and the target velocity estimated by the estimator 12. Is calculated. Thereby, the relationship between the rotation speed of the propeller and the water speed (propeller propulsion speed) of the ship at the rotation speed can be grasped relatively easily.

  Therefore, according to the ship characteristic estimation apparatus 1 according to the present embodiment, the relationship between the rotation speed of the propeller, which is a parameter affecting the water speed of the ship, and the water speed (propeller propulsion speed) of the ship corresponding thereto. Characteristics, that is, main engine characteristics can be easily grasped.

  Moreover, in the ship characteristic estimation apparatus 1, the estimator 12 is comprised using the neural network. Thereby, the estimator 12 capable of outputting the target speed can be appropriately configured.

  Moreover, in the ship characteristic estimation apparatus 1, the estimator 12 is updated so that an error between the output value of the estimator 12 and the ground speed calculated by the ground speed calculation unit 11 is reduced. Thereby, the estimator 12 provided with the learning function can be configured. In addition, since the ship characteristic estimation device 1 can accumulate a large amount of data necessary for estimating an accurate target speed during navigation, learning data (ground speed data under certain conditions) is prepared in advance. Save time and effort.

  Further, the ship characteristic estimation device 1 updates the coupling coefficient W stored in the estimator 12 so that an error between the output value of the estimator 12 and the ground speed calculated by the ground speed calculation unit 11 is reduced. ing. Thereby, the estimator 12 can be updated appropriately.

  Moreover, in the ship characteristic estimation apparatus 1, the propeller propulsion speed is calculated by subtracting the surface layer flow speed from the target speed. Thereby, the propeller propulsion speed can be calculated more easily.

Moreover, in the ship characteristic estimation apparatus 1, the ground speed can be accurately calculated using GNSS which is widely spread, particularly GPS. Moreover, since a general ship is often equipped with a GPS antenna, the ground speed can be calculated without introducing a new device or the like.

  Moreover, in the ship characteristic estimation apparatus 1, the rotation speed of the propeller and the surface layer flow velocity are detected by the propeller rotation speed detector 3 and the tide meter 4 mounted on the ship. Therefore, in the ship characteristic estimation apparatus 1, the propeller propulsion speed can be estimated using the propeller rotation speed detection unit 3 and the tide meter 4 that are widely used as general equipment.

  As mentioned above, although embodiment of this invention was described, this invention is not limited to these, A various change is possible unless it deviates from the meaning of this invention.

[Modification]
(1) FIG. 6 is a block diagram showing a configuration of a ship characteristic estimation device 1a according to a modification. Unlike the ship characteristic estimation apparatus 1 shown in FIG. 1, the ship characteristic estimation apparatus 1a according to the present modification has a configuration in which the coupling coefficient update unit 14 is omitted. That is, the ship characteristic estimation device 1a according to the present modification does not have a learning function. Further, in the ship characteristic estimation device 1a according to this modification, the GPS signal reception unit 2 and the ground speed calculation unit 11 are also omitted.

  In the ship characteristic estimation apparatus 1a according to the present modification, the output as the target speed is obtained by the estimator 12a in which the coupling coefficient W is determined based on a lot of previously acquired learning data (data on the ground speed at a certain condition). The value is output. Even with such a configuration, the propeller propulsion speed can be easily calculated as in the case of the above embodiment.

  (2) FIG. 7 is a block diagram showing a configuration of a ship characteristic estimation device 1b according to a modification. The ship characteristic estimation device 1b according to this modification is greatly different in the configuration of the estimator 12b from the ship characteristic estimation device 1 according to the above embodiment. Specifically, the estimator 12b is not configured using a neural network, but includes a storage unit 16 and an update unit 17.

  FIG. 8 is a diagram for explaining the estimator 12b shown in FIG. 7 in detail.

  As shown in FIG. 8, the storage unit 16 stores a matrix table. In this table, each condition (corresponding to each cell in the table) specified by a combination of each value (X1, X2, X3,...) Of the surface flow velocity and each value (R1, R2, R3,. ) The ground speed calculated at the time of) is stored. In FIG. 8, one ground speed value is indicated by one circle. That is, the storage unit 16 stores, for example, five ground speed values calculated when the surface flow velocity is X1 and the rotation speed is R1.

  When the rotational speed (for example, R2) detected by the propeller rotational speed detection unit 3 and the surface flow velocity (for example, X3) measured by the tide meter 4 are input to the estimator 12b, the estimator 12b An average value of the ground speeds (11 in the case of FIG. 8) included in the cell having the rotation speed R2 and the surface layer flow speed X3 is calculated. Then, the estimator 12b outputs the average value as an output value.

