Search

CN-121981476-A - Intelligent labor intensity regulation method and system based on multi-factor cost prediction

CN121981476ACN 121981476 ACN121981476 ACN 121981476ACN-121981476-A

Abstract

The invention discloses an intelligent labor intensity control method and system based on multi-factor cost prediction, and belongs to the technical field of labor intensity control. According to the method, the states of the worksheets and personnel are obtained in real time, a multi-factor dynamic cost model for comprehensively considering time in transit, skill matching degree, worksheet priority and service time limit SLA risk is constructed, and cost indexes are calculated for each potential scheduling scheme. The system adopts a variable neighborhood search VNS optimization algorithm to carry out global solution on the basis of a cost matrix so as to find an optimal task allocation scheme with the lowest total cost. According to the method, the optimization weight can be dynamically adjusted according to the operation states of real-time traffic, personnel load and the like, so that self-adaptive intelligent decision is realized, and finally, an optimization scheme is automatically distributed to personnel terminals, so that the labor-monotone efficiency, the resource utilization rate and the SLA performance rate are comprehensively improved.

Inventors

  • Zhong Zicheng
  • WANG YUANDONG

Assignees

  • 深圳市一应科技有限公司

Dates

Publication Date
20260505
Application Date
20260126

Claims (9)

  1. 1. The intelligent labor intensity modulation method based on multi-factor cost prediction is characterized by comprising the following steps: s100, acquiring a work order set to be processed and an adjustable personnel state set in real time; S200, constructing a multi-factor dynamic cost prediction model, and selecting a work order With personnel Matching every two of the two scheduling schemes and respectively calculating the cost index of each scheduling scheme; S300, constructing a cost matrix according to the cost index of each scheduling scheme, carrying out global optimization scheduling solution by adopting a variable neighborhood search VNS algorithm, and outputting an optimal labor-monotone scheme; And S400, distributing the output optimal labor intensity modulation scheme to the corresponding personnel terminal.
  2. 2. The intelligent industrial personal computer method based on multi-factor cost prediction according to claim 1, wherein in S100: each work order information in the work order set at least comprises a work order position, a required skill type and level, a priority coefficient and a service level agreement SLA time limit; Each person information in the person state set includes at least a real-time location, a skill matrix, and a current task load.
  3. 3. The intelligent industrial personal computer method based on multi-factor cost prediction according to claim 2, wherein in S200, the cost index is The calculation formula is as follows: ; Wherein, the For in-transit cost, based on personnel Real-time location and worksheet of (c) The estimated travel time is calculated by combining the work order position of the road condition data in real time; For basic skill mismatch cost, according to work order The type and level of skill required, and personnel A reference cost value for performing matching calculation on the skill matrix of the (a); to quantify the skill level difference according to the work order The required skill level and personnel The difference between the actual skill levels of the skill levels is calculated; estimating wait time for work order, referring to work order From the current moment to the personnel to be subjected The total waiting time required to begin processing; For SLA time margin, refer to predicted work order Time difference between completion time and its SLA time limit; Is a work order Is a preset constant with a value larger than 0; 、 、 And For dynamically adjusting the weight coefficient, adjusting according to the operation state; 、 And And the preset constant is larger than 0 and is used for controlling the skill gap amplifying effect, the waiting cost nonlinear increasing rate and the SLA risk nonlinear increasing rate respectively.
  4. 4. The intelligent worker-to-machine method based on multi-factor cost prediction of claim 3, wherein said base skills do not match cost Skill level gap quantization value Further adopting a regression model trained based on historical work order processing data to predict; the input features of the regression model include at least work order type, skill level required, skill level of the processor and historical average processing time.
  5. 5. The intelligent industrial personal computer method based on multi-factor cost prediction according to claim 3, wherein said dynamically adjusting weight coefficients 、 、 And The adjustment is performed according to the following operation states: When the average traffic congestion index reflected by the real-time road condition data exceeds a first preset threshold value, the weight coefficient is improved Is a value of (2); when the global load rate of the personnel with specific skills exceeds a second preset threshold value, the weight coefficient is improved Is a value of (2); When the backlog proportion of the high-priority worksheets exceeds a third preset threshold value, the weight coefficient is improved Is a value of (2); When the overall prediction SLA breach risk exceeds a fourth preset threshold, the weight coefficient is increased Is a value of (2).
  6. 6. The intelligent industrial personal computer method based on multi-factor cost prediction according to claim 2, wherein in S300, the global optimization scheduling solution is performed by adopting a variable neighborhood search VNS algorithm, and the method specifically comprises the following steps: S301, generating an initial scheduling scheme by adopting a greedy algorithm, wherein the specific strategy is to traverse all unallocated work orders, pair the work order with the lowest current cost index with personnel to determine an allocation relation, synchronously update personnel states, and iterate until all work orders are allocated or no allocable personnel; S302, defining a group of neighborhood operations for carrying out structural transformation on the scheduling scheme so as to form a neighborhood structure set Exchanging task sets distributed by two different persons, reassigning a work order from one person to another person, and internally rearranging task execution sequences distributed by the same person; s303, executing a variable neighborhood search main loop: Step a, randomly selecting a neighborhood structure from the neighborhood structure set N Performing the neighborhood operation of random times on the current optimal scheduling scheme to generate a new solution ; Step b, in the new solution As starting point, sequentially searching deterministically in the neighborhood structure set, and searching deterministically in the neighborhood structure Searching for neighborhood operations that reduce the total cost and performing; Step c, if found, updating the current solution and re-locating If in If the better solution cannot be found, then the method is sequentially carried out in the following steps , , ..., Is searched until a locally optimal solution is found ; Step d, local optimal solution Is lower than the total cost of the current optimal scheduling scheme, the current optimal scheduling scheme is updated to Returning to the step a to perform a new disturbance, otherwise, continuing to perform the disturbance in the next neighborhood structure; And S304, terminating the main loop when the preset iteration times or the calculated time threshold value are reached, and outputting the current optimal scheduling scheme as a final optimal work station scheduling scheme.
  7. 7. An intelligent industrial personal computer system based on multi-factor cost prediction, for implementing the intelligent industrial personal computer method based on multi-factor cost prediction as set forth in claim 1, comprising: the data input module is used for acquiring and maintaining work cell data and personnel state cell data in real time; the multi-factor dynamic cost prediction module is connected with the data input module and is used for pairing each pair of worksheet-personnel into a scheduling scheme and calculating a cost index; The system state monitoring module is used for monitoring the operation state of the system in real time, wherein the operation state at least comprises a real-time traffic congestion index, a specific skill personnel load rate, a high-priority work order backlog proportion and an overall SLA (service level agreement) performance risk; The dynamic weight adjustment module is connected with the system state monitoring module and the multi-factor dynamic cost prediction module and is used for dynamically adjusting each weight coefficient in the cost index formula according to the monitored system operation state; The variable neighborhood search optimizing and scheduling module is connected with the multi-factor dynamic cost prediction module and is used for receiving the cost matrix constructed by the cost index, executing a variable neighborhood search VNS algorithm and solving an optimal labor-hour algorithm proposal; The instruction dispatch module is connected with the variable neighborhood search optimization scheduling module and used for issuing the optimal labor-hour management scheme to terminal equipment of corresponding personnel.
  8. 8. An electronic device comprising a memory, a processor and a computer program stored on the memory and executable on the processor, characterized in that the processor implements the steps of the method according to any one of claims 1 to 6 when the program is executed.
  9. 9. A computer readable storage medium, on which a computer program is stored, characterized in that the program, when being executed by a processor, implements the steps of the method according to any one of claims 1 to 6.

