An information reminder method and system based on ship intelligent robot

Abstract

The present application proposes an information reminder method and system based on a ship intelligent robot, which belongs to the field of intelligent ship technology. First, the first ship risk information is generated based on the historical information of the first ship, and then the first monitoring frequency of each component of the ship is obtained, and then the second ship risk information is generated based on the actual navigation information of the ship and the navigation information of the surrounding ships. Then, the captain's basic information and driving experience information are input into the first reminder threshold determination model to determine the first reminder threshold of each component of the ship, and finally the first reminder information is output according to the first monitoring frequency and the first reminder threshold, and the second reminder information is output with the second ship risk information. The present application comprehensively analyzes the basic information of the ship, the real-time navigation information, and the real-time navigation information of the reference ship to determine the potential risks, and determines the reminder threshold according to the actual situation of the captain, so that the generated reminder information is more accurate and timely.

Description

Technical Field

[0001] The present invention belongs to the field of intelligent ships, and in particular relates to an information reminder method and system based on a ship intelligent robot. Background Art

[0002] With the continuous development of artificial intelligence technology, intelligent ship technology has become the mainstream research method of automatic navigation. However, the existing intelligent ship technology can only remind fault information and navigation information based on historical experience values. The generated reminder information cannot take into account the ship components themselves and navigation information such as weather, and cannot customize personalized reminder information for drivers, resulting in the generation of reminder information being too mechanical. Summary of the invention

[0003] The purpose of the present invention is to provide an information reminder method and system based on a ship intelligent robot to solve the technical problems in the prior art that the intelligent ship reminder information is not comprehensive in consideration and the information generation is more mechanized.

[0004] An information reminder method based on a ship intelligent robot, characterized in that the method comprises:

[0005] S1: determining first ship risk information according to first historical information of the first ship;

[0006] The S1 comprises the following sub-steps:

[0007] S11: generating a first ship portrait according to first historical information of the first ship;

[0008] S12: Acquire A second ship portraits whose similarity with the first ship portrait exceeds a first threshold;

[0009] S13: extracting B first ship fault information from A second ship portraits;

[0010] S14: performing cluster analysis on B pieces of the first ship fault information to obtain C first cluster centers, determining C first risk probabilities according to a first quantity of the first ship fault information corresponding to each of the first cluster centers, and determining the first ship risk information according to the C first cluster centers and the first risk probabilities;

[0011] A, B and C are all natural numbers greater than 1, where A <B;

[0012] S2: adjusting the information monitoring frequency according to the first ship risk information, thereby determining the first information monitoring frequency corresponding to each first component of the first ship;

[0013] S3: determining first route risk information and second route risk information according to the first route information and the first time information, and determining second ship risk information according to the first route risk information and the second route risk information;

[0014] S4: determining a first captain portrait according to first basic information and first driving experience information of the captain of the first ship, and inputting the first captain portrait into a first reminder threshold determination model, thereby determining first reminder thresholds corresponding to various components of the first ship;

[0015] S5: Acquire first reminder information according to the first information monitoring frequency and the first reminder threshold, and acquire second reminder information according to the second ship risk information.

[0016] Preferably, the first historical information includes a first type, a first driving age, a first historical maintenance record, a first driving mileage and a first historical fault record of the first ship.

[0017] Preferably, the first type describes the first ship from four dimensions: purpose type, structure type, size type and navigation area type, so that the first type is expressed in the form of a four-dimensional vector.

[0018] Preferably, S2 comprises the following sub-steps:

[0019] S21: determining an initial monitoring frequency of each of the first components according to the first type of the first ship;

[0020] S22: acquiring first adjustment range information corresponding to each of the first components according to the first ship risk information;

[0021] S23: Determine the first information monitoring frequency of each of the first components according to the initial monitoring frequency and the first adjustment amplitude information.

[0022] Preferably, the S22 comprises the following sub-steps:

[0023] S221: Decomposing the first ship risk information to obtain C first cluster centers and corresponding C first risk probabilities;

[0024] S222: determining the second component corresponding to each of the first cluster centers according to the correspondence between the ship fault information and the ship component information, wherein the C second components are a subset of all the first components on the first ship;

[0025] S223: Normalize the C first risk probabilities to obtain first adjustment range information of C second components.

