CN-122022413-A - Multi-target scheduling data processing method for distributed wire cutting machine group

CN122022413ACN 122022413 ACN122022413 ACN 122022413ACN-122022413-A

Abstract

The invention relates to the field of intelligent manufacturing and industrial data processing, in particular to a multi-target production scheduling data processing method for a distributed wire cutting machine group; the method comprises the steps of collecting task instruction data, equipment state time sequence data and actual execution time sequence data containing energy consumption records, carrying out time alignment and data integration to generate a reference production scheduling data set, constructing an ideal constraint model to generate an ideal production scheduling data set containing task starting time, task ending time and resource consumption tracks, correcting beat, waiting and resource occupation parameters based on recessive loss factors and cooperative conflict factors, simulating to generate a post-disturbance production state, respectively extracting a real residual vector and a theoretical residual vector, determining production deviation factors based on coupling similarity of the real residual vector and the theoretical residual vector, and outputting rescheduling, maintaining or rechecking instructions.

Inventors

YAO MING
DENG SHAOWEI
YANG XIAOBAO

Assignees

厦门姚明织带饰品有限公司

Dates

Publication Date: 20260512
Application Date: 20260414

Claims (10)

1. The multi-target production scheduling data processing method for the distributed wire cutting machine group is characterized by comprising the following steps of: Collecting task instruction data, equipment state time sequence data and actual execution time sequence data containing energy consumption records, performing time alignment and data integration, and generating a reference scheduling data set; An ideal constraint model is built based on a reference scheduling data set, and an ideal scheduling time sequence data set is generated, wherein the ideal scheduling time sequence data set comprises task starting time, task ending time, a resource consumption track, and associated beat parameters, waiting parameters and resource occupation parameters; Correcting beat parameters, waiting parameters and resource occupation parameters in an ideal scheduling time sequence data set based on preset recessive loss factors, cooperative conflict factors and combinations of the two, and generating various post-disturbance scheduling states through simulation, wherein the recessive loss factors are used for representing equipment beat attenuation and energy efficiency reduction, and the cooperative conflict factors are used for representing waiting effects caused by material distribution delay and resource contention; extracting a real residual vector based on the actual execution time sequence data and the ideal production scheduling time sequence data set, and extracting a plurality of theoretical residual vectors based on the production scheduling states after a plurality of disturbances and the ideal production scheduling time sequence data set, wherein the real residual vector and the plurality of theoretical residual vectors at least comprise a time delay residual and a resource consumption residual; and determining a production deviation cause based on coupling similarity of the real residual vector and the theoretical residual vectors, and outputting a rescheduling instruction, a maintenance instruction, a holding instruction or a rechecking instruction according to the determination result.
2. The method for processing multi-target production scheduling data for a distributed wire cutting cluster according to claim 1, further comprising, before integrating task instruction data, device state timing data, and actual execution timing data: Preprocessing task instruction data, equipment state time sequence data and actual execution time sequence data to obtain preprocessed data, wherein the preprocessing comprises missing value filling, abnormal value detection, time stamp alignment and data standardization.
3. The distributed line cutting cluster oriented multi-target production scheduling data processing method of claim 1, wherein constructing an ideal constraint model based on a reference production scheduling data set, generating an ideal production scheduling data set, comprises: Setting the health state of the equipment to correspond to the rated processing beat; setting the network issuing delay as a pre-calibrated reference transmission delay; Setting the material switching time length as a pre-calibrated standard switching time length; and solving the integrated data based on the mixed integer linear programming to generate an ideal scheduling time sequence data set.
4. The distributed wire-cutting cluster-oriented multi-target production scheduling data processing method according to claim 1, wherein the post-disturbance production scheduling state is generated by simulation, comprising: correcting beat parameters, waiting parameters and resource occupation parameters in an ideal scheduling time sequence data set according to a preset recessive loss factor and a preset cooperative conflict factor; Determining equipment health scores according to cutter wear data, fault interval duration or beat fluctuation indexes, and mapping the equipment health scores into beat attenuation factors; Mapping the material delivery delay and the resource contention status to a queue blocking factor; based on discrete event simulation, forward deduction is carried out on the corrected scheduling process, and a post-disturbance scheduling state is generated.
5. The multi-target production data processing method for a distributed wire cutting cluster according to claim 1, wherein extracting a real residual vector and a theoretical residual vector comprises: extracting a real residual vector based on the acquired actual execution time sequence data and the ideal production scheduling time sequence data set; Extracting a theoretical residual vector based on the post-disturbance production scheduling state and the ideal production scheduling data set; Wherein, the real residual vector and the theoretical residual vector comprise a time delay residual, a resource consumption residual and an energy consumption fluctuation residual.
6. The multi-target scheduling data processing method for the distributed line cutting machine group according to claim 5, wherein the coupling similarity of the real residual vector and the theoretical residual vector is obtained by respectively carrying out normalization processing on a time delay residual, a resource consumption residual and an energy consumption fluctuation residual by utilizing a maximum tolerance threshold value calibrated in advance based on dynamic time warping, and then carrying out time sequence alignment on the normalized data; mapping the aligned residual data into a topological structure representing the association relation of task nodes; And obtaining the coupling similarity by weighting calculation based on the time sequence alignment distance and the topological structure matching degree.
7. The method for processing multi-target production data for a distributed wire-cutting cluster according to claim 6, wherein determining production deviation causes based on coupling similarities comprises: Extracting a decision boundary based on the historical production abnormality sample distribution as a first threshold and a second threshold, and setting the first threshold to be greater than the second threshold; If the coupling similarity between the actual residual vector and a theoretical residual vector corresponding to the hidden loss factor in the multiple theoretical residual vectors is highest, judging that the current deviation corresponds to the hidden loss factor, and outputting a maintenance instruction; If the coupling similarity between the actual residual vector and the theoretical residual vector corresponding to the cooperative conflict factor in the multiple theoretical residual vectors is highest, judging that the current deviation corresponds to the cooperative conflict factor, and outputting a rescheduling instruction; When the coupling similarity is smaller than or equal to a second threshold value, judging that the current deviation corresponds to the business disturbance, and outputting a hold instruction, wherein the business disturbance comprises a newly added work order, a work order priority change, a traffic period adjustment or order cancellation; And when the coupling similarity is larger than the second threshold value and smaller than the first threshold value, outputting a rechecking instruction, and re-executing differential extraction and coupling judgment based on the updated equipment state time sequence data.
8. The multi-target scheduling data processing method for the distributed wire cutting machine group according to claim 1, wherein when the method is applied to a scene of the distributed wire cutting machine group, task instruction data comprises work order data, line data, option weight data, wire harness length data and quantity data; The equipment state time sequence data comprise cutting beat data, cutter abrasion data, wire coil allowance data and standby idle time data; the actual execution time sequence data comprises actual start time, actual finish time, actual resource occupation record and actual energy consumption record.
9. The multi-target scheduling data processing method for the distributed wire cutting cluster according to claim 4, wherein the cooperative conflict factors comprise a material distribution delay factor and a node resource preemption factor; the node resource preemption factors are used for representing the competition relation of a plurality of execution nodes to materials with the same specification; The queue blocking factor is used to characterize the latency variation caused by the material delivery delay factor and the node resource preemption factor.
10. The multi-target production scheduling data processing method for a distributed wire cutting cluster according to any one of claims 1 to 9, wherein after outputting the rescheduling instruction, the maintenance instruction, the hold instruction or the rechecking instruction, further comprising: Writing back the execution result to the reference scheduling data set; Updating a hidden loss factor and a cooperative conflict factor according to a preset recursive correction rule or a sliding window statistical rule based on the execution result after the write-back; and regenerating the post-disturbance production scheduling state based on the updated factors to form a closed-loop self-adaptive processing flow.

