Autonomous navigation method and apparatus for vessel

Abstract

The present invention discloses a method and apparatus for controlling autonomous navigation of a ship. By implementing an autonomous navigation control device for a ship using a discrete fuzzy PD controller instead of a conventional discrete PD controller, the conventional fixed control gain is varied according to the fuzzy control method. Consequently, the operating range of the autonomous navigation control device is not limited to a specific operating point, but rather appropriately controls the ship's heading angle across multiple operating ranges, thereby demonstrating superior performance in path estimation control.

Description

Technology Field

[0001] The present invention relates to a method and apparatus for controlling autonomous navigation of a ship, and more specifically, to a method and apparatus for controlling autonomous navigation of a ship through heading angle control. Background Technology

[0002] Autonomous navigation of a vessel is performed using an automatic controller to reduce the error between the target direction and the vessel's current direction. In this process, the automatic controller typically utilizes linear controllers, such as PID (Proportional Integral Derivative) controllers, and is designed to include a linear compensator to account for non-linear environmental influences such as wind and currents.

[0003] In particular, conventional autonomous navigation technology uses sway and yaw among the ship's 6 degrees of freedom motion to perform PID control of the ship's rudder angle so that the ship maintains a preset target heading angle.

[0004] An example of a ship autonomous navigation device using a PID controller according to the prior art is disclosed in Korean Patent No. 10-2167072 (Title of Invention: Control device and method of a free-running system for a water tank-based motion model test).

[0005] However, since PID controllers based on such conventional technology perform control using fixed control gains limited to a specific operating point, a problem arises where performance degrades in operating regions other than that specific operating point. The problem to be solved

[0006] The problem that the present invention aims to solve is to provide a method and apparatus for autonomous ship navigation control that can adaptively operate in various operating regions without being limited to a specific operating region through variable control gains, in order to resolve the problem that conventional PID controllers are limited to a specific operating point and perform control using fixed control gains.

[0007] In particular, the present invention aims to provide a method and apparatus for autonomous ship navigation control that effectively performs heading angle control to enable the ship to autonomously navigate a predetermined path while minimizing time and fuel consumption during autonomous ship navigation, thereby maintaining a minimum distance error for the predetermined path. means of solving the problem

[0008] A ship autonomous navigation control method according to a preferred embodiment of the present invention for solving the above-mentioned problem is a method performed in a ship autonomous navigation control device, comprising: (a) a control reference value of the k-th sampling interval and k-th sampling interval current state value The error between them Step of generating; (b) the error Delay by one sampling time interval, and the previous sampling interval ( Error calculated in ) Error rate by performing differentiation with respect to Step of generating; (c) the error rate Error acceleration by performing differentiation with respect to Step of generating; (d) the error rate and the above error acceleration Using, error Perform fuzzification and fuzzy inference processes to minimize the control signal increment A step of generating; and (e) a control signal output to the actuator in the previous sampling interval The above control signal increment Summing up the control signal It includes the step of outputting to an actuator.

[0009] In addition, the above step (d) includes (d1) the error rate and the above error acceleration A step of calculating the degree of membership of an input variable using as an argument of a membership function defined for each fuzzy rule; (d2) a step of selecting a degree of fit representing the firing strength of the corresponding rule among the fuzzy values ​​(degrees of membership) for each fuzzy rule; (d3) the error rate for each fuzzy rule and the above error acceleration A step of generating a conclusion value by applying pre-set parameters to; and (d4) performing defuscation using the conclusion value for each fuzzy rule and the suitability to the control signal increment It may include a step of generating.

[0010] Additionally, in step (d3) above, the parameters can be updated by applying an optimization algorithm after their initial values ​​are set using the proportional gain and derivative gain of the discrete PD controller for the vessel.

[0011] Meanwhile, a computer program according to a preferred embodiment of the present invention for solving the above-mentioned problem is stored in a non-transient storage medium and executed on a computer including a processor to perform the ship autonomous navigation control method.