[0026] Preferably, S3 includes the following sub-steps:

[0027] S31: determining first weather information and first waterway information according to the first route information and the first time information;

[0028] S32: inputting the first weather information and the first waterway information into a first route risk information prediction model to obtain the first route risk information;

[0029] S33: taking the current position of the first ship as the first position, obtaining information of D second ships within a first preset distance range, and obtaining second route risk information according to the D second ship information;

[0030] S34: determining second ship risk information according to the first route risk information and the second route risk information;

[0031] The D is a natural number.

[0032] Preferably, S5 comprises the following sub-steps:

[0033] S51: Acquire first monitoring information of each of the first components of the first ship according to the first information monitoring frequency;

[0034] S52: acquiring the first reminder threshold corresponding to the first component, generating the first reminder information and issuing a reminder according to a magnitude relationship between the first reminder threshold and the first monitoring information;

[0035] S53: Continuously reminding the risk type in the second ship risk information in the form of second reminder information.

[0036] The present application also proposes an information reminder system based on a ship intelligent robot, which is used to implement the above-mentioned information reminder method based on a ship intelligent robot.

[0037] The present application proposes an information reminder method and system based on a ship intelligent robot. First, the first ship risk information is generated based on the historical information of the first ship, and then the first monitoring frequency of each component of the ship is obtained. Then, the second ship risk information is generated based on the actual navigation information of the ship and the navigation information of surrounding ships. Then, the captain's basic information and driving experience information are input into the first reminder threshold determination model to determine the first reminder threshold of each component of the ship, and finally the first reminder information is output based on the first monitoring frequency and the first reminder threshold, and the second reminder information is output based on the second ship risk information. The present application comprehensively analyzes the basic information of the ship, the real-time navigation information, and the real-time navigation information of the reference ship to determine the potential risks, and determines the reminder threshold according to the actual situation of the captain, so that the generated reminder information is more accurate and timely. BRIEF DESCRIPTION OF THE DRAWINGS

[0038] In order to more clearly illustrate the implementation methods of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for the implementation methods or the description of the prior art. Obviously, the drawings in the following description are only exemplary, and for ordinary technicians in this field, other implementation drawings can be derived from the provided drawings without creative work.

[0039] Figure 1 The present invention is an execution flow chart of an information reminder method based on a ship intelligent robot. DETAILED DESCRIPTION

[0040] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0041] The present invention will be described in detail below in conjunction with the accompanying drawings and specific embodiments, wherein the illustrative embodiments and descriptions are only used to explain the present invention but are not intended to limit the present invention.

[0042] The following is a detailed description of an information reminder method and system based on a ship intelligent robot of the present invention.

[0043] This embodiment proposes an information reminder method based on a ship intelligent robot. The method flow is as follows: Figure 1 As shown, the specific steps include:

[0044] S1: Determine first ship risk information according to first historical information of the first ship.

[0045] The first ship refers to a ship for which risk information reminder is to be issued.

[0046] The S1 comprises the following sub-steps:

[0047] S11: Generate a first ship portrait according to the first historical information of the first ship.

[0048] The first historical information includes information such as the first type, the first driving years, the first historical maintenance record, the first driving mileage and the first historical fault record of the first ship.

[0049] The first type includes purpose type, structure type, size type and navigation area type. The purpose type includes commercial ships, passenger ships, fishing boats and warships, the structure type includes flat-bottomed boats, open cargo ships, container ships, cruise ships, etc., the size type includes large ships, medium-sized ships and small ships, etc., and the navigation area type includes offshore ships and ocean-going ships. The first type describes the first ship from four dimensions: purpose type, structure type, size type and navigation area type, so as to express the first type in the form of a four-dimensional vector.

[0050] The first driving years and the first driving mileage refer to the total years and total mileage of the first ship during its service.

[0051] The first historical maintenance record includes information such as the maintenance cycle and maintenance items of the first ship.

[0052] The first historical fault record includes information such as the number of faults, fault types, and maintenance status of the first ship during its service.

[0053] The first ship portrait is a multidimensional vector formed by the first type, the first driving years, the first historical maintenance record, the first mileage and the first historical fault record. The multidimensional vector can accurately and comprehensively describe the navigation status of the first ship.

[0054] S12: Acquire A second ship portraits whose similarity with the first ship portrait exceeds a first threshold.

