Post matching method and device based on machine learning
Through the job matching method based on machine learning, a person-job matching problem model is constructed and a heuristic algorithm is used to solve the problem of inefficient on-duty scheduling of ship personnel in traditional methods, and an efficient scheduling scheme generation under complex constraints is achieved.
Patent Information
- Application Number
- CN202410921556.8
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-07-10
- Publication Date
- 2024-11-08
- Estimated Expiration
- Not applicable · inactive patent
AI Technical Summary
The traditional method of on-duty scheduling of ship personnel is difficult to provide effective solutions when facing complex constraints and large-scale demands, resulting in inefficient solutions.
The job matching method based on machine learning is adopted, and the job matching problem model is constructed by obtaining personnel information, job information and scheduling constraint information, and the optimal scheduling solution is generated using heuristic algorithms and global search optimization methods.
It improves the flexibility and solution efficiency of ship personnel on duty scheduling under complex constraints, and quickly obtains the best scheduling solution, which is suitable for large-scale scheduling needs.
Smart Images

Figure CN118917575A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the field of ship management and operations research, and in particular to a job matching method and device based on machine learning. Background Art
[0002] In ship management, crew duty scheduling is an important part of ensuring safe operation and efficient management of ships. Traditional scheduling methods usually rely on manual experience or simple optimization algorithms, which often fail to provide effective solutions when faced with complex constraints and large-scale scheduling needs. In addition, factors such as the skill matching of ship personnel, work fatigue, and continuous duty restrictions also make the scheduling problem more complex and challenging. Therefore, there is an urgent need for a new method that can efficiently solve complex scheduling problems within a reasonable time. Summary of the invention
[0003] The technical problem to be solved by the present invention is to provide a job matching method and device based on machine learning, which improves the flexibility and solution efficiency of ship personnel duty scheduling with complex constraints, and thus quickly obtains the optimal scheduling plan.
[0004] In order to solve the above technical problems, the first aspect of the embodiment of the present invention discloses a job matching method based on machine learning, and the method includes:
[0005] S1, obtaining a first information set, a second information set and a third information set;
[0006] The first information set includes personnel information, position information and position change information v
[0007] The second information set includes a plurality of pieces of personnel information and personnel-position matching information; the personnel-position matching information is used to indicate the personnel-position matching degree;
[0008] The third information set includes the number of days in the shift cycle, the maximum number of shift days, the maximum number of consecutive shift days, specific shift information and specific non-shift information;
[0009] The specific value change information is used to indicate the value change information of a specific person, including person information, date information and change number information;
[0010] The specific unchangeable information is used to indicate that a specific person has changed the information that is unchangeable, including person information, date information and change information;
[0011] S2, using the first information set and the second information set to process the person-job matching problem model to obtain an optimized person-job matching problem model;
[0012] S3, using the optimization model for matching people with jobs, processing the third information set to obtain scheduling result information.
[0013] As an optional implementation, in the first aspect of the embodiment of the present invention, the model expression of the person-job matching problem is:
[0014]
[0015] Wherein, p represents the fitness value of person-job matching; a represents the person-job matching degree; x represents the first assigned value; f(1) represents the maximum continuous value constraint; f(2) represents the first humanized personalized constraint; f(3) represents the second humanized personalized constraint; f(4) represents the maximum value constraint; f(5) represents the minimum value constraint; o represents the second assigned value; p represents the third assigned value; tw represents the fourth assigned value; CW represents the first continuous value constraint; T + represents the second consecutive value change; T- represents the third consecutive value change; n represents the personnel index; k represents the position index; d represents the date index; s represents the change index; S represents the total number of daily value changes; N represents the total number of personnel; D represents the number of value change days vSOR represents the first set; DOR represents the second set; x ndsk represents the person-job matching constraint, and x ndsk +X nd(s+1)k indicates that discontinuous values are more constrained, and
[0016] As an optional implementation manner, in the first aspect of the embodiment of the present invention, the person-job matching problem model further includes:
[0017]
[0018] tw ndi ≤P nj , j = d, ..., d + i - 1;
[0019]
[0020]
[0021] Among them, tw represents the fourth allocation value; p represents the personnel date matching parameter; a represents the person-position adaptation parameter; x represents the first allocation value; o represents the second allocation value; p represents the third allocation value; j represents the date index; n represents the personnel index; k represents the position index; K represents the total number of positions; d represents the date index; s represents the change index; S represents the total number of daily value changes; N represents the total number of personnel; and D represents the number of value change days.