[0012] Meanwhile, a ship autonomous navigation control device according to a preferred embodiment of the present invention for solving the above-mentioned problem comprises a processor and a memory storing predetermined instructions, and the processor that executes the instructions stored in the memory comprises: (a) a control reference value of the k-th sampling interval and k-th sampling interval current state value The error between them Step of generating; (b) the error Delay by one sampling time interval, and the previous sampling interval ( Error calculated in ) Error rate by performing differentiation with respect to Step of generating; (c) the error rate Error acceleration by performing differentiation with respect to Step of generating; (d) the error rate and the above error acceleration Using, error Perform fuzzification and fuzzy inference processes to minimize the control signal increment A step of generating; and (e) a control signal output to the actuator in the previous sampling interval The above control signal increment Summing up the control signal The method for controlling autonomous ship navigation is performed by performing the step of outputting to an actuator.

[0013] In addition, the above step (d) includes (d1) the error rate and the above error acceleration A step of calculating the degree of membership of an input variable using as an argument of a membership function defined for each fuzzy rule; (d2) a step of selecting a degree of fit representing the firing strength of the corresponding rule among the fuzzy values ​​(degrees of membership) for each fuzzy rule; (d3) the error rate for each fuzzy rule and the above error acceleration A step of generating a conclusion value by applying pre-set parameters to; and (d4) performing defuscation using the conclusion value for each fuzzy rule and the suitability to the control signal increment It may include a step of generating.

[0014] Additionally, in step (d3) above, the parameters can be updated by applying an optimization algorithm after their initial values ​​are set using the proportional gain and derivative gain of the discrete PD controller for the vessel. Effects of the invention

[0015] The present invention implements a ship autonomous navigation control device using a discrete fuzzy PD controller instead of a conventional discrete PD controller. By varying the conventional fixed control gain according to the fuzzy control method, the operating range of the ship autonomous navigation control device is not limited to a specific operating point, and the ship's heading angle control is appropriately performed across multiple operating ranges, thereby demonstrating superior performance in path estimation control. Brief explanation of the drawing

[0016] FIG. 1 is a diagram defining variables for performing control to maintain a target heading angle in order to effectively reach a target point by utilizing fore-and-aft, side-and-side, and bow motions during the 6-degree-of-freedom motion of a ship according to the present invention. Figure 2 is a block diagram illustrating the structure of a PD controller in a discrete system. Figure 3 shows the error velocity component values, which are the variables of the fuzzy PD controller. and error acceleration component values Each fuzzy membership function and It is a drawing that illustrates. Figure 4 is a diagram illustrating the time response of a general second-order system for establishing fuzzy logic rules. Figure 5 is a conceptual diagram showing the entire inference process of a discrete fuzzy PD controller. FIG. 6 is a diagram illustrating the configuration of a ship autonomous navigation control device according to a preferred embodiment of the present invention. Figure 7 is a diagram showing the functions performed by the processor as function blocks when instructions stored in memory are executed by the processor to implement a ship autonomous navigation control device. FIG. 8 is a flowchart illustrating a ship autonomous navigation control method performed by a ship autonomous navigation control device according to a preferred embodiment of the present invention. FIGS. 9 and 10 are drawings illustrating the process of determining parameters through an optimization algorithm according to a preferred embodiment of the present invention. FIG. 11 is a diagram illustrating the result of performing heading angle control to minimize the sailing distance while sequentially following waypoints from a starting point to a destination point for a catamaran-type vessel according to a preferred embodiment of the present invention. Specific details for implementing the invention

[0017] Preferred embodiments of the present invention will be described below with reference to the attached drawings.

[0018] First, the present invention combines a proportional controller, which rapidly reduces heading angle error to ensure the heading angle is directed quickly and accurately toward the target waypoint when the vessel moves toward the target waypoint, with a differential controller, which maintains straightness toward the target waypoint.

[0019] In this case, when using a PD controller that combines a proportional controller and a derivative controller, if autonomous bow angle control of the ship is performed by fixing the controller gains set to match specific operating conditions, there is a risk that control performance will deteriorate under operating conditions other than those specific conditions.

[0020] Accordingly, in a preferred embodiment of the present invention, a ship autonomous navigation control device is implemented using a fuzzy PD controller that includes the features of a conventional PD controller, thereby allowing the controller gains to change flexibly according to the ship's operating state variables, so that the ship's heading angle control can be appropriately performed even under various operating conditions. Hereinafter, the ship autonomous navigation control device and the fuzzy PD controller are used interchangeably.

