Seaman evaluation method based on multi-source data
Through the crew evaluation method of multi-source data, combined with high-definition cameras and automatic identification systems, the multi-dimensional information of crew members is obtained, which solves the subjectivity and timeliness of traditional evaluation methods, and realizes the refined management of crew members and safe navigation guarantees.
Patent Information
- Application Number
- CN202410960097.4
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-07-17
- Publication Date
- 2024-11-22
AI Technical Summary
The prior art cannot meet the accuracy and timeliness of crew evaluation, and cannot achieve refined management. The traditional evaluation methods have subjectivity and limitations, and cannot accurately describe the crew's performance and navigation area information.
Using a crew evaluation method based on multi-source data, the crew’s basic information, navigation experience, traditional evaluation information and risk driving behavior information are obtained, and the crew is evaluated in different dimensions (postability, navigation experience and risk) and the crew is generated. Portrait, using high-definition cameras, CCTV systems and automatic identification systems to analyze crew members’ risk driving behavior in real time.
It realizes all-round, multi-dimensional and quantitative evaluation of crew members, improves the accuracy and timeliness of evaluation, and can actively discover potential dangerous driving behaviors and comprehensive quality problems, ensuring the safe navigation of the ship.
Smart Images

Figure CN119005677A_ABST
Abstract
Description
Technical Field
[0001] The invention belongs to the technical field of shipping safety, and in particular relates to a crew assessment method based on multi-source data. Background Art
[0002] According to statistics, by the end of 2020, there were 1,716,866 registered seafarers in my country, a year-on-year increase of 3.5%. There were 808,183 seafarers, a year-on-year increase of 3.0%; and 908,683 inland waterway ship crews, a year-on-year increase of 3.9%. In 2020, there were 377,638 seafarers with sea service qualifications, accounting for 46.7% of the total number of seafarers. The number of registered seafarers in my country has increased year by year. In 2020, the national seafarer education and training institutions enrolled a total of 19,184 people in navigation majors, including 10,451 in driving majors and 8,733 in engine (including electronic and electrical) majors, including 5,905 with bachelor's degrees, 10,758 with junior college (higher vocational) degrees, 2,237 with secondary technical (secondary vocational) degrees, and 284 with non-navigational engineering degrees. The proportion of highly educated people enrolled in seafarer education and training institutions has gradually increased, indicating that the overall educational level of registered seafarers in my country has gradually improved. The Rules for the Competency Examination and Certification of Seafarers of the People's Republic of China, issued by the Ministry of Transport on December 27, 2011, stipulate that seafarers must pass the competency examination and assessment before they can do onboard training. However, seafarer education, training and assessment are mainly completed by maritime schools and training institutions. The seafarer competency examination mode is carried out in the order of examination-assessment-onboard training. This mode makes it difficult to guarantee the practical skills training of seafarers and fails to meet the needs of high-quality seafarer training. From November 1, 2020, the new Rules for the Competency Examination and Certification of Seafarers of the People's Republic of China will come into effect. The new version of the rules changes the provision that seafarers must pass all theoretical examinations and assessments before they can do onboard training. It only requires that seafarers must pass theoretical examinations before onboard training, which increases the practical opportunities for seafarers on board and improves their competency. In the training of seafarers' practical operation ability, a large number of education and training methods combining real equipment and simulator equipment are adopted, requiring consistency between theoretical teaching and actual ship teaching. In the process of the instructor linking the navigation theory teaching with practical experience, the seafarers receiving training can unconsciously raise the practical operation to the level of theory. At the same time, the rapid development of electronic navigation theory has greatly enriched the content and training forms of navigation teaching, and promoted the transformation of navigation education models, evaluation methods and concepts.