  As described above with reference to FIG. 5, if the ground speed under a certain condition (a condition specified by a combination of a certain rotational speed and a certain surface layer flow speed) is averaged, the wind speed components included in the ground speed cancel each other. The average value is close to the target speed (the speed obtained by combining the propeller propulsion speed and the surface layer flow speed). Therefore, the target speed can be appropriately estimated also by the estimator 12b according to this modification.

The updating unit 17 updates the table stored in the storage unit 16 using the ground speed calculated at the timing when the rotational speed and the surface flow velocity input to the estimator 12b are detected. Specifically, the ground speed calculated for a predetermined rotation speed (for example, R3) and a predetermined surface layer flow speed (for example, X2) is added to the cell specified by R3 and X2. By performing this operation as needed, learning data is accumulated even during navigation, and the target speed can be estimated more accurately. That is, the estimator 12b according to this modification also has a learning function. As a result, the propeller propulsion speed can be calculated more accurately.

  In addition, the ship characteristic estimation apparatus 1c which does not have a learning function can be comprised by setting it as the structure which abbreviate | omitted the update part 17, the GPS signal receiving part 2, and the ground speed calculation part 11 in this modification (FIG. 9). In this case, it is necessary to store a plurality of previously acquired learning data (data corresponding to one circle in FIG. 8) in the storage unit 16.

  (3) FIG. 10 is a block diagram showing a configuration of a ship characteristic estimation apparatus 1d according to a modification. The ship characteristic estimation device 1d according to the present modification further includes a learning coefficient setting processing unit 20.

  As in the case of the above embodiment, the estimator 12d is configured using a neural network, and is configured so that the coupling coefficient is updated as needed by so-called supervised learning. In the ship characteristic estimation device 1d, an error between the output value from the estimator 12d and the teacher signal (ground speed) is calculated. Then, the ship characteristic estimation device 1d updates the coupling coefficient W while propagating the error as a learning signal from the unit on the output layer side to the unit on the input layer side. The correction amount of the coupling coefficient is given by the following equation (1).

[Equation 1]
ΔW _{i, j} ^{n, n−1} (t) = ηδ _i ⁿ X _j ⁿ⁻¹ + αΔW _{i, j} ^{n, n−1} (t−1) (1)

In Expression (1), ΔW _{i, j} ^{n, n−1} (t) is a correction amount for the weight of the coupling between the unit j of the n−1 layer and the unit i of the n layer, η is a learning coefficient, and δ _i ⁿ is a learning signal returned from the unit i of the nth layer to each unit of the n−1 layer, X _j ⁿ⁻¹ is an output value of the unit j of the n−1 layer, α is a stabilization coefficient, and ΔW _{i, j} ^{n , n-1} (t-1) indicates the previous correction amount. Note that the n-1th layer is one layer on the input side than the nth layer.

  FIG. 11 is a block diagram illustrating a configuration of the learning coefficient setting processing unit 20. The learning coefficient setting processing unit 20 is for setting the learning coefficient in Expression (1) as needed. As illustrated in FIG. 11, the learning coefficient setting processing unit 20 includes a storage unit 21, an SOM update unit 22, a count unit 23, a learning coefficient calculation unit 24, and a learning coefficient setting unit 25. .

  FIG. 12 is a diagram schematically showing a table stored in the storage unit 21 and a self-organizing map SOM stored corresponding to each cell of the table. As shown in FIG. 12, the storage unit 21 stores a table cut in a mesh shape for each predetermined propeller rotational speed and for each predetermined surface layer flow velocity. A corresponding self-organizing map SOM is stored in each cell of this table. Each self-organizing map SOM of this modification is a two-dimensional SOM composed of n × n units. Each unit stores a reference vector having the same dimension as the input vector. In an initial state (a state in which learning is not performed), an appropriate reference vector is set for each unit.

  The SOM update unit 22 updates the SOM according to an input vector (a vector composed of a propeller rotation speed, a surface flow velocity, a ground velocity, etc., input at every predetermined timing). Specifically, the SOM update unit 22 updates the SOM stored in the cell including the input rotation speed and surface flow velocity as follows.

  Specifically, the SOM update unit 22 sets the unit having the shortest Euclidean distance from the input vector as the winner unit, the reference vector stored in the winner unit, and the reference vector stored in units around the winner unit. Is updated based on the following equation (2).

[Equation 2]
m _i (t + 1) = m _i (t) + h _i (t) [x (t) −m _i (t)] (2)

Here, m _i is a reference vector, x (t) is an input vector, and h _i is a neighborhood function represented by c · exp (−dis ² / α ² ). In the neighborhood function, c is a learning rate coefficient and dis = | x−m _c |. Here, _mc is a reference vector that minimizes the Euclidean distance from x (t).

  The SOM updating unit 22 updates the self-organizing map SOM at any time using the above-described equation (2) according to the input vector that is input at any time.