[0022] As an optional implementation, in the first aspect of the embodiment of the present invention, the person-job matching problem model is processed using the first information set and the second information set to obtain an optimized person-job matching problem model. It includes:
[0023] S21, parsing the first information set to obtain personnel list information, position information and position change information;
[0024] S22, parsing the second information set to obtain personnel information and person-job matching information;
[0025] S23, inputting the personnel information, the job information, the job shift information and the person-job matching information into the person-job matching problem model to obtain an optimized person-job matching problem model.
[0026] As an optional implementation manner, in the first aspect of the embodiment of the present invention, the use of the optimization person-job matching problem model to process the third information set to obtain scheduling result information includes:
[0027] S31, encoding the third information set and the fourth information set to obtain a duty matrix set;
[0028] The person-post duty matrix set includes a number of person-post duty matrices;
[0029] The person-post duty matrix is used to describe the duty information v of the duty personnel
[0030] The duty information includes duty personnel information, duty date information, duty position information and duty shift information.
[0031] S32, using the optimization model of the person-job matching problem, processing the person-job duty matrix set to obtain scheduling result information.
[0032] As an optional implementation manner, in the first aspect of the embodiment of the present invention, the encoding process of the third information set and the fourth information set to obtain a person-on-duty matrix set includes:
[0033] S311, parsing the first information set to obtain personnel list information, position information and position change information;
[0034] S312, parsing the third information set to obtain the shift cycle days, the maximum shift days, the maximum continuous shift days, specific shift information and specific non-shift information;
[0035] S313, processing the personnel list information to obtain all on-duty personnel;
[0036] S314, for any of the on-duty personnel, using the person-job matching problem model, the personnel list information, the job information, the job shift information, the shift cycle days, the maximum number of shift days, the maximum number of consecutive shift days, the specific shift information and the specific non-shift information are processed to obtain a target person-job duty matrix;
[0037] S315, combining all target post duty matrices in order to obtain a post duty matrix set.
[0038] As an optional implementation, in the first aspect of the embodiment of the present invention, the optimization model of the person-job matching problem is used to process the person-job duty matrix set to obtain the scheduling result information, including:
[0039] S321, preset a first threshold; preset the number of iterations; preset the number of processing times to be 1; traverse the set of person-post duty matrices to obtain all person-post duty matrices;
[0040] S322, randomly selecting two of the man-post duty matrices to obtain a first target man-post duty matrix and a second target man-post duty matrix;
[0041] S323, using the objective function algorithm model, calculating and processing the first target person-post duty matrix and the second target person-post duty matrix to obtain a first processed value;
[0042] S324, determining whether the first processed value is greater than the first threshold value, and obtaining a first determination result;
[0043] When the first judgment result is yes, execute S325;
[0044] When the first judgment result is no, executing S326;
[0045] S325, using the optimization model of the person-job matching problem to obtain a target memory matrix;
[0046] Using the target memory matrix to update the first target person-on-duty matrix, to obtain a first target person-on-duty optimization matrix;
[0047] S326, using the second target person-post duty matrix to update the first target person-post duty matrix to obtain a first target person-post duty optimization matrix;
[0048] S327, deleting the first target person-post duty matrix from the person-post duty matrix set to obtain an updated person-post duty matrix set;
[0049] Increase the number of processing times by 1;
[0050] S328, determining whether the processing times are greater than the iteration times, and obtaining a second determination result;
[0051] When the second judgment result is yes, execute S329;
[0052] When the second judgment result is no, executing S322;
[0053] S429, sequentially combine all of the first target person-post duty optimization matrices to obtain scheduling result information.
[0054] A second aspect of an embodiment of the present invention discloses a job matching device based on machine learning, the device comprising:
[0055] An acquisition module, acquiring a first information set, a second information set and a third information set;
[0056] A first processing module is used to process the person-job matching problem model using the first information set and the second information set to obtain an optimized person-job matching problem model;
[0057] The second processing module is used to process the third information set using the optimization person-job matching problem model to obtain scheduling result information.
[0058] The third aspect of the present invention discloses another job matching device based on machine learning, the device comprising:
[0059] A memory storing executable program code;
[0060] a processor coupled to the memory;
[0061] The processor calls the executable program code stored in the memory to execute part or all of the steps in the machine learning-based job matching method disclosed in the first aspect of the embodiment of the present invention.
[0062] The fourth aspect of the present invention discloses a computer-readable storage medium, which stores computer instructions. When the computer instructions are called, they are used to execute part or all of the steps in the machine learning-based job matching method disclosed in the first aspect of an embodiment of the present invention.
[0063] Compared with the prior art, the present invention effectively solves the problem of low efficiency in solving the problem of large-scale ship personnel scheduling in the prior art, provides a flexible and efficient scheduling solution, and provides strong technical support for ship management. BRIEF DESCRIPTION OF THE DRAWINGS
[0064] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.