[0003] The refined management of crew training and assessment represents my country's gradual transformation from a country with a large number of crew members to a country with strong crew quality. Crew members are the main influencing factors for the safe navigation of ships, and the quality of crew members is the main determining factor for the safe navigation of ships. Crew training and competency evaluation represent the initial comprehensive quality of crew members. In the subsequent actual shipboard driving process of crew members, there is a lack of comprehensive and multi-dimensional methods to evaluate the safe driving behavior and comprehensive quality of crew members, which makes the accuracy of crew evaluation relatively low.
[0004] At present, the evaluation of crew members on board is mostly conducted by the captain to the driving personnel or by the heads of different departments to the crew members under their management. The evaluation content is mostly evaluated from the dimensions of work situation, comprehensive assessment of technical level, work attitude, safety awareness, work responsibility, labor discipline, foreign language level, health status, etc. The evaluation cycle is mostly quarterly evaluation, and the evaluation results are different levels such as excellent, good, general, and poor. The quarterly evaluation of crew members on board has a long evaluation cycle; the evaluation method is subjective through the captain or department head for grading evaluation; the evaluation scope is limited and cannot accurately describe the crew's on-board performance. Important aspects such as crew driving behavior for safe navigation of ships are not clearly reflected. In the analysis of crew sailing experience, traditional methods can only obtain the crew's historical on-board time, navigation area, and ship type, and cannot accurately obtain the crew's specific navigation area, port of call and other information.
[0005] In summary, the accuracy and timeliness of the current evaluation of seafarers cannot meet the needs of refined management of seafarers. Summary of the invention
[0006] The present invention solves the problem that the accuracy and timeliness of the current evaluation of crew members cannot meet the needs of refined management of crew members, and provides a crew evaluation method based on multi-source data.
[0007] The technical solution claimed in the present invention is as follows:
[0008] A crew assessment method based on multi-source data comprises the following steps:
[0009] S1: Acquire multi-source data of crew members, wherein the multi-source data of crew members include: basic information of crew members, information on crew members' sailing experience, traditional evaluation information of crew members, and information on risky driving behaviors of crew members; the traditional evaluation information of crew members includes information on the suitability of crew members for their positions;
[0010] S2: Evaluate the crew members in different dimensions according to the multi-source data of the crew members obtained in S1; the different dimensions include: job suitability, sailing experience and risk level; the job suitability is to score and grade the crew members on board according to the crew members' job suitability information; the sailing experience is to grade the crew members according to their sailing experience; the risk level includes monthly risk level and quarterly risk level; the monthly risk level is the crew members' risk score / the proportion of the crew members' time on board, the crew members' risk score is determined according to the type and number of the crew members' risky driving behavior information in the month, and the proportion of the crew members' time on board is determined according to the crew members' sailing information in the month; the quarterly risk level is the average value of the three monthly risk levels in the quarter;
[0011] S3: Visualize the crew evaluation results obtained in S2 and generate crew portraits.
[0012] In a specific embodiment of the present invention, the basic information of the crew in S1 includes: the crew's gender, age, marital status, educational level, English level, basic navigation experience and suitability information; the basic navigation experience of the crew includes: the crew's sailing time on board, position and service area; the crew's suitability information includes: certificate type, certificate position and validity date; the crew's sailing experience information includes: the crew's time on board, ship type, ship area and the main routes and main ports of call of the ship; the crew's job suitability information includes: the crew's work situation, technical level and other evaluation information; the work situation includes: comprehensive situation, work attitude, safety awareness, work responsibility and labor discipline; the technical level includes: comprehensive assessment, proficiency, maintenance ability, independent work ability and document management; the other evaluation information includes: foreign language level, teamwork spirit, health status, obedience awareness and management level; the crew's dangerous driving behavior information includes: unattended, stranger intrusion, overtime duty, violations, urgent encounters, deviations from the route, abnormal ship speed, abnormal steering; the violations include: smoking, playing with mobile phones on duty, and not dressing according to regulations.