[0021] FIG. 1 is a diagram defining variables for performing control to maintain a target heading angle in order to effectively reach a target point by utilizing fore-and-aft, side-and-side, and bow motions during the 6-degree-of-freedom motion of a ship according to the present invention.

[0022] In FIG. 1, variables for control to maintain a target heading angle (5) are defined to effectively reach a target point (4) using surge (1), sway (2), and yaw (3) during the 6-degree-of-freedom motion of the ship.

[0023] From the time the ship departs from the starting point (6) until it arrives at the target point (4), effective heading angle control must be performed to minimize the error heading angle (8), which is the difference between the target heading angle (5) and the current heading angle (7).

[0024] Mathematical formula 1 defines the target heading angle (5).

[0025]

[0026] In mathematical formula 1 means the target heading angle (5) for the target point (4), and (12) and (15) represents the forward and backward oscillation coordinates and the left and right oscillation coordinates of the target point (4), respectively. (11) and (14) represents the front-back and back-back oscillation coordinates and the left-right oscillation coordinates.

[0027]

[0028] In mathematical formula 2 This refers to the error angle (8), which is the difference between the target angle of descent (5) and the current angle of descent (7).

[0029] Figure 2 is a block diagram illustrating the structure of a PD controller in a discrete system.

[0030] With reference to FIG. 2, the basic configuration and operation of a PD controller in a discrete system that forms the basis of the present invention will be explained.

[0031] Mathematical Equation 3 is the basic equation of a PD controller in a general continuous-time system.

[0032]

[0033] In mathematical formula 3 (16) means proportional gain and (17) represents the differential gain.

[0034] A continuous-time system PD controller like Equation 3, sampling time (18) When expressed as a discrete system, it is as follows: Equation 4.

[0035]

[0036] In mathematical formula 4 means the number of samples and is a control input, The error The velocity component obtained by differentiating (22), and is speed It means the acceleration component obtained by differentiating (23).

[0037] Mathematical Equation 5 below is the sampling interval one step prior. Error velocity component at (20) expresses.

[0038]

[0039] In mathematical formula 5 (26) is Target value at the i-th sampling (24) and, Current state value at the nth sampling (25) The difference between, Error value at the i-th sampling (19) delayed by one step (28) (i.e., one step earlier) It refers to the error value at the nth sampling.

[0040] Meanwhile, mathematical equation 6 below is the sampling interval one step prior. Error acceleration component at (21) expresses.

[0041]

[0042] In mathematical formula 4 Control input at the i-th sampling interval (27) is Control input at the i-th sampling interval (29) and mathematical formula 5 of Velocity component in the i-th sampling interval (20) proportional gain (16) and sampling time (18) The value multiplied by the mathematical expression 6 Acceleration component in the i-th sampling interval (21) Differential gain (17) and sampling time (18) It can be seen that the result is the sum of the values ​​multiplied by each other.

[0043] The present invention applies a fuzzy control method to the discrete system PD controller described above, thereby allowing the controller gains to change flexibly according to the ship's operating state variables, so that the ship's autonomous navigation controller is not limited to a specific operating range and can appropriately perform ship heading angle control under various operating conditions.

[0044] Hereinafter, with reference to FIGS. 3 and 4, variables and fuzzy membership functions for applying a fuzzy PD controller according to a preferred embodiment of the present invention will be described first.

[0045] Figure 3 shows the error velocity component values, which are the variables of the fuzzy PD controller. (30) and error acceleration component value (36) Each fuzzy membership function (31) and (37) is a drawing that illustrates the drawing.

[0046] Error speed component value (30) fuzzy membership function The types of (31) include RP membership functions (32) that indicate the degree of positive (+) values ​​of speed and RN membership functions (33) that indicate the degree of negative (-) values ​​of speed, and the parameters that determine the shape of the RP membership functions (32) and RN membership functions (33) are It is (35,34). Parameter (35,34) can be set to a relatively large value (e.g., 100 or more) based on the controller designer's experience. Equations 7 and 8 below represent the RP membership function (32) and the RN membership function (33), respectively.

[0047]

[0048]

[0049] As can be seen from the above mathematical formulas 7 and 8, the sum of the two formulas is 1.

