Work order management system based on deep learning
Through the work order management system based on deep learning, the employee combination is dynamically allocated, which solves the problem of inefficiency under fixed team configuration in the field work order system, realizes the automation and intelligence of field work order personnel allocation, and improves the operation of the water supply company Efficiency and cost-effectiveness.
Patent Information
- Application Number
- CN202411554651.5
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-11-04
- Publication Date
- 2024-11-29
- Estimated Expiration
- Not applicable · inactive patent
AI Technical Summary
The existing field work ticket system is difficult to meet the needs of different work ticket projects under a fixed team configuration, resulting in inefficient efficiency and frequent personnel dispatch, resulting in high cost and low efficiency of labor allocation.
Using a work order management system based on deep learning, the work order task, employee skills and collaboration capabilities are collected through the data acquisition module, and the structure data generation module converts it into structured data. The deep learning model module dynamically allocates the optimal employee combination to form An efficient work order team can realize the automation and intelligence of team configuration.
Dynamic optimization of team configuration avoids personnel dispatch, saves labor allocation costs, improves operating efficiency and human resource utilization, and brings double improvements in economic benefits and operational performance.
Smart Images

Figure CN119047810A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the field of field dispatch, and more specifically, to a work order management system based on deep learning. Background Art
[0002] Water supply companies face various on-site work requirements such as pipe network leak detection, repairs, valve maintenance, etc. In order to efficiently handle these requirements, the company has developed a field work order system to receive and dispatch relevant work orders.
[0003] A field work order project will include several sub-projects, and the water supply company needs to provide a team to carry out the work. When configuring the team, most of the time, a fixed team is configured to implement the work. However, different field work order projects have different requirements. When a fixed team is configured to carry out the work, it is impossible to guarantee the good completion of the field work order project. Sometimes, personnel need to be dispatched to assist, resulting in low efficiency. Summary of the invention
[0004] The present invention provides a work order management system based on deep learning to solve the technical problems in related technologies.
[0005] The present invention provides a work order management system based on deep learning, comprising:
[0006] Data collection module: collects work order management related information, including all field work order task information, employee skill information and employee collaboration ability information in a water project. Each field work order task information includes several subtasks. Employee skill information includes several skill data and employee attributes. Employee attributes include employee ID and employee attendance status. The collaboration information between employees includes the employee collaboration ability score, the number of employee collaborations and the successful collaboration ratio.
[0007] The successful collaboration ratio includes the team’s overall collaboration completion rate, communication frequency, and communication quality indicators;
[0008] The employees in the employee skill information belong to the same employee database. Each employee in the employee database has a unique employee ID, and each employee has one or more corresponding skill data.
[0009] Structured data generation module: converts the collected work order management related information into structured data representation, including No. 1 one-dimensional structure data, No. 1 two-dimensional structure data, No. 2 one-dimensional structure data and No. 2 two-dimensional structure data;
[0010] Deep learning model module: input the structural data into the deep learning model, which includes a first hidden layer, a second hidden layer, a first feature fusion layer, a third hidden layer, a fourth hidden layer, a second feature fusion layer and an output layer. The first two-dimensional structural data is input into the first hidden layer, and the first hidden layer outputs the first hidden state to the first feature fusion layer. At the same time, the second one-dimensional structure is input into the second hidden layer, and the second hidden layer outputs the second hidden state to the first feature fusion layer. The third hidden layer outputs the third hidden state through the first feature fusion layer. At the same time, the third hidden layer also inputs the first one-dimensional structural data and outputs the fourth hidden state to the second feature fusion layer. The fourth hidden layer inputs the second two-dimensional structural data and outputs the fifth hidden state to the second feature fusion layer. The sixth hidden state is output through the second feature fusion layer, and then the output layer outputs the employee assigned to each field work order subtask;
[0011] Work order team generation module: A corresponding field work order team is formed according to the employees assigned to each field work order subtask. The field work order team works on all field work orders in a water project and completes the entire water project.
[0012] Furthermore, the structure data generation module also includes the following contents:
[0013] Generate a second one-dimensional structure data from the collected employee skill information, the second one-dimensional structure data includes n data items sorted by time, and the tth data item represents the employee skill information collected at the tth moment;
[0014] Based on the sorting of the field work order task information, the first one-dimensional structure data is obtained, the first one-dimensional structure data includes n data items sorted by time, and the t-th data item represents the first two-dimensional structure data generated by the field work order task information collected at the t-th time;
[0015] The No. 1 two-dimensional structure data includes a No. 1 data matrix and a No. 1 relationship matrix. A unit of the No. 1 data matrix represents the field work order task information of a subtask. The element of the No. 1 relationship matrix in the i-th row and the j-th column represents the association between the subtasks represented by the i-th unit and the j-th unit of the No. 1 data matrix. If there is an association, the value of the element of the No. 1 relationship matrix is 1, otherwise it is 0.