[0013] Preferably, the unattended operation, intrusion by strangers, overtime duty and illegal behaviors are obtained by analyzing and identifying the real-time video information of the ship's bridge; the real-time video information of the ship's bridge is obtained by installing a high-definition camera on the bridge or using an existing ship's CCTV system.
[0014] Preferably, the unmanned operation utilizes the background / foreground segmentation algorithm based on Gaussian mixture model and YOLO algorithm to identify whether there are people on the bridge; the stranger intrusion utilizes Dlib algorithm to recognize the face of the person on the bridge, which is obtained by comparing with the driver's face database; the overtime duty is combined with the information of the ship-side duty personnel and obtained through the ship-side CCTV system; the urgent encounter, deviation from the route, abnormal ship speed and abnormal turning are obtained based on the data analysis of the ship's automatic identification system AIS.
[0015] In a specific embodiment of the present invention, the crew's job suitability evaluation is specifically as follows: based on the crew's job suitability information, each subcategory of the crew's work situation, technical level and other evaluation information is assigned a score, and the scores are added up, and the grades are divided according to the sum of the scores.
[0016] Preferably, the crew's sailing experience evaluation is specifically: grading the crew according to their on-board sailing time and sailing mileage, and evaluating the crew's sailing experience according to the level of their on-board sailing time and sailing mileage.
[0017] Preferably, a ship with less than 1 year of sailing and a mileage of less than 10,000 nautical miles is classified as a novice; a ship with 1-5 years of sailing and a mileage of 10,000-50,000 nautical miles is classified as having certain experience; a ship with 5-10 years of sailing and a mileage of 50,000-100,000 nautical miles is classified as having rich experience; a ship with more than 10 years of sailing and a mileage of more than 100,000 nautical miles is classified as having rich experience; if the ship's sailing time meets the standard but the sailing mileage does not meet the requirements, it will be downgraded.
[0018] The voyage mileage is calculated based on the crew's historical on-board time and ship name, superimposed on the ship's historical AIS data.
[0019] The specific method for calculating the monthly risk level is as follows: (1) scoring the crew member's dangerous driving behaviors, and counting the number and type of dangerous driving behaviors that occurred in the month, and calculating the crew member's risk score by multiplying the number and accumulating the scores; (2) counting the proportion of the crew member's time on board, which is equal to the number of days on board in the month divided by the total number of days in the month; (3) the crew member's risk level is equal to the crew member's risk score divided by the proportion of the crew member's time on board.
[0020] Preferably, when a crew member takes a vacation in a certain month and the monthly risk is not calculated, the quarterly risk will not be counted for that month and will be automatically adjusted to the sum of the two months divided by 2; if a crew member takes a vacation for 2 months within a quarter, the risk of this quarter will be assumed to be the same as the monthly risk.
[0021] Beneficial effects:
[0022] The present invention provides a crew evaluation method based on multi-source data, which obtains multi-source information of crew, wherein the multi-source data of crew includes: basic information of crew, sailing experience information of crew, traditional crew evaluation information and risk driving behavior information of crew; the traditional crew evaluation information includes information on suitability of crew; on the basis of traditional crew evaluation information, the risk driving behavior information of crew is further obtained to realize automatic identification of risk driving behavior of crew, and the suitability of crew, sailing experience and risk degree of crew are evaluated based on the obtained multi-source data of crew, so as to realize all-round, multi-dimensional and quantitative evaluation of crew members on board, improve the accuracy of crew evaluation, and solve the problem that the accuracy of the current evaluation of crew members cannot meet the demand for refined management of crew members; the risk degree of crew members is calculated based on the risk type and quantity and the driving time of crew members on board, and the problem that the traditional evaluation method of grading and evaluating by captain or department head is subjective is abandoned; in addition, the risk degree of crew members includes monthly risk degree and quarterly risk degree, so as to improve the frequency of risk evaluation of crew members, and solve the problem that the timeliness of the traditional evaluation of crew members on board only in quarterly basis cannot meet the demand for refined management of crew members.