[0050] Meanwhile, the error acceleration component value Fuzzy membership function of (36) (37) The types include AP membership functions (38) representing the degree of positive (+) acceleration and AN membership functions (39) representing the degree of negative (-) acceleration, and the parameters determining the shape of the AP membership functions (38) and AN membership functions (39) are (40,41) (40,41) can be set to a relatively large value (e.g., 100 or more) based on the controller designer's experience. Equations 9 and 10 below represent the equations for the AP membership function (38) and the AN membership function (39), respectively.

[0051]

[0052]

[0053] As can be seen from the above mathematical formulas 9 and 10, the sum of the two formulas is 1.

[0054] Figure 4 is a diagram illustrating the time response of a general second-order system for establishing fuzzy logic rules. The activation region of each rule is determined through this time response curve.

[0055] In FIG. 4, the first graph (44) is the reference value System response to (42) (43) shows the curve. The second graph (46) is the reference value (42) and system response (43) The difference value of the error value (45) shows the curve. The third graph (48) shows the error value Error velocity component value obtained by differentiating (45) (47) shows the curve. The last, fourth graph (50) shows the error velocity component value Error acceleration component value obtained by differentiating (47) (49) represents.

[0056] The circular points in each graph are Rule 1 (51), the square points are rule 2 (52), triangle points are rule 3 (53) And the star points are rule 4 (54) indicates. If a rule is defined at each point, it is as follows: Equations 11 to 14.

[0057]

[0058]

[0059]

[0060]

[0061] Figure 5 is a conceptual diagram showing the entire inference process of a discrete fuzzy PD controller.

[0062] Error velocity component with fuzzy PD controller (55) and acceleration component of the error (56) When input, it is used as an argument to the membership function defined for each rule, and the degree of membership of the input variables is calculated. This process is called the fuzzification process (57).

[0063] Among the fuzzified values ​​(degrees of membership) for each rule, a suitability value representing the firing strength (58) of the corresponding rule must be selected. In this process, the AND operator used in the present invention is a minimum operator as shown in the following mathematical formula 15.

[0064]

[0065] In the above mathematical formula 15, the box portion is filled alternately with P(positive) and N(negative), as shown in Fig. 5.

[0066] The conclusion derivation process (Consequent part) (59) of each rule is a process of calculating the conclusion value for the conditions of the corresponding rule. In the present invention, a linear equation as shown in the following mathematical equation 16 is used, and the coefficients multiplied by each input variable become the parameters of the conclusion derivation process for each rule. These parameters of the conclusion derivation process are the main factors determining the performance of the fuzzy PD controller, and if the number of input variables is n and the number of rules is p, the total number of parameters of the conclusion derivation process is It becomes. Here, since there are 2 input variables (r(k-1), a(k-1)) and 4 rules (R1~R4), the number of parameters in the conclusion is This becomes.

[0067]

[0068] Finally, the defuzzification process (60) is the final process of fuzzy inference, and after the previously fuzzified input variables undergo the fuzzy inference process, the defuzzified output value is calculated in the defuzzification process. Defuzzification was performed using the central average method as shown in Equation 17 below.

[0069]

[0070] The final output value that is defuzzified (73) is the current sampling control value and previous sampling control value It is the difference, and is equal to the following mathematical formula 18.

[0071]

[0072] FIG. 6 is a diagram illustrating the configuration of a ship autonomous navigation control device according to a preferred embodiment of the present invention.

[0073] Referring to FIG. 6, a ship autonomous navigation control device according to a preferred embodiment of the present invention includes a processor (620) and a memory (610).

[0074] A memory (610) according to a preferred embodiment of the present invention may store instructions executable by a processor (620) and programs executed by the processor (620), and may also store data input to the processor (620) and data generated and output by the processor (620).

[0075] A processor (620) according to a preferred embodiment of the present invention executes instructions stored in a memory (610) to perform each step of the ship autonomous navigation control processes described above with reference to FIGS. 1 to 5 and the ship autonomous navigation control method described later with reference to FIGS. 7 and 8, thereby controlling reference values and current status value It receives input and outputs the control signal u(k) to an actuator that directly controls the direction of the vessel. At this time, the memory (610) may be replaced and operated by a web storage or cloud server that performs the function of a storage medium on the internet.

[0076] FIG. 7 is a diagram showing the functions performed by the processor (620) as function blocks when instructions stored in memory (610) are executed by the processor (620) to implement a ship autonomous navigation control device, and FIG. 8 is a flowchart explaining a ship autonomous navigation control method performed by the ship autonomous navigation control device according to a preferred embodiment of the present invention.