Description

Intelligent labor intensity regulation method and system based on multi-factor cost prediction Technical Field The invention relates to the technical field of worksheet management, in particular to an intelligent worksheet management method and system based on multi-factor cost prediction. Background In modern service and manufacturing operations, the industrial level is used as a core link for connecting the demand with the on-site service resources, and the efficiency directly determines the service response speed, the resource utilization rate and the operation cost. With the continuous expansion of service scale and increasing demand for service instantaneity, a scheduling system is required to realize real-time, globally optimal decision-making in a highly dynamic environment, which has become a key challenge in the industry digital transformation process. At present, the current mainstream scheduling technology has certain limitations. Firstly, the comprehensive quantification of multidimensional nonlinear cost factors such as skill matching degree, work order emergency degree, default risk and the like is difficult to realize mainly depending on static rules or single cost factors. Secondly, the optimization algorithm is easy to fall into a local optimal solution, and the accurate algorithm is difficult to meet the requirement of large-scale real-time calculation. Third, the system lacks learning ability based on historical data, and can not effectively predict implicit costs such as skill gaps. Fourth, each functional module is in a loose coupling state, and a closed loop optimization system of perception-decision-execution cannot be constructed, so that a scheduling strategy is stiff, and self-adapting to dynamically-changed service pressure is difficult. Disclosure of Invention The invention aims to provide an intelligent labor intensity modulation method and system based on multi-factor cost prediction, so as to solve the problems in the background technology. In order to solve the technical problems, the invention provides an intelligent industrial personal computer method based on multi-factor cost prediction, which comprises the following steps: and S100, acquiring a work order set to be processed and an adjustable personnel state set in real time. Each of the work order information in the work order set includes at least a work order location, a desired skill type and level, a priority coefficient, and a service level agreement SLA time limit. The SLA time limit and the priority coefficient of the work order together define the emergency degree of the work order, and are the core input of time penalty items in the follow-up cost model. Each person information in the person state set includes at least a real-time location, a skill matrix, and a current task load. The current task load determines the estimated waiting time of the work order, and directly influences the fairness and efficiency of scheduling. The real-time position of the personnel needs to be combined with external map API data, which is the key of the dynamic on-road cost for subsequent calculation. The real-time and comprehensive data base is established for the whole dispatching system, timeliness and integrity of information according to which the decision is optimized are ensured, and the method is a precondition for all subsequent calculation and optimization. S200, constructing a multi-factor dynamic cost prediction model, and selecting a work orderWith personnelTwo pairs of the cost indexes are matched and used as scheduling schemes, and the cost index of each scheduling scheme is calculated respectively: ; The cost index comprehensively converts the complex business influence involved in matching a potential person with a work order into a single numerical value which can be quantitatively compared. Various key costs and risks in scheduling decisions are finely characterized and balanced through a multi-dimensional weighting and nonlinear amplification mechanism. Wherein, the For in-transit cost, based on personnelReal-time location and worksheet of (c)The estimated travel time is calculated by combining the real-time road condition data. The in-transit cost function is to evaluate efficiency losses due to personnel going to the worksheet location. The method is not used for simply calculating the linear distance, but is used for dynamically reflecting the instant influence of the external environment on the scheduling efficiency by combining the real-time traffic road condition prediction travel time. For basic skill mismatch cost, according to work orderThe type and level of skill required, and personnelA reference cost value for performing a matching calculation. Skill mismatch costs act to quantify the inefficiency or quality risk of processing that may result from a mismatch in personnel skills and work order requirements. Including basic mismatch costs, and can be scaled up or down according to the level of gap be