  The counting unit 23 counts the number of units having a reference vector whose difference (Euclidean distance) from the input vector is equal to or less than a threshold.

  The learning coefficient calculation unit 24 takes the reciprocal of the value counted by the counting unit 23 and calculates the value as a learning coefficient. That is, the learning coefficient is small when the count value is large (when there are many similar input data), and the learning coefficient is large when the count value is small (when there are few similar input data).

  The learning coefficient setting unit 25 notifies the estimator 12d of the value calculated by the learning coefficient calculation unit 24, and sets it as the learning coefficient η in the equation (1). The estimator 12d uses the learning coefficient η to update the coupling coefficient based on the equation (1), and then calculates a water velocity vector based on the updated coupling coefficient.

According to this modification, when a large number of similar learning data (input vectors) are accumulated, the learning coefficient becomes small. In this case, as is clear from the above-described expression (1), the correction amount ΔW _{i, j} ^{n, n−1} (t) of the coupling coefficient becomes small. On the other hand, when similar learning data is not accumulated or only a little is accumulated, the learning coefficient is increased. In this case, as is clear from the equation (1), the correction amount of the coupling coefficient becomes large. Thereby, according to this modification, the bias | inclination of the output value from an estimator resulting from accumulating many similar learning data can be suppressed.

  (4) FIG. 13 is a block diagram illustrating a configuration of the learning coefficient setting processing unit 40 of the ship characteristic estimation device according to the modification. As in the case of the learning coefficient setting processing unit 20 of the above-described modification, the learning coefficient setting processing unit 40 according to the present modification has a learning coefficient η of Expression (1) used in an estimator configured using a neural network. Is for setting. However, the learning coefficient setting processing unit 40 according to the present modification is different in configuration from the learning coefficient setting processing unit 20 of the above-described modification. As illustrated in FIG. 13, the learning coefficient setting processing unit 40 according to the present modification includes a storage unit 41, a learning coefficient calculation unit 42, and a learning coefficient setting unit 43.

FIG. 14 is a diagram showing a table stored in the storage unit 41 and learning data stored corresponding to each cell (each area) of the table. As shown in FIG. 14, the storage unit 41 stores a table cut in a mesh shape for each predetermined propeller rotational speed and for each predetermined surface layer flow velocity, as in the case of the above-described modification. In this modification, the learning data stored in each area is mapped according to the ground speed of each learning data. Specifically, as shown in FIG. 14, in a map having a plurality of subareas cut in a mesh shape for each predetermined bow direction ground speed and each predetermined starboard direction ground speed, each learning data is It is mapped according to the speed.

  The learning coefficient calculation unit 42 calculates the number of learning data (4 in the case of FIG. 14) stored in the sub-area including the most recently input learning data among all the sub-areas including the sub-area. A value obtained by normalizing the reciprocal of the value divided by the number of learning data stored in the subarea with the largest number of learning data (10 in subarea A in the case of FIG. 14) is set as the learning coefficient. Then, the learning coefficient setting unit 43 notifies the estimator of the learning coefficient set by the learning coefficient calculation unit 42 as in the learning coefficient setting unit 25 of the above modification, and sets it as the learning coefficient η in the equation (1). . Even with such a configuration, the learning coefficient can be set appropriately.

  (5) FIG. 15 is a block diagram showing a configuration of a ship characteristic estimation device 1e according to a modification. In this modification, the GPS signal receiving unit 2 and the like constituting a part of the ship characteristic estimation device 1e and the calculation unit 10e are mounted at different places. Specifically, in this modification, the GPS signal receiving unit 2, the propeller rotation number detecting unit 3, and the tide meter 4 are mounted on the own ship, and the calculation unit 10e is installed in, for example, a land center. 30. And the ship characteristic estimation apparatus 1e which concerns on this modification is provided with the transmission part 19a and the receiving part 19b which were mounted in the own ship.

  The transmission unit 19a transmits various data detected by the GPS signal reception unit 2, the propeller rotation number detection unit 3, and the tide meter 4 to the data center 30 via the antenna. The calculation unit 10h of the data center 30 calculates the propeller propulsion speed based on the various data, as in the case of the above embodiment. The calculation unit 10 e calculates the propeller propulsion speed and stores them in the database unit 31. The receiving unit 19b receives the propeller propulsion speed stored in the database unit 31. This data is displayed on the display unit 5.

  In this modification, unlike the case of the above-described embodiment, a calculation unit with a relatively large calculation load can be provided at a place different from the own ship. Thereby, even if it does not mount a calculating part in own ship, the surface layer flow velocity in the own ship position can be estimated.

  (6) In the above-described embodiment and each modification, the learning data can be supplemented when the learning data is not sufficiently accumulated.