[0077] With reference to FIGS. 7 and 8, the function of a ship autonomous navigation control device and a ship autonomous navigation control method according to a preferred embodiment of the present invention will be described.

[0078] First, the ship autonomous navigation control device implemented by executing instructions stored in memory (610) in a processor (620) includes an error calculator, an adder, a first delayer (65-1), a second delayer (65-2), a first differentiator (67-1), a second differentiator (67-2), and a discrete fuzzy control module (69), and the discrete fuzzy control module (69) includes a fuzzying unit (70), a fuzzy inference unit (71), and a defuzzification unit (72).

[0079] First, the control reference value of the k-th sampling interval using the error calculator (62) and k-th sampling interval current state value (63) is input, and the error calculator is the control reference value of the k-th sampling interval (62) k-th sampling interval current state value Subtract (63) to get the error (64) is generated and output to the first delay unit (65-1) (S810).

[0080] The first delay unit (65-1) is an error (64) is delayed by one step (one sampling time interval) and output to the first differentiator (67-1) (S820).

[0081] The first differentiator (67-1) is the previous sampling interval ( Error calculated in ) (61) is input and differentiation is performed to obtain the error rate (66) is generated and output to the second differentiator (67-2) and the fuzzification unit (70) of the discrete fuzzy control module (69) (S830).

[0082] The second differentiator (67-2) is the error rate (66) is input and differentiation is performed to obtain the error acceleration (68) is generated and output to the fuzzy control module (69) of the discrete fuzzy control module (70) (S840).

[0083] The discrete fuzzy control module (69) is an error Perform fuzzification and fuzzy inference processes to minimize the control signal increment Generates and outputs to the adder (S850).

[0084] Here, the fuzzification unit (70) is the error speed (66) and error acceleration (68) is received as input, and a fuzzying process is performed as described with reference to reference numeral 57 of FIG. 5. The fuzzy inference unit (71) performs a fuzzy inference process by selecting a suitability that represents the firing strength (58) for each fuzzy rule as described with reference numerals 57 and 58 of FIG. 5, and the defuzzying unit (72) performs a defuzzying process as described with reference to reference numeral 59 of FIG. 5 and Equation 17, and the output value is the control signal increment (73) is output to the adder (S850).

[0085] The adder outputs to the actuator in the previous sampling interval and the actuator control signal delayed in the second delayer (65-2) The control signal increment input from the discrete fuzzy control module (69) (73) sum up the actuator control signal Output to the actuator (S860).

[0086] Meanwhile, the performance of the above-described ship autonomous navigation control device depends on how the parameters included in the fuzzy rules of Equations 11 to 14 are set. Below, with reference to FIGS. 9 and FIGS. 10, a method for setting the parameters included in the fuzzy rules to achieve optimal performance will be explained.

[0087] In the present invention, a linear equation as shown in Equation 16 above is used, and the coefficients multiplied by each input variable become the parameters in the process of deriving the conclusion of each rule. These parameters in the conclusion derivation process are the main factors determining the performance of the fuzzy PD controller, and if the number of input variables is n and the number of rules is p, the total number of parameters in the conclusion process is There are 2 input variables (r(k-1), a(k-1)) and 4 rules (R1~R4), so the number of parameters in the conclusion part is 2*4=8.

[0088] In the above mathematical formula 16 Igo If so, the above mathematical formula 16 is equal to the following mathematical formula 19.

[0089]

[0090] Under the same conditions as Equation 19, Equation 17 is equal to the following Equation 20.

[0091]

[0092] From the above mathematical formula 18 Therefore, the control value at the current k-th sampling is equal to the following mathematical equation 21.

[0093]

[0094] Here, comparing Equation 21 with the above Equation 4, since the equations are identical as the fundamental equations of the PD controller in a discrete system, Equation 21 is of mathematical formula 4 It is the same as, and of mathematical formula 21 is of mathematical formula 4 It is the same as.

[0095] From this, it can be seen that the discrete PD controller is a special case of the discrete fuzzy PD controller, and that the discrete fuzzy PD controller can possess at least the performance of the discrete PD controller, and that by adjusting the parameters of the discrete fuzzy PD controller, it can demonstrate superior performance compared to the discrete PD controller.