[0065] Figure 1 It is a flowchart of a job matching method based on machine learning disclosed in an embodiment of the present invention;
[0066] Figure 2 It is a structural schematic diagram of a job matching device based on machine learning disclosed in an embodiment of the present invention;
[0067] Figure 3 It is a structural schematic diagram of another job matching device based on machine learning disclosed in an embodiment of the present invention. DETAILED DESCRIPTION
[0068] In order to enable those skilled in the art to better understand the scheme of the present invention, the technical scheme in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings 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.
[0069] The terms "first", "second", etc. in the specification and claims of the present invention and the above-mentioned drawings are used to distinguish different objects, rather than to describe a specific order. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions. For example, a process, method, device, product or equipment that includes a series of steps or units is not limited to the listed steps or units, but may optionally include steps or units that are not listed, or may optionally include other steps or units that are inherent to these processes, methods, products or equipment.
[0070] Reference to "embodiments" herein means that a particular feature, structure, or characteristic described in conjunction with the embodiments may be included in at least one embodiment of the present invention. The appearance of the phrase in various places in the specification does not necessarily refer to the same embodiment, nor is it an independent or alternative embodiment that is mutually exclusive with other embodiments. It is explicitly and implicitly understood by those skilled in the art that the embodiments described herein may be combined with other embodiments.
[0071] The present invention discloses a job matching method and device based on machine learning, which improves the flexibility and solution efficiency of ship personnel duty scheduling with complex constraints, and thus quickly obtains the optimal scheduling plan. The following are detailed descriptions.
[0072] Embodiment 1
[0073] See also Figure 1 , Figure 1 : is a flow chart of a job matching method based on machine learning disclosed in an embodiment of the present invention. Figure 1 The described machine learning-based job matching method is applied to a ship management system, such as a local server or cloud server for ship management, and the embodiments of the present invention do not limit this. Figure 1 As shown, the job matching method based on machine learning may include the following operations:
[0074] S1, obtaining a first information set, a second information set and a third information set;
[0075] It should be noted that the first information set includes personnel information, position information and position shift information;
[0076] It should be noted that the second information set includes a number of pieces of personnel information and personnel-job matching information; the personnel-job matching information is used to indicate the personnel-job matching degree;
[0077] It should be noted that the third information set includes the shift cycle days, the maximum shift days, the maximum continuous shift days, specific shift information and specific non-shift information;
[0078] It should be noted that the specific value change information is used to indicate the value change information of a specific person, including person information, date information and change number information;
[0079] It should be noted that the specific unchangeable information is used to indicate that a specific person's change order is unchangeable information, including person information, date information and change order information;
[0080] S2, using the first information set and the second information set to process the person-job matching problem model to obtain an optimized person-job matching problem model;
[0081] It should be noted that the above-mentioned optimization model for the person-job matching problem represents a model with specific parameters;
[0082] It should be noted that the parameters include: personnel information, position information, position shift information, person-position matching information, shift cycle days, maximum shift days, maximum continuous shift days, specific shift information and specific non-shift information;
[0083] S3, using the optimization model for matching people with jobs, processing the third information set to obtain scheduling result information.
[0084] It can be seen that the implementation of the machine learning-based job matching method described in the embodiment of the present invention improves the flexibility and solution efficiency of ship personnel duty scheduling with complex constraints, thereby quickly obtaining the optimal scheduling plan.
[0085] In an optional embodiment, in the above step S2, the model expression of the person-job matching problem is:
[0086]
[0087]
[0088] Among them, p represents the fitness value of person-job matching; a represents the person-job matching degree; x represents the first distribution value; f(1) represents the maximum continuous value constraint; f(2) represents the first humanized personalized constraint; f(3) represents the second humanized personalized constraint; f(4) represents the maximum value constraint; f(5) represents the minimum value constraint; o represents the second distribution value; p represents the third distribution value; tw represents the fourth distribution value; CW represents the first continuous value constraint; T + represents the second consecutive value change; T- represents the third consecutive value change; n represents the personnel index; k represents the position index; d represents the date index; s represents the change index; S represents the total number of daily value changes; N represents the total number of personnel; D represents the number of value change days; SOR represents the first set; DOR represents the second set; x ndsk represents the person-job matching constraint, and x ndsk +x nd(s+1)k Indicates that discontinuous values are more constrained,
[0089] It should be noted that the person-job matching degree is a positive value indicating the matching degree between personnel and positions. The larger the person-job matching degree is, the higher the matching degree between personnel and positions is. If there is no match, the value is 0.