FIG. 16 is a schematic diagram for explaining complementation of learning data. For example, for a ship navigating under certain conditions, a surface flow from 45 degrees on the port side and a surface flow from 45 degrees on the starboard side (the magnitude of the surface flow) Is the same as each other), the traveling direction is expected to be symmetrical. Thus, with reference to FIG. 16, for example, propeller speed predetermined rotational speed, orientation port rear 45 degree surface current, when the surface layer flow speed is a predetermined magnitude, the ground speed is met V _G In such a case, the data can be supplemented as follows based on the learning data. Specifically, the size of the propeller speed and surface velocity is the same as the learning data, orientation starboard rear 45 degree surface current, as learning data when the, vector V _'G obtained by mirror-inverting Can be complemented. By supplementing the learning data in this way, the propeller propulsion speed can be accurately estimated even in the initial stage where the learning data is not sufficiently accumulated, for example.

  (7) In the above-described embodiment, the surface layer flow velocity is measured using the tide meter 4, but the present invention is not limited to this, and the surface layer flow estimation device or the like having a mechanism different from that of the tide meter 4 may be used.

  (8) In the above embodiment, the main unit characteristic graph is displayed on the display unit 5 to notify the user of the current state of the main unit. However, the present invention is not limited to this. Specifically, an indicator indicating the deterioration state of the main engine can be displayed on the display unit.

  FIG. 17 is a diagram illustrating an example of the tachometer 6 displayed on the display unit. The tachometer 6 is provided with a scale portion 6a on which a numerical value of the rotation speed is written, and the indicator needle 6b is configured to indicate the rotation speed of the propeller.

  In the present embodiment, the color of the indicator needle 6b functions as an index indicating the deterioration state of the main machine. Specifically, in a state where the main engine characteristics have not deteriorated, the color of the indicating needle 6b is displayed in, for example, blue (no hatching) as shown in FIG. In the state where the main engine characteristics are slightly deteriorated, the color of the indicating needle 6b is displayed, for example, in yellow (hatching with a wide interval) as shown in FIG. When the main machine characteristics are greatly deteriorated and repair is necessary, the color of the indicator needle 6b is displayed in, for example, red (hatching with a narrow interval) as shown in FIG. As a result, the user can easily visually recognize the deterioration of the main engine characteristics.

  For example, the deterioration state of the main engine characteristics described above is, for example, a ratio between the propeller propulsion speed at the predetermined speed of the main engine at the time of the predetermined speed of the main engine at the time of the predetermined speed of the main engine at the time of trial of a new boat or immediately after repair. Can be determined based on

  The index indicating the deterioration state of the main engine is not limited to the color as described above. For example, a pattern corresponding to the deterioration state can be used, or the brightness of the tachometer 6 is gradually increased according to the deterioration state. It can also be darkened.

  In FIG. 17, the color of the indicator needle 6b is changed according to the deterioration state of the main machine. However, the present invention is not limited to this. For example, when the rotation speed of the main machine is digitally displayed, The color may be changed according to the deterioration state of the main engine. In addition, as long as it is an index indicating the deterioration state of the main engine, any character such as an announcement indicating the deterioration state may be used.

  (9) FIG. 18 is a block diagram showing a configuration of a ship characteristic estimation device 1f according to a modification. In the above embodiment and each embodiment, the propeller propulsion speed and the main engine characteristics are calculated as the ship characteristics. On the other hand, according to this modification, it is possible to calculate the wind power propulsion speed (the speed of water against the ship due to the wind force with respect to the ship) as the characteristics of the ship. As shown in FIG. 18, the ship characteristic estimation device 1 f according to the present modification includes a GPS signal receiving unit 2, a propeller rotational speed detection unit 3, a tide meter 4, a calculation unit 10 f, and a display unit 5. I have.

  The GPS signal receiver 2, the propeller rotation speed detector 3, and the tide meter 4 have the same configuration as in the above embodiment, and operate in the same manner as in the above embodiment.

  The calculation unit 10 f includes a ground speed calculation unit 11, an estimator 12, a wind propulsion speed calculation unit 26, a coupling coefficient update unit 14, and a wind power characteristic calculation unit 27. The ground speed calculation unit 11, the estimator 12, and the coupling coefficient update unit 14 have the same configuration as in the above embodiment, and operate in the same manner as in the above embodiment.

  As described above, the wind power propulsion speed calculation unit 26 calculates the wind power propulsion speed, which is the water speed of the ship, which is caused by the wind power of the ship. In the present embodiment, the wind power propulsion speed calculator 26 calculates the wind power propulsion speed based on the output value as the target speed output from the estimator 12 and the ground speed calculated by the ground speed calculator. Specifically, the wind propulsion speed calculation unit 26 calculates the wind propulsion speed by subtracting the target speed (the output value from the estimator 12) from the ground speed.