[0023] The present invention classifies the crew's job suitability information in a refined manner, and assigns values and scores to the classifications, so that the crew's job suitability can be quantified; the crew's sailing experience information is further refined, and the crew's on-board sailing time and sailing history are quantitatively calculated; at the same time, the crew's dangerous driving behavior information is refined and classified, and the classifications are assigned values and scores, and the crew's monthly risk and quarterly risk are calculated; through the combination of the above three evaluation methods, the on-board work conditions, safe driving behaviors, technical levels and health conditions of different crew members are accurately analyzed, and crew members with dangerous driving behaviors, comprehensive qualities that do not meet the ship's navigation requirements, and problems with the crew's physical condition are actively discovered, ensuring that the ship management company or ship management personnel promptly and effectively deal with the crew accordingly to ensure the safe navigation of the ship. BRIEF DESCRIPTION OF THE DRAWINGS
[0024] Figure 1 It is a flow chart of the scheme of the present invention. DETAILED DESCRIPTION
[0025] The present invention is described in detail below in conjunction with the accompanying drawings:
[0026] The present invention provides a crew assessment method based on multi-source data, such as Figure 1 As shown, the following steps are included:
[0027] S1: Acquire multi-source data of crew members, wherein the multi-source data of crew members include: basic information of crew members, information on crew members' sailing experience, traditional evaluation information of crew members, and information on risky driving behaviors of crew members; the traditional evaluation information of crew members includes information on the suitability of crew members for their positions;
[0028] The basic information of the crew includes: the crew's gender, age, marital status, educational level, English proficiency, basic sailing experience and competency information; the basic sailing experience of the crew includes: the crew's sailing time on board, position and service area; the crew's competency information includes: certificate type, certificate position and validity date; the crew's sailing experience information includes: the crew's time on board, ship type, ship area and the main routes and main ports of call of the ship; the crew's job suitability information includes: the crew's work situation, technical level and other evaluation information; the work situation includes: comprehensive situation, work attitude, safety awareness, work responsibility and labor discipline; the technical level includes: comprehensive assessment, proficiency, maintenance ability, independent work ability and document management; the other evaluation information includes: foreign language level, teamwork spirit, health status, obedience awareness and management level; the crew's dangerous driving behavior information includes: unattended, stranger intrusion, overtime duty, illegal behavior, urgent encounter, deviation from the route, abnormal ship speed, abnormal steering; the illegal behavior includes: smoking, playing with mobile phones on duty, and not dressing according to regulations.
[0029] The unattended operation, intrusion by strangers, overtime duty and illegal behaviors are identified by analyzing the real-time video information of the ship's bridge; the real-time video information of the ship's bridge is obtained by installing a high-definition camera on the bridge or using the existing ship's CCTV system.
[0030] In some embodiments of the present invention, the unmanned operation uses a background / foreground segmentation algorithm based on a Gaussian mixture model and a YOLO algorithm to identify whether there are people on the bridge. The YOLO algorithm is a typical image recognition algorithm that can automatically identify people. If no people are identified, it is considered that there is no one; the stranger intrusion uses Dlib to recognize the face of the bridge, and obtains it by comparing the driver's face database. The driver's face database stores selfies entered by the driver; the overtime duty is combined with the ship-side duty personnel information and obtained through the ship-side CCTV system. When the ship is actually sailing, the ship-side duty personnel are on duty periodically. Usually, fixed time periods are fixed positions. The personnel use the ship-side software to configure the duty table once, and the software can determine the current duty officer according to the time. At the same time, CCTV shoots the bridge to analyze the current actual duty personnel and compares them with the scheduled personnel. If it is found that the actual duty time of the personnel on the bridge exceeds the duty schedule, it is considered to be overtime duty.