[0090] It should be noted that the first allocation value indicates whether the person is allocated to the designated position of the designated shift on the designated date. When the first allocation value is 1, it indicates that the person is allocated to the designated position of the designated shift on the designated date. When the first allocation value is 0, it indicates that the person is not allocated to the designated position of the designated shift on the designated date.
[0091] It should be noted that the first humanized personalized constraint means that the designated person is not on duty at the designated post on the designated date;
[0092] It should be noted that the second humanized personalized constraint means that the designated personnel are not on duty on the designated date;
[0093] It should be noted that the second allocation value indicates whether the person is assigned to a specified shift on a specified date; when the second allocation value is 1, it indicates that the person is assigned to a specified shift on a specified date; when the second allocation value is 01, it indicates that the person is not assigned to a specified shift on a specified date;
[0094] It should be noted that the third allocation value indicates whether the person is assigned a specified date; when the third allocation value is 1, it indicates that the person is assigned a specified date; when the third allocation value is 0, it indicates that the person is not assigned a specified date;
[0095] It should be noted that the fourth allocated value indicates whether the person has changed the value for a specified number of days continuously since the specified date; when the fourth allocated value is 1, it indicates that the person has changed the value for a specified number of days continuously since the specified date; when the fourth allocated value is 0, it indicates that the person has not changed the value for a specified number of days continuously since the specified date;
[0096] It should be noted that the first continuous value change value represents the maximum number of continuous value change days for a person;
[0097] It should be noted that the second continuous value change value represents the maximum number of continuous value change days for a person;
[0098] It should be noted that the third continuous value change value represents the minimum number of continuous value change days for personnel;
[0099] It should be noted that the first set represents a constraint set that the designated duty personnel are not on duty on the designated shift on the designated day, and is in the form of a triple of (n, d, s);
[0100] It should be noted that the second set represents a constraint set that the designated duty personnel are not on duty on the designated date, and is in the form of a tuple of (n, d);
[0101] It should be noted that p represents the fitness value of person-job matching, which is used to evaluate the quality of the shift scheduling. The fitness is calculated based on the total matching degree of the shift schedule matrix, and then the total value of the soft constraint is subtracted;
[0102] It should be noted that the soft constraints include maximum continuous value constraint f(1), humanized constraint f(2), personalized constraint f(3), maximum value constraint f(4), and minimum value constraint f(5).
[0103] It can be seen that the implementation of the machine learning-based job matching method described in the embodiment of the present invention uses a heuristic algorithm to provide a feasible solution in a relatively short time, thereby improving the flexibility and solution efficiency of ship personnel duty scheduling with complex constraints, and can quickly obtain the optimal scheduling plan. Compared with the traditional integer linear programming method, it is particularly suitable for large-scale scheduling needs.
[0104] In another optional embodiment, in the above step S2, the person-job matching problem model further includes:
[0105]
[0106] tw ndi ≤p nj , j = d, ..., d + i - 1;
[0107]
[0108]
[0109] Among them, tw represents the fourth allocation value; p represents the personnel date matching parameter; a represents the person-position adaptation parameter; x represents the first allocation value; o represents the second allocation value; p represents the third allocation value; j represents the date index; t represents; n represents the personnel index; k represents the position index; K represents the total number of positions; d represents the date index; s represents the change index; S represents the total number of daily value changes; N represents the total number of personnel; D represents the number of value change days;
[0110] It should be noted that the person-job adaptation parameter indicates the degree of adaptability of a person to a job;
[0111] It can be seen that the implementation of the machine learning-based job matching method described in the embodiment of the present invention can flexibly handle various complex scheduling constraints, including job matching degree, upper limit of duty times, continuous duty limit and duty requirements of specific personnel, etc., thereby improving the flexibility and solution efficiency of ship personnel duty scheduling with complex constraints, and the optimal scheduling plan can be quickly obtained.
[0112] In another optional embodiment, in the above step S2, the use of the first information set and the second information set to process the person-job matching problem model to obtain an optimized person-job matching problem model includes:
[0113] S21, parsing the first information set to obtain personnel list information, position information and position change information;
[0114] It should be noted that the parsing of the first information set to obtain the personnel list information, position information and position change information includes:
[0115] S211, traversing the first information set to obtain all structured first information;
[0116] S212, extracting any of the structured first information to obtain a personnel list, a position value, and a person-position matching value;
[0117] S213, arranging the personnel list, position values, and person-position matching values in order to obtain personnel information, position information, and person-position matching information;
[0118] S22, parsing the second information set to obtain personnel information and person-job matching information;
[0119] It should be noted that the parsing of the second information set to obtain personnel information and person-job matching information includes:
[0120] S221, traverse the second information set to obtain all structured second information;
[0121] S222, extracting any of the structured second information to obtain a personnel list and a person-position matching value;
[0122] S223, arranging the personnel list and the person-job matching values in order to obtain personnel information and person-job matching information;
[0123] S23, inputting the personnel information, the job information, the job shift information and the person-job matching information into the person-job matching problem model to obtain an optimized person-job matching problem model.