With reference to FIG. 3, the ground speed V _G is a vector obtained by synthesizing the target speed V _S and the wind propulsion speed V _Wind . Accordingly, the wind power propulsion speed calculation unit 26 subtracts the target speed V _S from the ground speed V _G as described above, thereby calculating the wind power propulsion speed V _Wind .

  The wind power characteristic calculation unit 27 is for calculating at what speed the ship moves by wind power with respect to the ship. In the present embodiment, the wind power characteristic calculation unit 27 calculates the wind propulsion speed with respect to the direction and magnitude of the wind force.

FIG. 19 is a diagram illustrating an example of wind power characteristics displayed on the display unit 5. For example, wind characteristics as shown in FIG. 19 are displayed on the display unit 5, for example. In FIG. 19, it is assumed that the wind is blowing from each direction in the horizontal direction when the wind speed is V ₁ [m / s], V ₂ [m / s], and V ₃ [m / s]. In this case, the moving speed of the ship (that is, the wind propulsion speed) is displayed. Thereby, the user can know how much the own ship is made to flow (moves) at what wind direction and wind speed. Then, for example, the user can estimate the future route more appropriately as an example.

  In addition, each modification demonstrated as a modification in the ship characteristic estimation apparatus 1 which concerns on the said embodiment can be applied suitably to this modification.

  (10) FIG. 20 is a block diagram showing a configuration of a ship characteristic estimation apparatus 1g according to a modification. In the embodiment and the modification, the main engine characteristic or the wind power characteristic is calculated as the ship characteristic. On the other hand, as shown in FIG. 20, according to this modification, both the main engine characteristics and the wind power characteristics can be calculated.

FIG. 21 is a diagram illustrating an example of a display example of the display unit 5 of the ship characteristic estimation device 1g according to the present modification. In the ship characteristic estimation device 1g according to this modification, the range of the ship position after a predetermined time has elapsed based on the main engine characteristic calculated by the main engine characteristic calculation unit 15 and the wind power characteristic calculated by the wind power characteristic calculation unit 27. Can be estimated. FIG. 21 shows the range of the ship position when the rotation speed of the main engine is maintained at a predetermined rotation speed. Specifically, the range indicated by the solid line in FIG. 21 is the case where the main engine continues navigating at a predetermined rotational speed, and the maximum wind speed V ₃ [m / s] continues to blow. The range in which the ship can exist is shown. Thereby, the user can predict the range in which the ship can exist after a predetermined time has elapsed.

  In the present modification, for example, when the user inputs a desired rotation number, the position range can be predicted based on the rotation number. Furthermore, in this modification, it is also possible to perform more accurate position prediction by taking into account the surface layer flow velocity.

  (11) FIG. 22 is a block diagram showing a configuration of a ship characteristic estimation device 1h according to a modification. The ship characteristic estimation device 1h according to this modification example calculates the propeller propulsion speed and the main engine characteristic as in the ship characteristic estimation apparatus according to the above-described embodiment and each modification example, but the calculation method is different.

  As shown in FIG. 22, the ship characteristic estimation device 1 h according to the present modification includes a propeller rotation speed detection unit 3, a hydrometer 7, an estimator 12 h, a coupling coefficient update unit 14, and a main engine characteristic calculation unit. 15. Since the propeller rotational speed detection unit 3 operates in the same manner as in the above embodiment, a detailed description thereof is omitted.

  The water speed meter 7 is for measuring the water speed near the ship and is mounted on the ship. As an example of the water velocimeter 7, an acoustic water velocimeter can be cited as an example. However, the water velocimeter 7 is not limited to this and may be of other types.

FIG. 23 is a diagram schematically illustrating an example of the configuration of the estimator 12h. In the present embodiment, the estimator 12h is configured using a neural network as in the case of the above-described embodiment (see FIG. 2). However, unlike the case of the above-described embodiment, one input unit U _{IN_1} to which the rotation speed of the propeller is input is provided. Note that the configuration of the estimator 12h is merely an example, and the number of units in each layer and the number of hidden layers are not limited to those illustrated in FIG.

  In the estimator 12h, appropriate initial values are set for the coupling coefficients W in the initial state. Each coupling coefficient W is updated by the coupling coefficient updating unit 14 as needed. Specifically, each coupling coefficient W is updated so that an error between the output value output from the estimator 12h and the water velocity (teacher signal) measured by the water velocity meter 7 is reduced. Updated by the unit 14.

  The coupling coefficient update unit 14 reduces the error between the output value output from the estimator 12h and the water speed (teacher signal) measured by the water speed meter 7 so that the coupling coefficient W of the estimator 12h is reduced. Update. For example, the coupling coefficient updating unit 14 updates the coupling coefficient W by using back propagation (error back propagation method), for example.

  The main machine characteristic calculation unit 15 calculates the main machine characteristic shown in FIG. 4 as an example by operating in the same manner as in the above embodiment.