[0055] In this step, it is necessary to maintain the first ship set in advance. The first ship set includes multiple historical ship information. Each historical ship information includes information obtained by describing the historical ship from five dimensions, such as ship type, age, historical maintenance record, mileage and historical fault record. Ship portraits are generated for all ships in the first ship set based on the historical ship information.

[0056] The first ship portrait is compared with all the ship portraits in the first ship set one by one for similarity, so as to determine A second ship portraits by taking the similarity exceeding a first threshold as a judgment criterion. Specifically, the cosine similarity can be used to calculate the similarity.

[0057] S13: Extracting B first ship fault information from A second ship portraits.

[0058] Each of the second ship portraits includes historical fault record information. After all the historical fault record information is extracted, B first ship fault information can be obtained.

[0059] S14: Perform cluster analysis on B pieces of the first ship fault information to obtain C first cluster centers, determine C first risk probabilities according to the first quantity of the first ship fault information corresponding to each of the first cluster centers, and determine the first ship risk information according to the C first cluster centers and the first risk probabilities.

[0060] In this step, each piece of the first ship fault information needs to be preprocessed to make a consistent representation.

[0061] Next, cluster analysis is performed on the B pre-processed first ship fault information to obtain C first cluster centers, and the first quantity of the corresponding first ship fault information is obtained for each first cluster center, and the C first quantities are normalized and sorted to determine C first risk probabilities. For example, if three first cluster centers are obtained through clustering, each of the first cluster centers corresponds to a type of ship fault, and the quantities of first ship fault information corresponding to the three first cluster centers are 1, 2, and 3, respectively, then the normalized first risk probabilities of the three first cluster centers are 33%, 66%, and 100%, respectively.

[0062] The first ship risk information is expressed as a 2×C dimensional vector consisting of C first cluster centers and corresponding C first risk probabilities.

[0063] A, B and C are all natural numbers greater than 1, where A <B。

[0064] S2: adjusting the information monitoring frequency according to the first ship risk information, thereby determining the first information monitoring frequency corresponding to each first component of the first ship.

[0065] The first components refer to the main components of the first ship, such as an engine, a navigator, etc. During the travel of the first ship, it is necessary to obtain monitoring information corresponding to each of the first components at a certain time interval.

[0066] The S2 comprises the following sub-steps:

[0067] S21: Determine an initial monitoring frequency of each of the first components according to the first type of the first ship.

[0068] Different types of ships usually have different monitoring frequency requirements for various components on the ship. Therefore, in this step, it is first necessary to determine the initial monitoring frequency of each component based on the first type of the first ship, where the initial monitoring frequency is obtained by empirical values.

[0069] S22: Acquire first adjustment range information corresponding to each of the first components according to the first ship risk information.

[0070] The S22 specifically includes the following sub-steps:

[0071] S221: Decompose the first ship risk information to obtain C first cluster centers and corresponding C first risk probabilities.

[0072] S222: Determine the second component corresponding to each of the first cluster centers according to the correspondence between the ship fault information and the ship component information, and the C second components are a subset of all the first components on the first ship.

[0073] Since the correspondence between the ship fault information and the ship component information can be obtained from historical experience, the second component causing the ship fault information can be found according to the first cluster center representing the ship fault information.

[0074] S223: Normalize the C first risk probabilities to obtain first adjustment range information of C second components.

[0075] The first adjustment range information is positively correlated with the first risk probability, that is, the greater the first risk probability is, the greater the first adjustment range information is.

[0076] For the first component that is not included in the C second components, its first adjustment amplitude information is set to 0, that is, there is no corresponding fault information at present, so there is no need to adjust its initial monitoring frequency.

[0077] S23: Determine the first information monitoring frequency of each of the first components according to the initial monitoring frequency and the first adjustment amplitude information.

[0078] Among them, each of the first information monitoring frequencies is obtained by adjusting the corresponding initial monitoring frequency according to the first adjustment amplitude information.

[0079] The overall adjustment rule is: for the first component with a greater probability of failure, the adjustment range of its information monitoring frequency is greater, and vice versa. For the first component with no probability of failure, the adjustment probability is 0.

[0080] S3: Determine first route risk information and second route risk information according to the first route information and the first time information, and determine second ship risk information according to the first route risk information and the second route risk information.

[0081] The first route information refers to the route information on which the first ship is to travel, and the first time information refers to the time information corresponding to when the first ship travels through the first route information.