[0016] The association between subtasks means that the subtasks belong to the same work order task, and each subtask is collaboratively performed by several employees;
[0017] Generate No. 2 two-dimensional structure data based on the information on the inter-employee collaboration capability, the No. 2 two-dimensional structure data includes a No. 2 data matrix and a No. 2 relationship matrix, a unit of the No. 2 data matrix represents the inter-employee collaboration capability information of an employee; an element of the No. ith row and the No. ith column of the No. 2 relationship matrix represents the association between the employees represented by the No. ith unit and the No. ith unit of the No. 2 data matrix, if there is an association, the value of the element of the No. 2 relationship matrix is 1, otherwise it is 0;
[0018] The existence of association between employees means that the employees have one or more skill information required in the subtask, and the employees collaborate to complete one or more subtasks.
[0019] Furthermore, the calculation formula of the first hidden layer is as follows:
[0020]
[0021] in represents the first hidden state of the vth unit of the data matrix No. 1, represents the state weight parameter of the first hidden layer, and They represent the field work order task information represented by the vth unit and the uth unit of the data matrix No. 1, respectively. represents the set of cells of data matrix No. 1 that are associated with the v-th cell of data matrix No. 1, express The total number of cells in ;
[0022]
[0023] Where tanh represents the hyperbolic tangent function, T represents the transpose, Indicates adjustable parameters; The default value is 1 / M.
[0024] Furthermore, the calculation formula of the second hidden layer is as follows:
[0025]
[0026] in represents the second hidden state, its cth component value represents the skill data of the cth employee, X represents the input employee skill information, Represents the convolution function.
[0027] Furthermore, the calculation formula of the first feature fusion layer is as follows:
[0028]
[0029] in, represents the tth third hidden state, It represents the first hidden state of the vth unit output when the tth data item of the one-dimensional structure data is input into the first hidden layer. represents the set of all cells of data matrix number one, represents the concatenation function, represents the summation weight parameter, Represents the summed bias parameter.
[0030] Furthermore, the calculation formula of the third hidden layer is as follows:
[0031]
[0032]
[0033]
[0034]
[0035] in, , , represents the first, second, and third weight parameters of the third hidden layer, , , represents the first, second, and third bias parameters of the third hidden layer, represents the dot product, , and Represent the first, second, and third intermediate states of the third hidden layer, respectively. , and Represent the t-th and t-1-th first hidden states respectively, n≥t≥1, n represents the total number of data items of the one-dimensional structure data, and when t=1 , tanh is the hyperbolic tangent function, Represents the sigmoid function.
[0036] Furthermore, the calculation formula of the fourth hidden layer is as follows:
[0037]
[0038] in represents the fifth hidden state of the vth unit of the second data matrix, is the set of cells of data matrix No. 2 that are associated with the vth cell of data matrix No. 2, represents the information on the inter-employee collaboration capability represented by the u-th unit of the second data matrix, is a normalizing constant (such as the unit degree of the data matrix), represents the weight parameter of the fourth hidden layer, represents the bias parameter of the fourth hidden layer, is the sigmoid function.
[0039] Furthermore, the calculation formula of the second feature fusion layer is as follows:
[0040]
[0041] in, represents the sixth hidden state, The nth fourth hidden state, represents the set of all cells of data matrix No. 2, represents the concatenation function, represents the summation weight parameter, Represents the summed bias parameter.
[0042] The present invention also proposes a work order management method based on deep learning, comprising:
[0043] Step 100 is used to collect work order management related information, including field work order task information, employee skill information, and employee collaboration capability information;
[0044] Step 200, converting the collected work order management related information into a structured data representation, including No. 1 one-dimensional structure data, No. 1 two-dimensional structure data, No. 2 one-dimensional structure data and No. 2 two-dimensional structure data;
[0045] Step 300, inputting structured data through an end-to-end deep neural network model, and outputting the employee assigned to each field work order subtask;
[0046] Step 400, according to the employee assignment information of each work order subtask output by the deep neural network model, a corresponding work order team is formed to collaborate to complete the field work order tasks of the entire project.
[0047] The present invention also proposes a storage medium storing non-temporary computer-readable instructions for executing one or more steps in the aforementioned deep learning-based work order management method.
[0048] The beneficial effects of the present invention are:
[0049] The present invention does not adopt a fixed team configuration, but dynamically generates the optimal personnel combination according to the specific needs of each field work order project in the project, subtask information, and employee skills and collaboration capabilities, and dynamically optimizes the team configuration, avoiding frequent personnel dispatches and saving a lot of manpower deployment costs. Through deep learning technology, the automation and intelligent deployment of field work order personnel is realized, which can greatly improve work efficiency and optimize the utilization of human resources, thereby bringing a double improvement in economic benefits and operational performance to water supply companies. BRIEF DESCRIPTION OF THE DRAWINGS
[0050] Figure 1 It is a flow chart of a work order management method based on deep learning proposed by the present invention;
[0051] Figure 2 It is a structural block diagram of a work order management system based on deep learning proposed by the present invention;
[0052] Figure 3 It is a model structure diagram of the deep learning model proposed in the present invention.