[Output value output from the estimator]
FIG. 24 is a diagram for explaining the reason why the output value output from the estimator 12h converges to the propeller propulsion speed every time the coupling coefficient W of the estimator 12 is updated. As described above, each coupling coefficient W stored in the estimator 12h is combined so that an error between a value output from the estimator 12h at any time and a water speed as a teacher signal measured at any time is small. Updated by the coefficient updating unit 14.

  As described above in the embodiment, the wind speed has different sizes and directions due to the sea area, time, weather conditions, and the like. Therefore, it is considered that the wind speed component of any magnitude and direction is included in the water speed when the rotation speed of the propeller is the same. Therefore, when these are averaged, the wind speed components cancel each other, and the propeller propulsion speed remains. That is, when the coupling coefficient W of the estimator 12h is updated so as to reduce the error between the output value of the estimator 12h and the water speed, as described above, the influence of the wind speed component included in the water speed. Is gradually reduced, the output value of the estimator 12h converges to the propeller propulsion speed. Therefore, when the learning is sufficiently advanced (that is, when the coupling coefficient is updated a sufficient number of times), the output value from the estimator 12h can be estimated as the target speed.

  As described above, in the ship characteristic estimation device 1h according to this modification, the propeller propulsion speed is calculated based on the rotation speed of the propeller and the water speed. Thereby, the relationship between the rotation speed of the propeller and the water speed (propeller propulsion speed) of the ship at the rotation speed can be grasped relatively easily. In addition, according to the ship characteristic estimation device 1h according to the present modification, the number of parameters required for estimating the propeller propulsion speed is smaller than in the case of the above-described embodiment, so that the main engine characteristic can be obtained more easily. be able to.

  Moreover, in the ship characteristic estimation apparatus 1h, the estimator 12h is configured using a neural network. Thereby, the estimator 12h that can output the propeller propulsion speed can be appropriately configured.

Moreover, in the ship characteristic estimation apparatus 1h, the estimator 12h is updated so that the error between the output value of the estimator 12h and the water speed is reduced. Thereby, the estimator 12h provided with the learning function can be configured. Moreover, since the ship characteristic estimation device 1h can accumulate a large amount of data necessary for estimating an accurate propeller propulsion speed during navigation, learning data (ground speed data under certain conditions) is prepared in advance. This saves you time and effort.

  Further, in the ship characteristic estimation device 1h, the coupling coefficient W stored in the estimator 12h is updated so that an error between the output value of the estimator 12h and the water velocity is reduced. Thereby, the estimator 12h can be updated appropriately.

  Note that the learning coefficient setting processing unit described with reference to FIGS. 10 to 14 may be provided in the ship characteristic estimation device 1h according to the present modification. Accordingly, an appropriate learning coefficient can be set as in the modification described with reference to FIGS. 10 to 14.

  (12) FIG. 25 is a block diagram showing a configuration of a ship characteristic estimation device 1i according to a modification. Unlike the ship characteristic estimation apparatus 1h shown in FIG. 22, the ship characteristic estimation apparatus 1i according to the present modification has a configuration in which the coupling coefficient update unit 14 is omitted. That is, the ship characteristic estimation device 1i according to this modification does not have a learning function. Moreover, the ship characteristic estimation apparatus 1i which concerns on this modification becomes a structure by which the water speedometer 7 was also abbreviate | omitted.

  In the ship characteristic estimation device 1i according to the present modification, the propeller is determined by the estimator 12i in which the coupling coefficient W is determined based on a large amount of learning data acquired in advance (water speed data at each rotation speed). An output value as a propulsion speed is output. Even with such a configuration, the propeller propulsion speed can be easily calculated as in the case of the above embodiment.

  (13) FIG. 26 is a block diagram showing a configuration of a ship characteristic estimation apparatus 1j according to a modification. The ship characteristic estimation device 1j according to the present modification is greatly different from the ship characteristic estimation device 1h shown in FIG. 22 in the configuration of the estimator 12j. Specifically, the estimator 12j is not configured using a neural network, but includes a storage unit 16j and an update unit 17j.

  FIG. 27 is a diagram for explaining the estimator 12j shown in FIG. 26 in detail.

  In the storage unit 16j, as shown in FIG. 27, a plurality of water speeds measured at the time of detecting the detected number of revolutions are stored for each corresponding number of revolutions (in the case of FIG. 27, R1, R2, R3, ... everything). In FIG. 27, one water speed value is indicated by one circle. That is, the storage unit 16j stores, for example, five water speed values calculated when the rotation speed is R1.

  When the rotational speed (for example, R2) detected by the propeller rotational speed detection unit 3 is input to the estimator 12j, the estimator 12j measures the water speed (10 in the case of FIG. 27) corresponding to the rotational speed R2. The average value of is calculated. Then, the estimator 12j outputs the average value as an output value.