[0124] It can be seen that the machine learning-based job matching method described in the embodiment of the present invention is implemented, and parameters are adjusted according to actual needs, which improves the flexibility and solution efficiency of ship personnel duty scheduling with complex constraints, and thus quickly obtains the optimal scheduling plan.
[0125] In another optional embodiment, in the above step S3, the third information set is processed using the optimized person-job matching problem model to obtain the scheduling result information, including:
[0126] S31, encoding the third information set and the fourth information set to obtain a duty matrix set;
[0127] The person-post duty matrix set includes a number of person-post duty matrices;
[0128] The personnel duty matrix is used to describe the duty information of the duty personnel;
[0129] The duty information includes duty personnel information, duty date information, duty post information and duty shift information;
[0130] S32, using the optimization model of the person-job matching problem, processing the person-job duty matrix set to obtain scheduling result information.
[0131] It can be seen that the implementation of the machine learning-based job matching method described in the embodiment of the present invention improves the flexibility and solution efficiency of ship personnel duty scheduling with complex constraints, thereby quickly obtaining the optimal scheduling plan.
[0132] In an optional embodiment, in the above step S31, the encoding process is performed on the third information set and the fourth information set to obtain a person-on-duty matrix set, including:
[0133] S311, parsing the first information set to obtain personnel list information, position information and position change information;
[0134] It should be noted that the parsing process refers to extracting information according to information type to obtain personnel list information, position information and position change information;
[0135] S312, parsing the third information set to obtain the shift cycle days, the maximum shift days, the maximum continuous shift days, specific shift information and specific non-shift information;
[0136] It should be noted that the parsing of the third information set means extracting information according to information type to obtain the shift cycle days, the maximum shift days, the maximum continuous shift days, specific shift information and specific non-shift information;
[0137] S313, processing the personnel list information to obtain all on-duty personnel;
[0138] It should be noted that the processing of the personnel list information means extracting the personnel list in the personnel list information, and obtaining all the on-duty personnel by combining the serial number and the list;
[0139] S314, for any of the on-duty personnel, using the person-job matching problem model, the personnel list information, the job information, the job shift information, the shift cycle days, the maximum number of shift days, the maximum number of consecutive shift days, the specific shift information and the specific non-shift information are processed to obtain a target person-job duty matrix;
[0140] S315, combining all target post duty matrices in order to obtain a post duty matrix set.
[0141] It can be seen that the implementation of the machine learning-based job matching method described in the embodiment of the present invention improves the flexibility and solution efficiency of ship personnel duty scheduling with complex constraints, thereby quickly obtaining the optimal scheduling plan.
[0142] In another optional embodiment, in the above step S32, the optimization model of the person-job matching problem is used to process the person-job duty matrix set to obtain the scheduling result information, including:
[0143] S321, preset a first threshold; preset the number of iterations; preset the number of processing times to be 1; traverse the set of person-post duty matrices to obtain all person-post duty matrices;
[0144] It should be noted that the first threshold represents the set follow probability, which is used to indicate the probability of whether a human post duty matrix chooses to follow another human post duty matrix. A lower value can increase the breadth of the exploration space, while a higher value can increase the depth of developing known excellent solutions.
[0145] It should be noted that, in this embodiment, the first threshold is set to 0.25;
[0146] S322, randomly selecting two of the man-post duty matrices to obtain a first target man-post duty matrix and a second target man-post duty matrix;
[0147] S323, using the objective function algorithm model, calculating and processing the first target person-post duty matrix and the second target person-post duty matrix to obtain a first processed value;
[0148] S324, determining whether the first processed value is greater than the first threshold value, and obtaining a first determination result;
[0149] When the first judgment result is yes, execute S325;
[0150] When the first judgment result is no, executing S326;
[0151] S325, using the optimization model of the person-job matching problem to obtain a target memory matrix;
[0152] Using the target memory matrix to update the first target person-on-duty matrix, to obtain a first target person-on-duty optimization matrix;
[0153] S326, using the second target person-post duty matrix to update the first target person-post duty matrix to obtain a first target person-post duty optimization matrix;
[0154] S327, deleting the first target person-post duty matrix from the person-post duty matrix set to obtain an updated person-post duty matrix set;
[0155] Increase the number of processing times by 1;
[0156] S328, determining whether the processing times are greater than the iteration times, and obtaining a second determination result;
[0157] When the second judgment result is yes, execute S329;
[0158] When the second judgment result is no, executing S322;
[0159] S329, sequentially combining all the first target person-post duty optimization matrices to obtain scheduling result information;
[0160] As shown in the following table, assuming there are D days in total, each position has a maximum of S shifts per day, and there are K positions in total, then this table has D*S rows and K columns, and the cell content is the on-duty personnel.