[0053] In the figure: 101, data collection module; 102, structure data generation module; 103, deep learning model module; 104, work order team generation module. DETAILED DESCRIPTION
[0054] The subject matter described herein will now be discussed with reference to example embodiments. It should be understood that the discussion of these embodiments is only to enable those skilled in the art to better understand and implement the subject matter described herein, and the functions and arrangements of the elements discussed may be changed without departing from the scope of protection of the contents of this specification. Each example may omit, replace or add various processes or components as needed. In addition, the features described relative to some examples may also be combined in other examples.
[0055] Example 1
[0056] like Figure 1-Figure 3 As shown, a work order management method based on deep learning includes:
[0057] Step 100, collecting work order management related information, which includes all field work order task information, employee skill information and employee collaboration capability information in a project, each field work order task information includes several subtasks, employee skill information includes several skill data and employee attributes, employee attributes include employee ID and employee attendance status, and employee collaboration information includes employee collaboration capability score, employee collaboration times and successful collaboration ratio;
[0058] Among them, the successful collaboration ratio includes the team's overall collaboration completion rate, communication frequency, and communication quality indicators;
[0059] The employees in the employee skill information belong to the same employee database. Each employee in the employee database has a unique employee ID, and each employee has one or more corresponding skill data.
[0060] It should be noted that the employee attendance status includes the attendance status and the absence status. When the employee attendance status is the absence status, the skill data of the corresponding employee in the employee skill information is zero;
[0061] Step 200, generating structure data from the collected work order management related information;
[0062] Generate a second one-dimensional structure data from the collected employee skill information, the second one-dimensional structure data includes n data items sorted by time, and the tth data item represents the employee skill information collected at the tth moment;
[0063] Based on the sorting of the field work order task information, the first one-dimensional structure data is obtained, the first one-dimensional structure data includes n data items sorted by time, and the t-th data item represents the first two-dimensional structure data generated by the field work order task information collected at the t-th time;
[0064] The No. 1 two-dimensional structure data includes a No. 1 data matrix and a No. 1 relationship matrix. A unit of the No. 1 data matrix represents the field work order task information of a subtask. The element of the No. 1 relationship matrix in the i-th row and the j-th column represents the association between the subtasks represented by the i-th unit and the j-th unit of the No. 1 data matrix. If there is an association, the value of the element of the No. 1 relationship matrix is 1, otherwise it is 0.
[0065] The association between subtasks means that the subtasks belong to the same work order task, and each subtask is collaboratively performed by several employees;
[0066] Generate No. 2 two-dimensional structure data based on the information on the inter-employee collaboration capability, the No. 2 two-dimensional structure data includes a No. 2 data matrix and a No. 2 relationship matrix, a unit of the No. 2 data matrix represents the inter-employee collaboration capability information of an employee; an element of the No. ith row and the No. ith column of the No. 2 relationship matrix represents the association between the employees represented by the No. ith unit and the No. ith unit of the No. 2 data matrix, if there is an association, the value of the element of the No. 2 relationship matrix is 1, otherwise it is 0;
[0067] The relationship between employees means that the employees have one or more skill information required in the subtask, and the employees collaborate to complete one or more subtasks;
[0068] Step 300, input the structure data into the deep learning model, the deep learning model includes a first hidden layer, a second hidden layer, a first feature fusion layer, a third hidden layer, a fourth hidden layer, a second feature fusion layer and an output layer, the first two-dimensional structure data is input into the first hidden layer, the first hidden layer outputs the first hidden state to the first feature fusion layer, at the same time, the second one-dimensional structure data is input into the second hidden layer, the second hidden layer outputs the second hidden state to the first feature fusion layer, and the third hidden state is output to the third hidden layer through the first feature fusion layer, at the same time, the third hidden layer also inputs the first one-dimensional structure data, outputs the fourth hidden state to the second feature fusion layer, the fourth hidden layer inputs the second two-dimensional structure data, outputs the fifth hidden state to the second feature fusion layer, and outputs the sixth hidden state through the second feature fusion layer, and then outputs the employee assigned to each field work order subtask through the output layer.
[0069] In one embodiment of the present invention, a corresponding field work order team is formed according to the employees assigned to each field work order subtask, and the field work order team works on all field work orders in a project, thereby completing the entire water project.
[0070] In one embodiment of the present invention, the calculation formula of the first hidden layer is as follows:
[0071]
[0072] in represents the first hidden state of the vth unit of the data matrix No. 1, represents the state weight parameter of the first hidden layer, and They represent the field work order task information represented by the vth unit and the uth unit of the data matrix No. 1, respectively. represents the set of cells of data matrix No. 1 that are associated with the v-th cell of data matrix No. 1, express The total number of cells in ;
[0073]
[0074] Where tanh represents the hyperbolic tangent function, T represents the transpose, Indicates adjustable parameters; The default value is 1 / M.
[0075] In one embodiment of the present invention, the calculation formula of the second hidden layer is as follows:
[0076]
[0077] in represents the second hidden state, its cth component value represents the skill data of the cth employee, X represents the input employee skill information, Represents the convolution function.