  As described above with reference to FIG. 24, when the water speed at a certain number of revolutions is averaged, the wind speed components included in the water speed cancel each other, so the average value is close to the propeller propulsion speed. Become. Therefore, the propeller propulsion speed can be appropriately estimated also by the estimator 12j according to this modification.

The update unit 17j updates the table stored in the storage unit 16j using the water speed calculated at the timing when the rotation speed input to the estimator 12j is detected. Specifically, the water speed calculated at a predetermined rotation speed (for example, R3) is added to the cell specified by R3. By performing this operation as needed, learning data is accumulated even during navigation, and the target speed can be estimated more accurately. That is, the estimator 12j according to this modification also has a learning function. As a result, the propeller propulsion speed can be estimated more accurately.

  In addition, the ship characteristic estimation apparatus 1k which does not have a learning function can be comprised by setting it as the structure which abbreviate | omitted the update part 17j and the water speedometer 7 in this modification (refer FIG. 28). In this case, it is necessary to store a plurality of learning data obtained in advance (data corresponding to one circle in FIG. 27) in the storage unit 16k.

  (14) FIG. 29 is a block diagram showing a configuration of a ship characteristic estimation apparatus 11 according to a modification. The ship characteristic estimation device 11 according to the present modification has a configuration in which a wind propulsion speed calculation unit 26a and a wind characteristic calculation unit 27 are added to the ship characteristic estimation device 1h according to the modification described with reference to FIG. Yes. Thereby, in the ship characteristic estimation apparatus 1l which concerns on this embodiment, in addition to a main machine characteristic, a wind force characteristic can also be calculated.

FIG. 30 is a vector diagram showing a relationship among the water speed V _W , the propeller propulsion speed V _p, and the wind power propulsion speed V _Wind . The wind power propulsion speed calculation unit 26a calculates the wind power propulsion speed V _Wind based on the output value output from the estimator 12l (that is, the propeller propulsion speed V _p ) and the water speed measured by the water speed meter 7. calculate. Specifically, the wind power propulsion speed calculation unit 26a calculates the wind power propulsion speed by subtracting the propeller propulsion speed from the water speed.

  The wind force characteristic calculation unit 27 operates in the same manner as the wind force characteristic calculation unit 27 according to the modification described with reference to FIG. 18. For example, the wind force characteristic calculation unit 27 displays the wind force characteristic as illustrated in FIG. 19 on the display unit 5.

  Therefore, according to the present modification, wind power characteristics can be easily calculated as in the modification described with reference to FIG. Moreover, according to the ship characteristic estimation device 11 according to the present modification, the number of parameters required for estimating the wind power propulsion speed is smaller than in the case shown in FIG. Can do.

  (15) FIG. 31 is a block diagram showing the configuration of the water speed calculation unit for calculating the water speed. In each of the above-described modified examples (FIGS. 22, 26, 29, etc.), the water speed meter 7 is used to calculate the water speed. However, the present invention is not limited to this. For example, a configuration as shown in FIG. The water velocity calculation unit 8 may be used.

  The ground speed calculated by the ground speed calculator 11 and the surface layer speed measured by the tide meter 4 are input to the water speed calculator 8. Then, the water speed calculation unit 8 calculates the water speed by subtracting the surface layer flow speed from the ground speed. Thereby, the water velocity can be calculated appropriately.

1, 1a, 1b, 1c,..., 1l Ship characteristic estimation device 12, 12a,..., 12d, 12h,.

Claims (17)