[0082] The S3 comprises the following sub-steps:

[0083] S31: Determine first weather information and first waterway information according to the first route information and the first time information.

[0084] Since the navigation risk of a ship is constrained by weather information and channel information, the channel on which the first ship is to travel can be obtained based on the first route information, and the weather conditions that the first ship will experience while traveling on the first channel can be obtained based on the first time information.

[0085] S32: Inputting the first weather information and the first waterway information into a first route risk information prediction model to obtain the first route risk information.

[0086] The first route risk information includes specific risk conditions and probability of occurrence.

[0087] The first route risk information prediction model is obtained based on CNN model training, specifically by obtaining sample data based on historical navigation data, taking corresponding weather information and waterway information as input, and taking specific risk situations and probability of occurrence as output, to train the CNN model.

[0088] S33: Taking the current position of the first ship as the first position, obtaining D pieces of second ship information within a first preset distance range, and obtaining second route risk information based on the D pieces of second ship information.

[0089] In this step, the current position of the first ship is used as the center of the circle, D second ships within the first preset distance range are obtained, and the risk information encountered by the D second ships during the actual navigation process is used as the second ship risk information. Specifically, the second ship risk information may include the extreme bad weather or other abnormal conditions actually encountered during the navigation of the second ship.

[0090] According to the relative position relationship between the first ship and the second ship, the information of D second ships is integrated to obtain the second route risk information.

[0091] The D is a natural number.

[0092] S34: Determine second ship risk information according to the first route risk information and the second route risk information.

[0093] In this step, the second ship risk information is determined based on the predicted first route risk information and the second ship risk information actually occurring during the navigation of other ships.

[0094] Specifically, the content of the first route risk information is modified and corrected according to the main risk information reflected in the second route risk information. For example, if the predicted probability of the risk type of iceberg floating objects in the first route risk information is very low, but the second ship within the first preset distance actually encounters iceberg floating objects, then the risk type and the corresponding occurrence probability in the first route risk information need to be corrected.

[0095] S4: determining a first captain's portrait according to the first basic information and the first driving experience information of the captain of the first ship, and inputting the first captain's portrait into a first reminder threshold determination model to determine first reminder thresholds corresponding to the components of the first ship.

[0096] Affected by personal basic information and driving experience information, different captains have different abilities to discover potential problems from the same warning information. Therefore, in this step, it is necessary to predict the warning threshold of the captain of the first ship.

[0097] The first basic information includes gender, education background, personality, etc., and the first driving experience information includes driving years, driving mileage, historical driving routes, etc.

[0098] The first reminder threshold determination model is obtained based on CNN model training. Based on historical navigation data as sample data, the portrait composed of basic information and driving experience information in the sample data is used as input, and the reminder thresholds of different ship components are used as output to train the CNN model, thereby obtaining the first reminder threshold determination model.

[0099] In this step, the first reminder threshold corresponding to each component of the first ship can be determined by inputting the first captain portrait into the first reminder threshold determination model.

[0100] S5: Acquire first reminder information according to the first information monitoring frequency and the first reminder threshold, and acquire second reminder information according to the second ship risk information.

[0101] The S5 comprises the following sub-steps:

[0102] S51: Acquire first monitoring information of each of the first components of the first ship according to the first information monitoring frequency.

[0103] S52: Acquire the first reminder threshold corresponding to the first component, generate the first reminder information and issue a reminder according to the magnitude relationship between the first reminder threshold and the first monitoring information.

[0104] In this step, for each of the first components, the magnitude relationship between the corresponding first reminder threshold and the first monitoring information is compared, so that when the first monitoring information exceeds the first reminder threshold, the first reminder information is generated and a reminder is given.

[0105] S53: Continuously reminding the risk type in the second ship risk information in the form of second reminder information.

[0106] The present application also proposes an information reminder system based on a ship intelligent robot, which is used to execute the above-mentioned information reminder method based on a ship intelligent robot.

[0107] The present application proposes an information reminder method and system based on a ship intelligent robot. First, the first ship risk information is generated based on the historical information of the first ship, and then the first monitoring frequency of each component of the ship is obtained. Then, the second ship risk information is generated based on the actual navigation information of the ship and the navigation information of surrounding ships. Then, the captain's basic information and driving experience information are input into the first reminder threshold determination model to determine the first reminder threshold of each component of the ship, and finally the first reminder information is output based on the first monitoring frequency and the first reminder threshold, and the second reminder information is output based on the second ship risk information. The present application comprehensively analyzes the basic information of the ship, the real-time navigation information, and the real-time navigation information of the reference ship to determine the potential risks, and determines the reminder threshold according to the actual situation of the captain, so that the generated reminder information is more accurate and timely.