[0096] Accordingly, a preferred embodiment of the present invention sets initial values ​​of the parameters of the fuzzy rule using the control gains of the discrete PD controller of Equation 4 ( , After that, the parameter values ​​of the final discrete fuzzy PD controller are updated through an optimization algorithm.

[0097] FIGS. 9 and 10 are drawings illustrating the process of determining parameters through an optimization algorithm according to a preferred embodiment of the present invention.

[0098] First, when the initial parameter values ​​(77) of the fuzzy PD controller are input into the optimization algorithm (78), M-1 objects are derived (80) from the initial parameters (77), and a total of M objects are generated by combining the initial parameters (77).

[0099] In the first generation (81), M objects compete with each other (82) to find the object (83) with the highest fitness score. The most fit object becomes the first object (84) of the next generation, and from this first object (84), M-1 objects are derived (85) again, and the optimal parameters of the previous generation are combined to create a total of M objects.

[0100] In this way, M objects compete with each other (87) in the nth generation (86) to find the object (88) with the highest fitness score in the nth generation. In this way, the final optimal parameter (79) is found through an optimization process that is repeated N times (89) in which M objects compete in each generation over a total of N generations.

[0101] FIG. 10 describes a fitness determination method for selecting the optimal object by having objects compete (82, 87) in the optimization algorithm (78) of FIG. 9. The fitness function (90) for finding the optimal object in the competition among objects is as shown in Equation 22 below.

[0102]

[0103] Mathematical formula 22 gives a high score to an object that produces a response (96) with reduced overshoot (93), rise time (94), and vibration (95) compared to the existing step response (92) in the step response based on error (91).

[0104] In mathematical formula 22, the squared mean of the error (97) is entered to compare the total error, such as overshoot, and the error value (98) at a specific time when the error is reduced is entered to compare the speed of the system response, and the squared mean of the error derivative (99) is entered to compare the degree of system vibration.

[0105] By doing this, objects with larger values ​​in Equation 22 are more likely to be evaluated as superior objects. In Equation 22 It refers to a weight value that adjusts whether to focus on reducing the error or reducing the rate of change of the error.

[0106] FIG. 11 is a diagram illustrating the result of performing heading angle control to minimize the sailing distance while sequentially following waypoints from a starting point to a destination point for a catamaran-type vessel according to a preferred embodiment of the present invention.

[0107] In the example illustrated in FIG. 11, the result of controlling a catamaran-type vessel with two propellers installed on the right (starboard) and left (port) at the stern to minimize the sailing distance while following the waypoints (103, 104, 105, 106) in sequence from the starting point (102) to the destination point (107) is shown.

[0108] The horizontal axis of FIG. 11 represents sway (101) and the vertical axis represents surge (100). In the graph of FIG. 11, the dotted line (108) is the result of performing path estimation control with a discrete PD controller, and this is the parameters of the discrete fuzzy PD controller and It is the same as the result set identically.

[0109] In the graph of FIG. 11, the solid line (109) represents the error velocity component value coefficients in the discrete fuzzy PD controller according to the present invention. The parameters and error acceleration component value coefficients This is the result of performing path estimation control based on the optimal parameter values ​​found by the optimization algorithm.

[0110] As can be seen from the graph in Figure 11, the path estimation control performance of the discrete fuzzy PD controller is superior to that of the discrete PD controller or the discrete fuzzy PD controller in which all parameters are identical. This is because, unlike the conventional PD controller which operates only with control gains set at a specific operating point and suffers a loss in control performance under other operating conditions, the discrete fuzzy PD controller demonstrates superior performance in path estimation control by flexibly changing control parameters according to the ship's operating state variables and appropriately controlling the ship's heading angle under various operating conditions.

[0111] The ship autonomous navigation control method according to the preferred embodiment of the present invention described so far can be implemented as a computer program stored in a non-transient storage medium, implemented as computer-executable instructions.

[0112] Storage media include all types of recording devices in which data that can be read by a computer system is stored. Examples of computer-readable storage media include ROM, RAM, CD-ROM, magnetic tape, floppy disk, and optical data storage devices. In addition, computer-readable storage media are distributed across networked computer systems, allowing computer-readable code to be stored and executed in a distributed manner.