[0078] In one embodiment of the present invention, the calculation formula of the first feature fusion layer is as follows:
[0079]
[0080] in, represents the tth third hidden state, It represents the first hidden state of the vth unit output when the tth data item of the one-dimensional structure data is input into the first hidden layer. represents the set of all cells of data matrix number one, represents the concatenation function, represents the summation weight parameter, Represents the summation bias parameter.
[0081] In one embodiment of the present invention, the calculation formula of the third hidden layer is as follows:
[0082]
[0083]
[0084]
[0085]
[0086] in, , , represents the first, second, and third weight parameters of the third hidden layer, , , represents the first, second, and third bias parameters of the third hidden layer, represents the dot product, , and Represent the first, second, and third intermediate states of the third hidden layer, respectively. , and Represent the t-th and t-1-th first hidden states respectively, n≥t≥1, n represents the total number of data items of the one-dimensional structure data, and when t=1 , tanh is the hyperbolic tangent function, Represents the sigmoid function.
[0087] In one embodiment of the present invention, the calculation formula of the fourth hidden layer is as follows:
[0088]
[0089] in represents the fifth hidden state of the vth unit of the second data matrix, is the set of cells of data matrix No. 2 that are associated with the vth cell of data matrix No. 2, represents the information on the inter-employee collaboration capability represented by the u-th unit of the second data matrix, is a normalizing constant (such as the unit degree of the data matrix), represents the weight parameter of the fourth hidden layer, represents the bias parameter of the fourth hidden layer, is the sigmoid function.
[0090] In one embodiment of the present invention, the calculation formula of the second feature fusion layer is as follows:
[0091]
[0092] in, represents the sixth hidden state, The nth fourth hidden state, represents the set of all cells of data matrix No. 2, represents the concatenation function, represents the summation weight parameter, Represents the summed bias parameter.
[0093] In one embodiment of the present invention, the calculation formula of the output layer is as follows:
[0094]
[0095] in represents the output vector, whose cth component value represents the employee assigned to the subtask in the cth field work order. represents the kth sixth hidden state, is the weight parameter in the output layer, is the bias parameter in the output layer, Represents the sigmoid function.
[0096] like Figure 1-Figure 3 As shown, at least one embodiment of the present disclosure provides a work order management system based on deep learning, including:
[0097] Data collection module: used to collect information related to work order management, including field work order task information, employee skill information, and information on collaboration capabilities between employees.
[0098] Structural data generation module: converts the collected work order management related information into structured data representation, including No. 1 one-dimensional structural data, No. 1 two-dimensional structural data (data matrix and relationship matrix), No. 2 one-dimensional structural data and No. 2 two-dimensional structural data (data matrix and relationship matrix).
[0099] Deep learning model module: This module is the core of the system and contains an end-to-end deep neural network model that inputs the above structured data and ultimately outputs the employee assigned to each field work order subtask. The model includes multiple hidden layers, feature fusion layers, etc.
[0100] The main components of this deep learning model include:
[0101] The first hidden layer: inputs the first two-dimensional structure data and outputs the first hidden state;
[0102] The second hidden layer: inputs the second one-dimensional structure data and outputs the second hidden state;
[0103] The first feature fusion layer: fuses the first and second hidden states and outputs the third hidden state;
[0104] The third hidden layer: inputs the first one-dimensional structure data and the third hidden state, and outputs the fourth hidden state;
[0105] The fourth hidden layer: inputs the second two-dimensional structure data and outputs the fifth hidden state;
[0106] The second feature fusion layer: fuses the fourth and fifth hidden states and outputs the sixth hidden state;
[0107] Output layer: Output the employee information assigned to each work order subtask according to the sixth hidden state;
[0108] Work order team generation module: According to the employee information assigned to each work order subtask output by the model, a corresponding work order team is formed to collaborate to complete the field work order tasks of the entire project.
[0109] At least one embodiment of the present disclosure provides a storage medium storing non-transitory computer-readable instructions for executing one or more steps in the aforementioned deep learning-based work order management method.
[0110] The computer program may be stored / distributed on a suitable medium, such as an optical storage medium or a solid-state medium supplied together with or as part of other hardware, but may also be distributed in other forms, such as via the Internet or other wired or wireless telecommunication systems. Any reference signs in the claims should not be construed as limiting the scope.