[0031] In some embodiments of the present invention, the urgent encounter, deviation from the route, abnormal ship speed, and abnormal turning are obtained according to the analysis of the ship's automatic identification system AIS data; wherein, urgent encounter: the urgent encounter of ships (urgent danger stage) refers to the process of ship navigation, two or more ships approach each other at a certain speed and heading, reaching a critical state where a collision may occur. At this stage, the ship needs to take emergency collision avoidance actions to avoid the occurrence of a collision. The AIS data includes the position, heading, speed, and bow parameters of the two ships, which can calculate the closest encounter distance DCPA and the closest encounter time TCPA of the two ships. These two parameters can comprehensively determine whether the current encounter is an urgent encounter. Deviation from the route: the degree of deviation between the actual position of the ship and the planned route, wherein the planned route needs to be manually entered into the system before sailing, and the actual position of the ship is obtained through AIS during the actual navigation process, and the distance from the planned route is calculated, and the judgment is made with the set threshold. Abnormal speed: The AIS data includes the ship's navigation speed, which is compared with the set speed threshold to determine whether the ship is overspeeding or underspeeding. Abnormal turning: AIS data includes the ship's heading or bow direction. The rate of change of the ship's heading or bow direction per unit time is calculated to determine whether the turning is too fast.
[0032] S2: Evaluate the crew members in different dimensions according to the multi-source data of the crew members obtained in S1; the different dimensions include: job suitability, sailing experience and risk level; the job suitability is to score and grade the crew members on board according to the crew members' job suitability information; the sailing experience is to grade the crew members according to their sailing experience; the risk level includes monthly risk level and quarterly risk level; the monthly risk level is the crew members' risk score / the proportion of the crew members' time on board, the crew members' risk score is determined according to the type and number of the crew members' risky driving behavior information in the month, and the proportion of the crew members' time on board is determined according to the crew members' sailing information in the month; the quarterly risk level is the average value of the three monthly risk levels in the quarter;
[0033] For suitability, the crew's work situation, technical level and other evaluation information are assigned values and scores for each subcategory based on the crew's suitability information, and the scores are added up, and the grades are divided according to the sum of the scores. In a specific embodiment of the present invention, based on the crew's suitability information, the types involved are scored in turn, and each category is assigned with excellent, good, general and poor grades, where excellent is 5 points, good is 4 points, general is 2 points, and poor is 0 points. The three major categories of on-board evaluation indicators such as work situation, technical level and others are generated by adding and calculating the sum of each subcategory, 22-25 points for excellent, 16-21 for good, 6-15 for general, and 0-5 for poor; the overall suitability of the crew is the sum of the scores of work situation, technical level and other three, 66-75 points for excellent, 48-65 points for good, 18-47 points for general, and 0-17 for poor. Among them, the crew's suitability information comes from the quarterly assessment information of the crew reported by the ship. For example: in the crew's suitability information of a certain crew member: the scores of comprehensive situation, work attitude, safety awareness, work responsibility and labor discipline are 5 points, 4 points, 5 points, 3 points and 5 points respectively, then the total score of the crew's work situation is 22 points, and the level of work situation is excellent; the scores of comprehensive assessment, proficiency, maintenance ability, independent work ability and document management are 4 points, 4 points, 5 points, 3 points and 4 points respectively, then the total score of the crew's technical level is 20 points, and the level of technical level is good; the scores of foreign language level, teamwork spirit, health status, obedience awareness and management level are 3 points, 2 points, 3 points, 3 points and 3 points respectively, then the total score of the crew's other scores is 14 points, and the other levels are general; therefore, the total score of the crew member is 54 points, and the overall level of suitability is good.