[0113] The present invention has been described above with reference to its preferred embodiments. Those skilled in the art will understand that the present invention may be embodied in modified forms without departing from the essential characteristics of the invention. Therefore, the disclosed embodiments should be considered in an illustrative rather than a restrictive sense. The scope of the invention is defined by the claims, not by the foregoing description, and all variations within the scope of the claims should be interpreted as being included in the invention. Explanation of the symbols

[0114] 65-1: 1st delay period 65-2: 2nd delay period 67-1: First differentiator 67-2: Second differentiator 69: Discrete Fuzzy Control Module 70: Fuzzy Unit 71: Fuzzy Inference Unit 72: De-fuzzy Unit 610:Memory 620:Processor

Claims

Claim 1 A method for autonomous ship control performed in an autonomous ship control device, comprising: (a) a control reference value of the k-th sampling interval and k-th sampling interval current state value The error between them Step of generating; (b) the error Delay by one sampling time interval, and the previous sampling interval ( Error calculated in ) Error rate by performing differentiation with respect to Step of generating; (c) the error rate Error acceleration by performing differentiation with respect to Step of generating; (d) the error speed and the above error acceleration Using, error Perform fuzzification and fuzzy inference processes to minimize the control signal increment A step of generating; and (e) a control signal output to the actuator in the previous sampling interval The above control signal increment Summing up the control signal It includes the step of outputting to an actuator, and in step (d), a mathematical formula Control signal increment according to is generated, and the parameters of the above mathematical formula is the proportional gain of the discrete PD controller for the vessel An initial value is set using, and the parameter of the above mathematical formula is the derivative gain of the discrete PD controller for the vessel A ship autonomous navigation control method characterized by being updated by applying an optimization algorithm after an initial value is set. Claim 2 In claim 1, the above step (d) is (d1) the error rate and the above error acceleration A step of calculating the degree of membership of an input variable using as an argument to a membership function defined for each fuzzy rule; (d2) a step of selecting a degree of fit representing the firing strength of the corresponding rule among the degrees of membership for each fuzzy rule; (d3) the error rate for each fuzzy rule and the above error acceleration A step of generating a conclusion value by applying pre-set parameters to; and (d4) performing defuscation using the conclusion value for each fuzzy rule and the suitability to the control signal increment A ship autonomous navigation control method characterized by including a step of generating Claim 3 delete Claim 4 A computer program stored in a non-transient storage medium and executed on a computer including a processor to perform the ship autonomous navigation control method of claim 1 or 2. Claim 5 A ship autonomous navigation control device comprising a processor and a memory storing predetermined instructions, wherein the processor that executes the instructions stored in the memory (a) a control reference value of the k-th sampling interval and k-th sampling interval current state value The error between them Step of generating; (b) the error Delay by one sampling time interval, and the previous sampling interval ( Error calculated in ) Error rate by performing differentiation with respect to Step of generating; (c) the error rate Error acceleration by performing differentiation with respect to Step of generating; (d) the error speed and the above error acceleration Using, error Perform fuzzification and fuzzy inference processes to minimize the control signal increment A step of generating; and (e) a control signal output to the actuator in the previous sampling interval The above control signal increment Summing up the control signal Performing a ship autonomous navigation control method by performing a step of outputting to an actuator, wherein in step (d), mathematical formula Control signal increment according to is generated, and the parameters of the above mathematical formula is the proportional gain of the discrete PD controller for the vessel An initial value is set using, and the parameter of the above mathematical formula is the derivative gain of the discrete PD controller for the vessel A ship autonomous navigation control device characterized by being updated by applying an optimization algorithm after an initial value is set. Claim 6 In claim 5, the above step (d) is (d1) the error rate and the above error acceleration A step of calculating the degree of membership of an input variable using as an argument to a membership function defined for each fuzzy rule; (d2) a step of selecting a degree of fit representing the firing strength of the corresponding rule among the degrees of membership for each fuzzy rule; (d3) the error rate for each fuzzy rule and the above error acceleration A step of generating a conclusion value by applying pre-set parameters to; and (d4) performing defuscation using the conclusion value for each fuzzy rule and the suitability to the control signal increment A ship autonomous navigation control device characterized by including a step of generating Claim 7 delete

Images