Description

Multi-target scheduling data processing method for distributed wire cutting machine group Technical Field The invention relates to the field of intelligent manufacturing and industrial data processing, in particular to a multi-target production scheduling data processing method for a distributed wire cutting machine group. Background In the scene of a distributed wire cutting machine such as automobile wire harness production, a plurality of wire cutting devices need to cooperatively arrange production around different wire work orders, exchange period requirements and material resources, and the arrangement result directly relates to the device utilization rate, order standard exchange rate and production process stability, so that accurate data processing and deviation recognition are carried out on the arrangement process, and the method is an important basis for guaranteeing the efficient operation of workshops; The existing production scheduling analysis method has a plurality of problems, for example, the judgment is carried out by depending on static rules, manual experience or simple work order distribution results, the task planning and the equipment availability are usually only concerned, the unified alignment and integration of task instruction data, equipment state time sequence data and actual execution time sequence data are difficult, especially when hidden factors such as equipment beat attenuation, tool sub-health, material distribution delay and common specification resource contention exist continuously, deviation sources between ideal production and actual execution cannot be effectively described, and differences between equipment loss, cooperative conflict and normal business disturbance are difficult to distinguish, so that rescheduling, maintenance or maintenance decision lack of pertinence is easy to cause. Disclosure of Invention In order to solve the technical problems, the invention provides a multi-target production scheduling data processing method for a distributed wire cutting machine group, and specifically, the technical scheme of the invention comprises the following steps: Collecting task instruction data, equipment state time sequence data and actual execution time sequence data containing energy consumption records, performing time alignment and data integration, and generating a reference scheduling data set; An ideal constraint model is built based on a reference scheduling data set, and an ideal scheduling time sequence data set is generated, wherein the ideal scheduling time sequence data set comprises task starting time, task ending time, a resource consumption track, and associated beat parameters, waiting parameters and resource occupation parameters; Correcting beat parameters, waiting parameters and resource occupation parameters in an ideal scheduling time sequence data set based on preset recessive loss factors, cooperative conflict factors and combinations of the two, and generating various post-disturbance scheduling states through simulation, wherein the recessive loss factors are used for representing equipment beat attenuation and energy efficiency reduction, and the cooperative conflict factors are used for representing waiting effects caused by material distribution delay and resource contention; extracting a real residual vector based on the actual execution time sequence data and the ideal production scheduling time sequence data set, and extracting a plurality of theoretical residual vectors based on the production scheduling states after a plurality of disturbances and the ideal production scheduling time sequence data set, wherein the real residual vector and the plurality of theoretical residual vectors at least comprise a time delay residual and a resource consumption residual; and determining a production deviation cause based on coupling similarity of the real residual vector and the theoretical residual vectors, and outputting a rescheduling instruction, a maintenance instruction, a holding instruction or a rechecking instruction according to the determination result. Preferably, before integrating the task instruction data, the device state time sequence data and the actual execution time sequence data, preprocessing is further included to obtain preprocessed data, wherein preprocessing includes missing value filling, outlier detection, time stamp alignment and data standardization. Preferably, the process of constructing an ideal constraint model based on a reference scheduling data set and generating an ideal scheduling time sequence data set comprises the steps of setting the health state of equipment to correspond to rated processing beats, setting network issuing delay to be a pre-calibrated reference transmission delay, setting material switching time to be a pre-calibrated standard switching time, and solving integrated data based on mixed integer linear programming to generate the ideal scheduling time sequence data set. Preferably,