[0034] Said evaluation of crew sailing experience is specifically as follows: grading the crew according to the sailing time and mileage of the crew on board, and evaluating the sailing experience of the crew according to the level of the sailing time and mileage of the crew on board. In a specific embodiment of the present invention, the sailing time on board is less than 1 year and the mileage is less than 10,000 nautical miles, which is classified as a novice; the sailing time on board is 1-5 years and the mileage is 10,000-50,000 nautical miles, which is classified as having certain experience; the sailing time on board is 5-10 years and the mileage is 50,000-100,000 nautical miles, which is classified as having rich experience; the sailing time on board is greater than 10 years and the mileage is greater than 100,000 nautical miles, which is classified as having rich experience; wherein if the sailing time on board meets the standard but the mileage does not meet the requirements, the downgrade classification is performed. Sailing mileage is calculated based on the crew's historical time on board and the name of the ship, and the historical AIS data of the ship is superimposed. The dynamic data in the AIS data includes the ship track data, which is composed of a series of ship coordinate points (latitude and longitude). By calculating the distance between two adjacent coordinate points one by one, the ship's mileage can be accumulated.
[0035] In another embodiment of the present invention, in order to fully understand the sailing ability of the crew, based on the crew's historical time on board and the name of the ship, the historical AIS data of the ship is superimposed, and the historical voyage mileage, historical ports of call and its main navigation area of the crew are analyzed. The longer the historical voyage mileage, the more ports of call and the more navigation areas reflect the richer the sailing experience, and the stronger the navigation ability of the ship. The historical voyage mileage is calculated according to the distance of the AIS data point. The main navigation area determines different navigation areas according to its AIS data. The historical port of call obtains the port of call information according to the destination port information contained in the historical AIS static data; if the destination port information is not included in some historical AIS data, the judgment logic is added to obtain the port of call information. The ship's navigation status changes from being underway or anchored to berthing, and the current position of the ship is superimposed and analyzed with the port position of the ship to obtain the ship's berthing port information; if the ship's navigation status does not change, the ship's speed is less than 0.2 knots, the ship's navigation mileage is less than 100 meters in 30 minutes, and the duration is greater than 30 hours, and the ship's position information is superimposed and analyzed with the port position to obtain the ship's berthing port information.
[0036] As for the crew risk, in a specific embodiment of the present invention, the main content of the crew risk assessment is the dangerous ship behavior related to the driver during the navigation of the ship. Specifically, the corresponding risk scores are: unmanned score 8-10 points (advanced warning 10 points, intermediate 9 points, low 8 points), stranger 6 points, overtime duty 2-4 points (advanced warning 4 points, intermediate 3 points, low 2 points), violation 2 points, urgent encounter score 4-6 points (advanced warning 6 points, intermediate 5 points, low 4 points), deviation from the route score 4-6 points (advanced warning 6 points, intermediate 5 points, low 4 points), abnormal speed score 2-4 points (advanced warning 4 points, intermediate 3 points, low 2 points), abnormal steering score 2-4 points (advanced warning 4 points, intermediate 3 points, low 2 points).
[0037] In actual voyages, crew members may need to take leave or have the ship repaired. In this case, the crew risk algorithm needs to exclude the above special circumstances and then calculate the crew monthly risk and crew quarterly risk respectively.
[0038] The monthly risk calculation method is as follows:
[0039] The monthly risk assessment of crew members is conducted on the 1st of each month. The monthly risk assessment of the crew members in the previous month is conducted. The statistical algorithm is as follows:
[0040] (1) Score the dangerous driving behaviors of the crew members, and count the number and types of dangerous driving behaviors of the crew members in the month, and calculate the crew risk score by multiplying the number and accumulating the scores; specifically, count the number and types of all risk types of the crew members in this month (within 1-30 / 31 days), and calculate by multiplying the number and accumulating the scores. For example: Crew member A, in March, there was 1 unattended operation (senior), 2 strangers broke in, and 2 overtime shifts (senior), then the crew risk score of the crew member in that month is: 1*10+2*8+2*4=34 points.