[0161]
[0162] The individual code for each on-duty personnel is a matrix with D*S rows and K columns, and each value in the matrix corresponds to the personnel number. For example, the matrix corresponding to the above table is:
[0163]
[0164] The first row is the first shift of the first day, the second row is the second shift of the first day, the third row is the first shift of the second day, and so on. In this way, each shift schedule corresponds to a matrix. At the same time, the right person must be placed in the right position, that is, the person must go to the position with a proficiency not equal to 0. The personnel are regarded as point set A, and the positions are regarded as point set B. If person a is competent for position b, then there is an edge connection between a and b. The Hungarian algorithm is used to solve the data to obtain the scheduling result information;
[0165] It can be seen that the implementation of the machine learning-based job matching method described in the embodiment of the present invention adopts a heuristic search optimization method that combines global search and local search, which effectively improves the search efficiency of the algorithm and the quality of the solution, and can quickly obtain the optimal scheduling plan.
[0166] In another optional embodiment, in the above step S323, the objective function algorithm model expression is:
[0167] f=I-Tn-M d -G r ;
[0168]
[0169] M d =max(C_n)for n∈{1, 2,...,N};
[0170]
[0171] Where I represents the total matching degree; Tn represents the total number of work shifts of person n; M d Indicates the maximum number of consecutive working shifts; G r represents the penalty value of personalized demand; d represents the duty days index; S represents the total number of daily shifts; s represents the duty shift; D represents the total number of duty days; K represents the number of posts; N represents the number of on-duty personnel; n represents the on-duty personnel index; k represents the post index; x {s,d,k} represents the personnel number of the kth position in the sth shift on the dth day; a{n, k} represents the matching degree of personnel n to position k; y {n,s,d} Said crew member n was on duty on the sth shift on the dth day; z {n,d,s} Indicates a specific shift duty; W {n,d} Indicates whether person n is on duty on any shift on day d. If not, it is 1, otherwise it is 0. SOR represents the first set; DOR represents the second set; C_n represents the maximum number of consecutive working days for person n;
[0172] It should be noted that the total matching degree means maximizing the overall matching degree of personnel under the premise of satisfying various constraints;
[0173] It should be noted that when the duty value = 1, it means that crew member n is on duty on the sth shift on the dth day; when the duty value = 0, it means that crew member n is not on duty on the sth shift on the dth day;
[0174] It should be noted that the specific shift duty value is used to indicate whether person n is on duty in the sth shift on the dth day; when the specific shift duty value is 1, person n is not on duty; when the specific shift duty value is 0, person n is on duty;
[0175] It should be noted that the personalized demand penalty value is used to indicate that if person n violates the requirement of not being on duty, the penalty value will be increased;
[0176] It should be noted that the maximum number of consecutive working days means that for each on-duty personnel n, the number of days d and the number of shifts s are traversed, the number of consecutive working days is recorded, and the maximum value is obtained;
[0177] It can be seen that the implementation of the machine learning-based job matching method described in the embodiment of the present invention improves the flexibility and solution efficiency of ship personnel duty scheduling with complex constraints, thereby quickly obtaining the optimal scheduling plan.
[0178] In order to verify the effect of the algorithm, the performance of the algorithm of the present invention was compared with that of HiGHS. We randomly generated test data of different sizes, with a fixed scheduling cycle of 14 days and 3 shifts per day. The open source solver HiGHS, which is recognized to have good performance and high versatility, was used for simulation. The test was conducted on the Winll platform, and the processor was Intel(R) Core(TM) i7-12700H. HiGHS (1.5.3) was called in Python3.10 to solve the problem. In the simulation experiment, the objective function values obtained by the two methods are the same and are no longer listed separately.
[0179] As can be seen from the table below, compared with open source solvers, this algorithm has a greater advantage in large-scale problems.
[0180] Number of people N Number of positions K Design Algorithms HkDJ 12 3 0.049 0.084 60 10 0.430 7.722 100 30 1.702 9.849 200 30 3.774 45.764 200 60 6.335 135.175 300 90 15.562 546.400
[0181] Embodiment 2
[0182] See also Figure 2 , Figure 2 is a schematic diagram of the structure of a job matching device based on machine learning disclosed in an embodiment of the present invention. Figure 2 The described device can be applied to a ship management system, such as a local server or a cloud server for ship management, etc., which is not limited in the embodiments of the present invention. Figure 2 As shown, the device may include:
[0183] An acquisition module 101 acquires a first information set, a second information set, and a third information set;
[0184] A first processing module 102 is used to process the person-job matching problem model using the first information set and the second information set to obtain an optimized person-job matching problem model;
[0185] The second processing module 103 is used to process the third information set using the optimized person-job matching problem model to obtain scheduling result information.