[0108] The above description is only a preferred embodiment of the present invention, so all equivalent changes or modifications made according to the structure, characteristics and principles described in the scope of the patent application of the present invention are included in the scope of the patent application of the present invention.

Claims

1. An information reminder method based on a ship intelligent robot, characterized in that: The method includes: S1: determining first ship risk information according to first historical information of the first ship; The S1 comprises the following sub-steps: S11: generating a first ship portrait according to first historical information of the first ship; S12: Acquire A second ship portraits whose similarity with the first ship portrait exceeds a first threshold; S13: extracting B first ship fault information from A second ship portraits; S14: performing cluster analysis on B pieces of the first ship fault information to obtain C first cluster centers, determining C first risk probabilities according to a first quantity of the first ship fault information corresponding to each of the first cluster centers, and determining the first ship risk information according to the C first cluster centers and the first risk probabilities; A, B and C are all natural numbers greater than 1, where A <B; S2: adjusting the information monitoring frequency according to the first ship risk information, thereby determining the first information monitoring frequency corresponding to each first component of the first ship; S3: determining first route risk information and second route risk information according to the first route information and the first time information, and determining second ship risk information according to the first route risk information and the second route risk information; S4: determining a first captain portrait according to first basic information and first driving experience information of the captain of the first ship, and inputting the first captain portrait into a first reminder threshold determination model, thereby determining first reminder thresholds corresponding to various components of the first ship; S5: Acquire first reminder information according to the first information monitoring frequency and the first reminder threshold, and acquire second reminder information according to the second ship risk information.

2. The information reminder method based on a ship intelligent robot according to claim 1 is characterized in that: The first historical information includes a first type, a first driving age, a first historical maintenance record, a first driving mileage, and a first historical fault record of the first ship.

3. The information reminder method based on a ship intelligent robot according to claim 2 is characterized in that: The first type describes the first ship from four dimensions: purpose type, structure type, size type and navigation area type, so that the first type is expressed in the form of a four-dimensional vector.

4. The information reminder method based on a ship intelligent robot according to claim 3 is characterized in that: The S2 comprises the following sub-steps: S21: determining an initial monitoring frequency of each of the first components according to the first type of the first ship; S22: acquiring first adjustment range information corresponding to each of the first components according to the first ship risk information; S23: Determine the first information monitoring frequency of each of the first components according to the initial monitoring frequency and the first adjustment amplitude information.

5. The information reminder method based on a ship intelligent robot according to claim 4 is characterized in that: The S22 comprises the following sub-steps: S221: Decomposing the first ship risk information to obtain C first cluster centers and corresponding C first risk probabilities; S222: determining the second component corresponding to each of the first cluster centers according to the correspondence between the ship fault information and the ship component information, wherein the C second components are a subset of all the first components on the first ship; S223: Normalize the C first risk probabilities to obtain first adjustment range information of C second components.

6. The information reminder method based on a ship intelligent robot according to claim 5 is characterized in that: The S3 comprises the following sub-steps: S31: determining first weather information and first waterway information according to the first route information and the first time information; S32: inputting the first weather information and the first waterway information into a first route risk information prediction model to obtain the first route risk information; S33: taking the current position of the first ship as the first position, obtaining information of D second ships within a first preset distance range, and obtaining second route risk information according to the D second ship information; S34: determining second ship risk information according to the first route risk information and the second route risk information; The D is a natural number.

7. The information reminder method based on a ship intelligent robot according to claim 6 is characterized in that: The S5 comprises the following sub-steps: S51: Acquire first monitoring information of each of the first components of the first ship according to the first information monitoring frequency; S52: acquiring the first reminder threshold corresponding to the first component, generating the first reminder information and issuing a reminder according to the magnitude relationship between the first reminder threshold and the first monitoring information; S53: Continuously reminding the risk type in the second ship risk information in the form of second reminder information.

8. An information reminder system based on a ship intelligent robot, used to implement an information reminder method based on a ship intelligent robot as described in any one of claims 1 to 7.

Images