[0041] (2) Count the proportion of the crew's time on board, which is equal to the number of days on board in the month divided by the total number of days in the month; specifically, count the proportion of the crew's time on board, crew proportion of time on board = this month's time on board (days) / total number of days in the month (days). By default, as long as the crew is on board and has duty schedule information, it can be counted as on-board driving time, and no longer refined to specific duty time. For example: in the month (assuming there are 31 days in the month) the crew has a duty schedule for 20 days, then the crew's time on board ratio = 20 / 31.
[0042] (3) Monthly risk level of crew = crew risk score / percentage of crew time on board.
[0043] If the crew is on leave as a whole, no scoring is required.
[0044] The quarterly risk calculation method is:
[0045] Quarterly risk assessment for crew members is conducted on the 1st of each quarter for the previous quarter.
[0046] The quarterly risk score is scored using the tie-break method: quarterly risk = the sum of the risk scores for the three months of each quarter / 3.
[0047] If a crew member takes leave in a certain month and the risk calculation for that month is not performed, the quarterly risk will not be counted for that month and will be automatically adjusted to the sum of February / 2.
[0048] If a crew member takes a two-month vacation during a quarter, the risk level for this quarter is assumed to be the same as the monthly risk level.
[0049] If the crew's driving behavior is relatively standard, the monthly risk is 0, and quarterly risk calculations also need to be statistically analyzed.
[0050] By calculating the crew's monthly, quarterly or annual risk level, if the crew's risk level is >80 points, it is considered high-risk driving; if the risk level is between 50-80 points, it is considered dangerous driving; if the risk level is between 20-50 points, it is considered risky driving; if the risk level is <20 points, it is considered safe driving.
[0051] S3: Visualize the evaluation results of the crew obtained in S2 and generate a crew portrait. In a specific embodiment of the present invention, job suitability is divided into excellent (A), good (B), average (C), and poor (D); crew sailing experience is divided into novice (D), with some experience (C), experienced (B), and experienced (A); crew risk is divided into high-risk driving (D), dangerous driving (C), risky driving (B), and safe driving (A). Evaluate the crew in the above three dimensions and generate a crew portrait to facilitate the relevant ship management company to better manage the crew. The crew portrait includes basic information of the crew.
[0052] The above embodiments are only used to illustrate the technical solutions of the present invention rather than to limit the present invention. Although the present invention has been described in detail through the above preferred embodiments, those skilled in the art should understand that various changes can be made in form and details without departing from the scope defined by the claims of the present invention.
Claims
1. A crew assessment method based on multi-source data, characterized in that: The steps include: S1: Acquire multi-source data of crew members, wherein the multi-source data of crew members include: basic information of crew members, information on crew members' sailing experience, traditional evaluation information of crew members, and information on risky driving behaviors of crew members; the traditional evaluation information of crew members includes information on the suitability of crew members for their positions; S2: Evaluate the crew members in different dimensions according to the multi-source data of the crew members obtained in S1; the different dimensions include: job suitability, sailing experience and risk level; the job suitability is to score and grade the crew members on board according to the crew members' job suitability information; the sailing experience is to grade the crew members according to their sailing experience; the risk level includes monthly risk level and quarterly risk level; the monthly risk level is the crew members' risk score / the proportion of the crew members' time on board, the crew members' risk score is determined according to the type and number of the crew members' risky driving behavior information in the month, and the proportion of the crew members' time on board is determined according to the crew members' sailing information in the month; the quarterly risk level is the average value of the three monthly risk levels in the quarter; S3: Visualize the crew evaluation results obtained in S2 and generate crew portraits.