[0111] In Example 1, an example of a work order management system based on deep learning is as follows:
[0112] Input data:
[0113] Two-dimensional structure data No. 1:
[0114] Data matrix No. 1:
[0115] Subtask 1: Replace water meter (skill requirements: water meter installation, pipe connection);
[0116] Subtask 2: Inspect pipelines (skill requirements: pipeline inspection, leak detection);
[0117] Subtask 3: Repair water leaks (skill requirements: pipe repair, welding);
[0118] Relationship Matrix No. 1: ;
[0119] One-dimensional structure data of No. 2:
[0120] Employee 1: water meter installation, pipe connection;
[0121] Employee 2: pipeline inspection, pipeline maintenance;
[0122] Employee 3: leak detection, welding;
[0123] Employee 4: water meter installation, pipeline maintenance;
[0124] One-dimensional structure data:
[0125] Work order task 1: subtask 1, subtask 2, subtask 3;
[0126] Two-dimensional structure data No. 2:
[0127] Data matrix No. 2:
[0128] Employee 1: collaboration ability score 85, collaboration times 20, successful collaboration rate 95%;
[0129] Employee 2: collaboration ability score 90, collaboration times 25, successful collaboration rate 98%;
[0130] Employee 3: collaboration ability score 80, collaboration times 18, successful collaboration rate 93%;
[0131] Employee 4: collaboration ability score 88, collaboration times 22, successful collaboration rate 96%;
[0132] Relationship Matrix No. 2: ;
[0133] Output:
[0134] Subtask 1 is assigned to employee 1 and employee 4;
[0135] Subtask 2 is assigned to employee 2 and employee 3;
[0136] Subtask 3 is assigned to employee 2 and employee 3;
[0137] Field work order team: Employee 1, Employee 2, Employee 3, Employee 4.
[0138] Example 2
[0139] In this embodiment, a work order management system based on deep learning is proposed, including:
[0140] Data collection module: collects work order management related information, including all field work order task information, employee skill information and employee collaboration ability information in a water project. Each field work order task information includes several subtasks. Employee skill information includes several skill data and employee attributes. Employee attributes include employee ID and employee attendance status. The collaboration information between employees includes the employee collaboration ability score, the number of employee collaborations and the successful collaboration ratio.
[0141] In Example 2, it should be supplemented that the task of the work order is to clean the water storage tank, and at the same time, the water quality needs to be tested to improve the water quality before and after cleaning;
[0142] Structured data generation module: converts the collected work order management related information into structured data representation;
[0143] Deep learning model module: inputs the structured data into the deep learning model, and outputs the employee assigned to each field work order subtask;
[0144] Work order team generation module: A corresponding field work order team is formed according to the employees assigned to each field work order subtask. The field work order team works on all field work orders in a water project and completes the entire water project.
[0145] In Example 2, a field work order team can eventually be obtained to complete the water storage tank cleaning operation. The field work order team also obtains a task allocation plan, which includes personnel allocation and task allocation;
[0146] In Example 2, an example of a work order management system based on deep learning is as follows:
[0147] Input data:
[0148] Two-dimensional structure data No. 1:
[0149] Data matrix No. 1:
[0150] Subtask 1: Cleaning the water storage tank (skill requirements: high-altitude work, tank wall cleaning);
[0151] Subtask 2: Water quality testing (skill requirements: water quality testing, data analysis);
[0152] Relationship Matrix No. 1: ;
[0153] One-dimensional structure data of No. 2:
[0154] Employee 1: high-altitude work, warehouse wall cleaning;
[0155] Employee 2: Water quality testing;
[0156] Employee 3: Working at height;
[0157] Employee 4: Data analysis;
[0158] Employee 5: water quality testing, data analysis;
[0159] One-dimensional structure data:
[0160] Work order task 1: subtask 1, subtask 2;
[0161] Two-dimensional structure data No. 2:
[0162] Data matrix No. 2:
[0163] Employee 1: collaboration ability score 82, collaboration times 18, successful collaboration rate 94%;
[0164] Employee 2: collaboration ability score 86, collaboration times 20, successful collaboration rate 96%;
[0165] Employee 3: collaboration ability score 80, collaboration times 16, successful collaboration rate 92%;
[0166] Employee 4: collaboration ability score 84, collaboration times 20, successful collaboration rate 93%;
[0167] Employee 5: collaboration ability score 90, collaboration times 22, successful collaboration rate 98%;
[0168] Relationship Matrix No. 2: ;
[0169] Output:
[0170] Subtask 1 is assigned to employee 1 and employee 3;
[0171] Subtask 2 is assigned to employee 5;
[0172] Field work order team: Employee 1, Employee 3, Employee 5.
[0173] Example 3
[0174] In this embodiment, a work order management system based on deep learning is proposed, including:
[0175] Data collection module: collects work order management related information, including all field work order task information, employee skill information and employee collaboration ability information in a water project. Each field work order task information includes several subtasks. Employee skill information includes several skill data and employee attributes. Employee attributes include employee ID and employee attendance status. The collaboration information between employees includes the employee collaboration ability score, the number of employee collaborations and the successful collaboration ratio.
[0176] In Example 3, it should be noted that the tasks of the work order are to clean the water storage tank, replace the filter, check the water supply pipe and repair the water supply pump. At the same time, the water quality needs to be tested to ensure that the water supply pipe can stably transport water, the pipes, filters and water pumps on the water supply route can operate normally, and the water quality in the temporary water storage tank and the water supply meet the water quality requirements;
[0177] Structured data generation module: converts the collected work order management related information into structured data representation;
[0178] Deep learning model module: inputs the structured data into the deep learning model, and outputs the employee assigned to each field work order subtask;
[0179] Work order team generation module: A corresponding field work order team is formed according to the employees assigned to each field work order subtask. The field work order team works on all field work orders in a water project and completes the entire water project.