[0186] It can be seen that the implementation of the machine learning-based job matching method described in the embodiment of the present invention improves the flexibility and solution efficiency of ship personnel duty scheduling with complex constraints, thereby quickly obtaining the optimal scheduling plan.
[0187] Embodiment 3
[0188] See also Figure 3 , Figure 3 is a structural diagram of another position matching device based on machine learning disclosed in an embodiment of the present invention. Figure 3 The described device can be applied to a ship management system, such as a local server or a cloud server for ship management, etc., which is not limited in the embodiments of the present invention. Figure 3 As shown, the device may include:
[0189] A memory 201 storing executable program codes;
[0190] a processor 202 coupled to the memory 201;
[0191] The processor 202 calls the executable program code stored in the memory 201 to execute the steps in the machine learning-based job matching method described in Example 1.
[0192] Embodiment 4
[0193] An embodiment of the present invention discloses a computer-readable storage medium, which stores a computer program for electronic data exchange, wherein the computer program enables a computer to execute the steps in the machine learning-based job matching method described in Embodiment 1.
[0194] The device embodiments described above are only illustrative, wherein the modules described as separate components may or may not be physically separated, and the components displayed as modules may or may not be physical modules, i.e., they may be located in one place, or they may be distributed on multiple network modules. Some or all of the modules may be selected according to actual needs to achieve the purpose of the scheme of this embodiment. Those of ordinary skill in the art may understand and implement it without creative work.
[0195] Through the specific description of the above embodiments, those skilled in the art can clearly understand that each implementation method can be implemented by means of software plus a necessary general hardware platform, and of course, it can also be implemented by hardware. Based on such an understanding, the above technical solution can be essentially or partly contributed to the prior art in the form of a software product, and the computer software product can be stored in a computer-readable storage medium, and the storage medium includes a read-only memory (ROM), a random access memory (RAM), a programmable read-only memory (PROM), an erasable programmable read-only memory (EPROM), a one-time programmable read-only memory (OTPROM), an electronically erasable rewritable read-only memory (EEPROM), a compact disc (CD-ROM) or other optical disc storage, a magnetic disk storage, a magnetic tape storage, or any other computer-readable medium that can be used to carry or store data.
[0196] Finally, it should be noted that the machine learning-based job matching method and device disclosed in the embodiments of the present invention only disclose the preferred embodiments of the present invention, which are only used to illustrate the technical solution of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that the technical solutions described in the aforementioned embodiments can still be modified, or some of the technical features can be replaced by equivalents. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention.
Claims
1. A job matching method based on machine learning, characterized in that: The method comprises: S1, obtaining a first information set, a second information set and a third information set; The first information set includes personnel list information, position information and position change information; The second information set includes a plurality of personnel list information and personnel-position matching information; the personnel-position matching information is used to indicate the personnel-position matching degree; The third information set includes the number of days in the shift cycle, the maximum number of shift days, the maximum number of consecutive shift days, specific shift information and specific non-shift information; The specific value change information is used to indicate the value change information of a specific person, including person information, date information and change number information; The specific unchangeable information is used to indicate that a specific person has changed the information that is unchangeable, including person information, date information and change information; S2, using the first information set and the second information set to process the person-job matching problem model to obtain an optimized person-job matching problem model; S3, using the optimization model for matching people with jobs, processing the first information set and the third information set to obtain scheduling result information.
2. The job matching method based on machine learning according to claim 1, characterized in that: The model expression of the person-job matching problem is: f(2)=∑ (n,d,s)∈SOR o nds ; Among them, p represents the fitness value of person-job matching; a represents the person-job matching degree; x represents the first distribution value; f(1) represents the maximum continuous value constraint; f(2) represents the first humanized personalized constraint; f(3) represents the second humanized personalized constraint; f(4) represents the maximum value constraint; f(5) represents the minimum value constraint; o represents the second distribution value; p represents the third distribution value; tw represents the fourth distribution value; CW represents the first continuous value constraint; T + Indicates that the second consecutive value is more valuable; T - represents the third consecutive value change; n represents the personnel index; k represents the position index; d represents the date index; s represents the change index; S represents the total number of daily value changes; N represents the total number of personnel; D represents the number of value change days; SOR represents the first set; DOR represents the second set; x ndsk represents the person-job matching constraint, and x ndsk +x nd(s+1)k indicates that discontinuous values are more constrained, and 3. The job matching method based on machine learning according to claim 2, characterized in that: The person-job matching problem model also includes: tw ndi ≤p nj ,j=d,...,d+i-1; Among them, tw represents the fourth allocated value; p represents the personnel date matching parameter; a represents the person-position adaptation parameter; x represents the first allocated value; o represents the second allocated value; p represents the third allocated value; j represents the date index; t represents; n represents the personnel index; k represents the position index; K represents the total number of positions; d represents the date index; s represents the change index; S represents the total number of daily value changes; N represents the total number of personnel; and D represents the number of value change days.