2. The crew assessment method based on multi-source data according to claim 1, characterized in that: The basic information of the crew in S1 includes: the crew's gender, age, marital status, educational level, English proficiency, basic sailing experience and competency information; the basic sailing experience of the crew includes: the crew's sailing time on board, position and service area; the crew's competency information includes: certificate type, certificate position and validity date; the crew's sailing experience information includes: the crew's time on board, ship type, ship area and the main routes and main ports of call of the ship; the crew's job suitability information includes: the crew's work situation, technical level and other evaluation information; the work situation includes: comprehensive situation, work attitude, safety awareness, work responsibility and labor discipline; the technical level includes: comprehensive assessment, proficiency, maintenance ability, independent work ability and document management; the other evaluation information includes: foreign language level, teamwork spirit, health status, obedience awareness and management level; the crew's dangerous driving behavior information includes: unattended, stranger intrusion, overtime duty, violation, urgent encounter, deviation from the route, abnormal ship speed, abnormal steering; the violation includes: smoking, playing with mobile phones on duty, and not dressing according to regulations.
3. The crew assessment method based on multi-source data according to claim 2, characterized in that: The unattended operation, intrusion by strangers, overtime duty and illegal behaviors are identified by analyzing the real-time video information of the ship's bridge; the real-time video information of the ship's bridge is obtained by installing a high-definition camera on the bridge or using the existing ship's CCTV system.
4. The crew assessment method based on multi-source data according to claim 3, characterized in that: The unmanned operation uses the background / foreground segmentation algorithm based on Gaussian mixture model and the YOLO algorithm to identify whether there are people on the bridge; the stranger intrusion uses the Dlib algorithm to recognize the face of the person on the bridge, and obtains it by comparing with the driver's face database; the overtime duty is combined with the information of the ship-side duty personnel through the ship-side CCTV system; the urgent encounter, deviation from the route, abnormal ship speed, and abnormal turning are obtained based on the data analysis of the ship's automatic identification system AIS.
5. The crew assessment method based on multi-source data according to claim 2, characterized in that: The crew's job suitability evaluation is specifically as follows: based on the crew's job suitability information, each subcategory of the crew's work situation, technical level and other evaluation information is assigned a score, the scores are added up, and the grades are divided according to the sum of the scores.
6. The crew assessment method based on multi-source data according to claim 3, characterized in that: The crew's sailing experience evaluation is specifically as follows: the crew is graded according to their on-board sailing time and sailing mileage, and the crew's sailing experience is evaluated according to the level of their on-board sailing time and sailing mileage.
7. The crew assessment method based on multi-source data according to claim 6, characterized in that: Sailing time on board is less than 1 year and the mileage is less than 10,000 nautical miles, they are classified as novices; sailing time on board is 1-5 years and the mileage is 10,000-50,000 nautical miles, they are classified as having certain experience; sailing time on board is 5-10 years and the mileage is 50,000-100,000 nautical miles, they are classified as having rich experience; sailing time on board is more than 10 years and the mileage is more than 100,000 nautical miles, they are classified as rich experience; if the sailing time on board meets the standard but the mileage does not meet the requirements, they will be downgraded.
8. The crew assessment method based on multi-source data according to claim 6 or 7, characterized in that: The voyage mileage is calculated based on the crew's historical on-board time and ship name, superimposed on the ship's historical AIS data.
9. The crew assessment method based on multi-source data according to claim 1, characterized in that: The specific method for calculating the monthly risk level is as follows: (1) scoring the crew member's dangerous driving behaviors, and counting the number and type of dangerous driving behaviors that occurred in the month, and calculating the crew member's risk score by multiplying the number and accumulating the scores; (2) counting the proportion of the crew member's time on board, which is equal to the number of days on board in the month divided by the total number of days in the month; (3) the crew member's risk level is equal to the crew member's risk score divided by the proportion of the crew member's time on board.
10. The crew assessment method based on multi-source data according to claim 9, characterized in that: If a crew member takes leave in a certain month and the monthly risk calculation is not performed, the quarterly risk will not be counted for that month and will be automatically adjusted to the sum of the two months divided by 2. If a crew member takes leave for two months within a quarter, the quarterly risk will be assumed to be the same as the monthly risk.
Citation Information
Patent Citations
Method for realizing end-to-end service quality reliability guarantee
CN1716866A
Cited By
User portrait construction method and system, electronic equipment and storage medium
CN120524284A