[0180] In Example 3, a field work order team can eventually be obtained to complete many tasks on the work order. The field work order team also obtains a task allocation plan, which includes personnel allocation and task allocation in the team;
[0181] In Example 3, an example of a work order management method based on deep learning is as follows:
[0182] Input data:
[0183] Two-dimensional structure data No. 1:
[0184] Data matrix No. 1:
[0185] Subtask 1: Cleaning the water storage tank (skill requirements: high-altitude work, tank wall cleaning);
[0186] Subtask 2: Water quality testing (skill requirements: water quality testing, data analysis);
[0187] Subtask 3: Replace filter (skills required: mechanical maintenance, equipment installation);
[0188] Subtask 4: Check water supply pipes (skills required: pipe inspection, leak detection);
[0189] Subtask 5: Repair water pump (skill requirements: mechanical maintenance, electrical maintenance);
[0190] Relationship Matrix No. 1: ;
[0191] One-dimensional structure data of No. 2:
[0192] Employee 1: high-altitude work, warehouse wall cleaning, machinery maintenance;
[0193] Employee 2: water quality testing, data analysis;
[0194] Employee 3: high-altitude work, pipeline inspection;
[0195] Employee 4: data analysis, electrical maintenance;
[0196] Employee 5: water quality testing, machinery maintenance, equipment installation;
[0197] Employee 6: leak detection, pipeline inspection;
[0198] Employee 7: Mechanical maintenance, electrical maintenance, equipment installation;
[0199] One-dimensional structure data:
[0200] Work order task 1: subtask 1, subtask 2, subtask 3, subtask 4, subtask 5;
[0201] Two-dimensional structure data No. 2:
[0202] Data matrix No. 2:
[0203] Employee 1: collaboration ability score 85, collaboration times 25, successful collaboration rate 96%;
[0204] Employee 2: collaboration ability score 88, collaboration times 22, successful collaboration rate 95%;
[0205] Employee 3: collaboration ability score 82, collaboration times 20, successful collaboration rate 94%;
[0206] Employee 4: collaboration ability score 84, collaboration times 23, successful collaboration rate 93%;
[0207] Employee 5: collaboration ability score 90, collaboration times 28, successful collaboration rate 98%;
[0208] Employee 6: collaboration ability score 86, collaboration times 24, successful collaboration rate 95%;
[0209] Employee 7: collaboration ability score 92, collaboration times 30, successful collaboration rate 97%;
[0210] Relationship Matrix No. 2: ;
[0211] Output:
[0212] Subtask 1 is assigned to employee 1 and employee 3;
[0213] Subtask 2 is assigned to employee 2 and employee 5;
[0214] Subtask 3 is assigned to employee 5 and employee 7;
[0215] Subtask 4 is assigned to employee 2 and employee 6;
[0216] Subtask 5 is assigned to employee 4 and employee 7;
[0217] Field work order team: Employee 1, Employee 2, Employee 3, Employee 4, Employee 5, Employee 6, Employee 7.
[0218] Example 4
[0219] In this embodiment, a work order management system based on deep learning is proposed, including:
[0220] Data collection module: collects work order management related information, including all field work order task information, employee skill information and employee collaboration ability information in a water project. Each field work order task information includes several subtasks. Employee skill information includes several skill data and employee attributes. Employee attributes include employee ID and employee attendance status. The collaboration information between employees includes the employee collaboration ability score, the number of employee collaborations and the successful collaboration ratio.
[0221] In Example 4, it should be noted that the task of the work order is the laying of a pipe network, and the tasks in the laying of the pipe network include but are not limited to pipe laying, valve installation, water pressure testing, pipe cleaning, water quality sampling, and leakage detection, so that the laying of the pipe network can be completed efficiently, and there is no leakage in the laid pipe network, and the water quality output through the pipe network also meets the requirements;
[0222] Structured data generation module: converts the collected work order management related information into structured data representation;
[0223] Deep learning model module: inputs the structured data into the deep learning model, and outputs the employee assigned to each field work order subtask;
[0224] Work order team generation module: A corresponding field work order team is formed according to the employees assigned to each field work order subtask. The field work order team works on all field work orders in a water project and completes the entire water project.