Accepts the input of the surface flow velocity vector, which is the velocity vector of the surface flow at the target point where the vessel to be navigated, and the rotation speed of the propeller of the ship at the target point, and the value of the received surface flow velocity vector And a value corresponding to each condition specified by the combination of the rotation speed values, the propeller propulsion speed vector, which is the speed vector of the ship resulting from the rotation of the propeller, and the surface flow velocity vector are combined. An estimator that estimates and outputs a velocity vector;
A propeller propulsion speed calculation unit that calculates the propeller propulsion speed vector based on the output value as the target speed vector estimated by the estimator and the surface flow velocity vector;
The ship characteristic estimation apparatus characterized by comprising. In the ship characteristic estimation device according to claim 1,
The propeller propulsion speed calculation unit calculates the propeller propulsion speed vector by subtracting the surface flow velocity vector from the output value output from the estimator. A ground speed calculation unit for calculating a ground speed vector of the ship at a target point where the navigating ship is located;
Accepting the input of the surface flow velocity vector, which is the velocity vector of the surface flow at the target point, and the rotation speed of the propeller of the ship at the target point, the value of the received surface flow velocity vector and the rotation number A value corresponding to each condition specified by a combination of values is estimated as a target speed vector obtained by combining a propeller propulsion speed vector that is a speed vector of the ship due to rotation of the propeller and the surface flow speed vector. Output estimator,
Based on the output value as the target speed vector estimated by the estimator and the ground speed vector calculated by the ground speed calculation unit, the speed vector of the ship resulting from wind power for the ship at the target point A wind power propulsion speed calculation unit for calculating a wind power propulsion speed vector,
The ship characteristic estimation apparatus characterized by comprising. In the ship characteristic estimation apparatus according to claim 3,
The ship propulsion speed calculation unit calculates the wind power propulsion speed vector by subtracting the output value output from the estimator from the ground speed vector. In the ship characteristic estimation apparatus according to any one of claims 1 to 4,
The ship estimator, wherein the estimator is configured using a neural network. In the ship characteristic estimation apparatus according to any one of claims 1 to 4,
A ground speed calculation unit for calculating a ground speed vector of the ship;
The ship estimator, wherein the estimator is configured to output an average value of a plurality of ground speed vectors corresponding to the respective conditions as the output value. In the ship characteristic estimation apparatus according to any one of claims 1 to 6,
A ground speed calculation unit for calculating a ground speed vector of the ship;
The output value from the estimator and the ground speed vector calculated by the ground speed calculation unit are compared, and the estimator is adjusted so that an error between the output value and the ground speed vector is reduced. An update section to update;
The ship characteristic estimation apparatus characterized by further including. In the ship characteristic estimation device according to claim 7,
The estimator is configured using a neural network,
Each of the plurality of input units to which either the information on the surface flow velocity vector and the rotation speed are input,
An output unit that outputs the output value as the target velocity vector;
The value output from the input side unit in the neural network is multiplied by a coupling coefficient and then transmitted to the output side unit.
The update unit compares the output value with the ground speed vector as a teacher signal, and updates the coupling coefficient so that an error between the output value and the teacher signal is reduced. A ship characteristic estimation device. In the ship characteristic estimation apparatus according to any one of claims 1 to 8,
A ground speed calculation unit for calculating a ground speed vector of the ship;
A GNSS signal receiving unit that is mounted on the ship and receives a GNSS signal;
The ground speed calculation unit calculates the ground speed vector based on a GNSS signal received by the GNSS signal reception unit and a time when the GNSS signal is received. In the ship characteristic estimation apparatus of any one of Claims 1-9,
A propeller rotation number detection unit for detecting the rotation number;
A tidal meter for measuring the surface flow velocity vector,
The ship characteristic estimation device, wherein the estimator receives the rotation speed detected by the propeller rotation speed detection unit and the surface layer flow velocity vector measured by the tide meter.   The propeller which receives the rotation speed of the propeller of the ship at the target point where the vessel to be navigated is located, and a value corresponding to each value of the received rotation speed is a speed vector of the ship due to the rotation of the propeller. A ship characteristic estimation device comprising an estimator that estimates and outputs a propulsion speed vector. In the ship characteristic estimation apparatus according to claim 11,
The ship estimator, wherein the estimator is configured using a neural network. In the ship characteristic estimation apparatus according to claim 11,
A water velocity calculation unit for calculating a water velocity vector of the ship;
The estimator calculates an average value of the plurality of water speed vectors corresponding to each value of the rotation speed at the time of detecting the rotation speed, and the propeller propulsion speed vector. It is comprised so that it may output as, The ship characteristic estimation apparatus characterized by the above-mentioned. In the ship characteristic estimation apparatus according to any one of claims 11 to 13,
A water speed calculation unit for calculating a water speed vector of the ship;
The output value from the estimator is compared with the water speed vector calculated by the water speed calculation unit, and the estimation is performed so that an error between the output value and the water speed vector is reduced. An update unit for updating the container;
The ship characteristic estimation apparatus characterized by further including. In the ship characteristic estimation device according to claim 14,
The estimator is configured using a neural network,
An input unit for inputting the rotational speed;
An output unit for outputting the output value,
The value output from the input side unit in the neural network is multiplied by a coupling coefficient and then transmitted to the output side unit.
The updating unit compares the output value with the water velocity vector as a teacher signal, and updates the coupling coefficient so that an error between the output value and the teacher signal is reduced. A ship characteristic estimation device. In the ship characteristic estimation apparatus according to any one of claims 11 to 15,
Based on the output value output from the estimator and the water velocity vector of the ship at the target point, a wind propulsion speed vector that is a speed vector of the ship due to wind power with respect to the ship at the target point, A ship characteristic estimation device, further comprising a wind power propulsion speed calculation unit. In the ship characteristic estimation device according to claim 16,
The ship propulsion speed calculation unit calculates the wind propulsion speed vector by subtracting the output value output from the estimator from the water speed vector.

Images