4. The job matching method based on machine learning according to claim 1, characterized in that: The method of processing the person-job matching problem model by using the first information set and the second information set to obtain an optimized person-job matching problem model includes: S21, parsing the first information set to obtain personnel list information, position information and position change information; S22, parsing the second information set to obtain personnel list information and person-position matching information; S23, inputting the personnel information, the job information, the job shift information and the person-job matching information into the person-job matching problem model to obtain an optimized person-job matching problem model.
5. The job matching method based on machine learning according to claim 1, characterized in that: The third information set is processed using the optimization model for matching people with positions to obtain scheduling result information, including: S31, encoding the first information set and the third information set to obtain a person-post duty matrix set; The person-post duty matrix set includes a number of person-post duty matrices; The personnel duty matrix is used to describe the duty information of the duty personnel; The duty information includes duty personnel information, duty date information, duty post information and duty shift information; S32, using the optimization model of the person-job matching problem, processing the person-job duty matrix set to obtain scheduling result information.
6. The job matching method based on machine learning according to claim 5, characterized in that: The encoding process of the first information set and the third information set to obtain a person-on-duty matrix set includes: S311, parsing the first information set to obtain personnel list information, position information and position change information; S312, parsing the third information set to obtain the shift cycle days, the maximum shift days, the maximum continuous shift days, specific shift information and specific non-shift information; S313, processing the personnel list information to obtain all on-duty personnel; S314, for any of the on-duty personnel, using the person-job matching problem model, the personnel list information, the job information, the job shift information, the shift cycle days, the maximum number of shift days, the maximum number of consecutive shift days, the specific shift information and the specific non-shift information are processed to obtain a target person-job duty matrix; S315, combining all target post duty matrices in order to obtain a post duty matrix set.
7. The job matching method based on machine learning according to claim 5, characterized in that: The optimization model for matching people with positions is used to process the person-position duty matrix set to obtain the scheduling result information, including: S321, preset a first threshold; preset the number of iterations; preset the number of processing times to be 1; traverse the set of person-post duty matrices to obtain all person-post duty matrices; S322, randomly selecting two of the man-post duty matrices to obtain a first target man-post duty matrix and a second target man-post duty matrix; S323, using the objective function algorithm model, calculating and processing the first target person-post duty matrix and the second target person-post duty matrix to obtain a first processed value; S324, determining whether the first processed value is greater than the first threshold value, and obtaining a first determination result; When the first judgment result is yes, execute S325; When the first judgment result is no, executing S326; S325, using the optimization model of the person-job matching problem to obtain a target memory matrix; Using the target memory matrix to update the first target person-on-duty matrix, to obtain a first target person-on-duty optimization matrix; S326, using the second target person-post duty matrix to update the first target person-post duty matrix to obtain a first target person-post duty optimization matrix; S327, deleting the first target person-post duty matrix from the person-post duty matrix set to obtain an updated person-post duty matrix set; Increase the number of processing times by 1; S328, determining whether the processing times are greater than the iteration times, and obtaining a second determination result; When the second judgment result is yes, execute S329; When the second judgment result is no, executing S322; S329, sequentially combine all of the first target person-post duty optimization matrices to obtain scheduling result information.
8. A job matching device based on machine learning, characterized in that: The device comprises: An acquisition module, acquiring a first information set, a second information set and a third information set; A first processing module is used to process the person-job matching problem model using the first information set and the second information set to obtain an optimized person-job matching problem model; The second processing module is used to process the third information set using the optimization person-job matching problem model to obtain scheduling result information.
9. A job matching device based on machine learning, characterized in that: The device comprises: A memory storing executable program code; a processor coupled to the memory; The processor calls the executable program code stored in the memory to execute the machine learning-based job matching method as described in any one of claims 1-7.
10. A computer-readable storage medium, wherein the computer-readable storage medium stores computer instructions, and when the computer instructions are called, they are used to execute the machine learning-based job matching method as described in any one of claims 1-7.
Citation Information
Cited By
Dynamic matching method and device for ship post demand information
CN121481068A