[0225] In Example 4, a field work order team can eventually be obtained to complete many tasks on the work order. The field work order team also obtains a task allocation plan, which includes personnel allocation and task allocation in the team;
[0226] In Example 4, an example of a work order management method based on deep learning is as follows:
[0227] Input data:
[0228] Two-dimensional structure data No. 1:
[0229] Data matrix No. 1:
[0230] Subtask 1: Pipeline laying (skill requirements: pipeline installation, welding);
[0231] Subtask 2: Valve installation (skill requirements: valve installation, pipe connection);
[0232] Subtask 3: Water pressure test (skill requirements: water pressure test, data recording);
[0233] Subtask 4: Pipeline cleaning (skill requirements: pipeline cleaning, chemical treatment);
[0234] Subtask 5: Water quality sampling (skill requirements: water quality sampling, laboratory analysis);
[0235] Subtask 6: Leak detection (skill requirements: leak detection, pipeline inspection);
[0236] Relationship Matrix No. 1: ;
[0237] One-dimensional structure data of No. 2:
[0238] Employee 1: Pipeline installation, welding, valve installation;
[0239] Employee 2: water pressure testing, data recording, water quality sampling;
[0240] Employee 3: Pipe cleaning, chemical treatment;
[0241] Employee 4: Water sampling, laboratory analysis, leak detection;
[0242] Employee 5: Pipeline installation, pipe connection, leak detection;
[0243] Employee 6: valve installation, pipeline connection, pipeline inspection;
[0244] Employee 7: water pressure testing, pipeline inspection, data recording;
[0245] Employee 8: Pipe cleaning, chemical treatment, water sampling;
[0246] One-dimensional structure data:
[0247] Work order task 1: subtask 1, subtask 2, subtask 3, subtask 4, subtask 5, subtask 6;
[0248] Two-dimensional structure data No. 2:
[0249] Data matrix No. 2:
[0250] Employee 1: collaboration ability score 90, collaboration times 30, successful collaboration rate 98%;
[0251] Employee 2: collaboration ability score 88, collaboration times 28, successful collaboration rate 96%;
[0252] Employee 3: collaboration ability score 85, collaboration times 25, successful collaboration rate 95%;
[0253] Employee 4: collaboration ability score 92, collaboration times 32, successful collaboration rate 97%;
[0254] Employee 5: collaboration ability score 89, collaboration times 27, successful collaboration rate 96%;
[0255] Employee 6: collaboration ability score 87, collaboration times 26, successful collaboration rate 95%;
[0256] Employee 7: collaboration ability score 91, collaboration times 29, successful collaboration rate 97%;
[0257] Employee 8: collaboration ability score 86, collaboration times 24, successful collaboration rate 94%;
[0258] Relationship Matrix No. 2: ;
[0259] Output:
[0260] Subtask 1 is assigned to employee 1 and employee 5;
[0261] Subtask 2 is assigned to employee 1 and employee 6;
[0262] Subtask 3 is assigned to employee 2 and employee 7;
[0263] Subtask 4 is assigned to employee 3 and employee 8;
[0264] Subtask 5 is assigned to employee 2 and employee 4;
[0265] Subtask 6 is assigned to employee 5 and employee 7;
[0266] Field work order team: Employee 1, Employee 2, Employee 3, Employee 4, Employee 5, Employee 6, Employee 7, Employee 8.
[0267] Examples 1 to 4 all optimize personnel allocation through a deep learning model. The deep learning model comprehensively considers multiple factors such as employee skills, collaboration capabilities, task relevance, etc., and rationally allocates tasks to the most suitable combination of employees to achieve the goal of efficiently completing field work orders.
[0268] The above describes an embodiment of the present embodiment, but the present embodiment is not limited to the above-mentioned specific implementation methods. The above-mentioned specific implementation methods are merely illustrative and not restrictive. Under the guidance of the present embodiment, ordinary technicians in this field can also make many forms, all of which are protected by the present embodiment.
Claims
1. A work order management system based on deep learning, characterized in that: include: Data collection module: collects work order management related information, including all field work order task information, employee skill information and employee collaboration ability information in a water project. Each field work order task information includes several subtasks. Employee skill information includes several skill data and employee attributes. Employee attributes include employee ID and employee attendance status. The collaboration information between employees includes the employee collaboration ability score, the number of employee collaborations and the successful collaboration ratio. The successful collaboration ratio includes the team’s overall collaboration completion rate, communication frequency, and communication quality indicators; The employees in the employee skill information belong to the same employee database. Each employee in the employee database has a unique employee ID, and each employee has one or more corresponding skill data. Structured data generation module: converts the collected work order management related information into structured data representation, including No. 1 one-dimensional structure data, No. 1 two-dimensional structure data, No. 2 one-dimensional structure data and No. 2 two-dimensional structure data; Deep learning model module: input the structural data into the deep learning model, which includes a first hidden layer, a second hidden layer, a first feature fusion layer, a third hidden layer, a fourth hidden layer, a second feature fusion layer and an output layer. The first two-dimensional structural data is input into the first hidden layer, and the first hidden layer outputs the first hidden state to the first feature fusion layer. At the same time, the second one-dimensional structure is input into the second hidden layer, and the second hidden layer outputs the second hidden state to the first feature fusion layer. The third hidden layer outputs the third hidden state through the first feature fusion layer. At the same time, the third hidden layer also inputs the first one-dimensional structural data and outputs the fourth hidden state to the second feature fusion layer. The fourth hidden layer inputs the second two-dimensional structural data and outputs the fifth hidden state to the second feature fusion layer. The sixth hidden state is output through the second feature fusion layer, and then the output layer outputs the employee assigned to each field work order subtask; Work order team generation module: A corresponding field work order team is formed according to the employees assigned to each field work order subtask. The field work order team works on all field work orders in a water project and completes the entire water project.
2. A work order management system based on deep learning according to claim 1, characterized in that: The structure data generation module also includes the following: Generate a second one-dimensional structure data from the collected employee skill information, the second one-dimensional structure data includes n data items sorted by time, and the tth data item represents the employee skill information collected at the tth moment; Based on the sorting of the field work order task information, the first one-dimensional structure data is obtained, the first one-dimensional structure data includes n data items sorted by time, and the t-th data item represents the first two-dimensional structure data generated by the field work order task information collected at the t-th time; The No. 1 two-dimensional structure data includes a No. 1 data matrix and a No. 1 relationship matrix. A unit of the No. 1 data matrix represents the field work order task information of a subtask. The element in the i-th row and j-th column of the first relationship matrix represents the association between the i-th unit and the subtask represented by the j-th unit of the first data matrix. If there is an association, the value of this element of the first relationship matrix is 1, otherwise it is 0; The association between subtasks means that the subtasks belong to the same work order task, and each subtask is collaboratively performed by several employees; Generate No. 2 two-dimensional structure data based on the information on the inter-employee collaboration capability, the No. 2 two-dimensional structure data includes a No. 2 data matrix and a No. 2 relationship matrix, a unit of the No. 2 data matrix represents the inter-employee collaboration capability information of an employee; an element of the No. ith row and the No. ith column of the No. 2 relationship matrix represents the association between the employees represented by the No. ith unit and the No. ith unit of the No. 2 data matrix, if there is an association, the value of the element of the No. 2 relationship matrix is 1, otherwise it is 0; The existence of association between employees means that the employees have one or more skill information required in the subtask, and the employees collaborate to complete one or more subtasks.
3. A work order management system based on deep learning according to claim 2, characterized in that: The calculation formula for the first hidden layer is as follows: ; in represents the first hidden state of the vth unit of the data matrix No. 1, represents the state weight parameter of the first hidden layer, and They represent the field work order task information represented by the vth unit and the uth unit of the data matrix No. 1, respectively. represents the set of cells of data matrix No. 1 that are associated with the v-th cell of data matrix No. 1, express The total number of cells in ; ; Where tanh represents the hyperbolic tangent function, T represents the transpose, Indicates adjustable parameters; The default value is 1 / M.
4. A work order management system based on deep learning according to claim 3, characterized in that: The calculation formula for the second hidden layer is as follows: ; in represents the second hidden state, its cth component value represents the skill data of the cth employee, X represents the input employee skill information, Represents the convolution function.
5. A work order management system based on deep learning according to claim 4, characterized in that: The calculation formula of the first feature fusion layer is as follows: ; in, represents the tth third hidden state, It represents the first hidden state of the vth unit output when the tth data item of the one-dimensional structure data is input into the first hidden layer. represents the set of all cells of data matrix number one, represents the concatenation function, represents the summation weight parameter, Represents the summed bias parameter.
6. A work order management system based on deep learning according to claim 5, characterized in that: The calculation formula for the third hidden layer is as follows: ; ; ; ; in, , , represents the first, second, and third weight parameters of the third hidden layer, , , represents the first, second, and third bias parameters of the third hidden layer, represents the dot product, , and Represent the first, second, and third intermediate states of the third hidden layer, respectively. , and Represent the t-th and t-1-th first hidden states respectively, n≥t≥1, n represents the total number of data items of the one-dimensional structure data, and when t=1 , tanh is the hyperbolic tangent function, Represents the sigmoid function.
7. A work order management system based on deep learning according to claim 6, characterized in that: The calculation formula for the fourth hidden layer is as follows: ; in represents the fifth hidden state of the vth unit of the second data matrix, is the set of cells of data matrix No. 2 that are associated with the vth cell of data matrix No. 2, represents the information on the inter-employee collaboration capability represented by the u-th unit of the second data matrix, is a normalizing constant, represents the weight parameter of the fourth hidden layer, represents the bias parameter of the fourth hidden layer, is the sigmoid function.
8. A work order management system based on deep learning according to claim 7, characterized in that: The calculation formula of the second feature fusion layer is as follows: ; in, represents the sixth hidden state, The nth fourth hidden state, represents the set of all cells of data matrix No. 2, represents the concatenation function, represents the summation weight parameter, Represents the summed bias parameter.
9. A work order management method based on deep learning, characterized in that: Executed by one or more modules in a deep learning-based work order management system as described in any one of claims 1 to 8, including: Step 100 is used to collect work order management related information, including field work order task information, employee skill information, and employee collaboration capability information; Step 200, converting the collected work order management related information into a structured data representation, including No. 1 one-dimensional structure data, No. 1 two-dimensional structure data, No. 2 one-dimensional structure data and No. 2 two-dimensional structure data; Step 300, inputting structured data through an end-to-end deep neural network model, and outputting the employee assigned to each field work order subtask; Step 400, according to the employee assignment information of each work order subtask output by the deep neural network model, a corresponding work order team is formed to collaborate to complete the field work order tasks of the entire project.
10. A storage medium, characterized in that: Non-transitory computer-readable instructions are stored for executing one or more steps in a deep learning-based work order management method as described in claim 9.
Citation Information
Cited By
Traffic automobile image processing system and method based on deep learning
CN119693891A
