Pump body installation scheduling management method based on process optimization

Abstract

The invention provides a pump body installation scheduling management method based on process optimization, and relates to the technical field of data processing, and the method comprises the steps: obtaining basic scheduling data of pump body installation, identifying a preposition relation and a subsequent relation between processes, extracting a state change node of each process, carrying out the statistics of the distribution condition of the state change node in unit time, and carrying out the scheduling management of the pump body installation. Calculating the concentration degree of the process state change in the time dimension to obtain a state aggregation index, depicting a change fluctuation relationship between adjacent processes, calculating the state instability degree of the processes in the execution process to obtain a process fluctuation value, dividing each process into a plurality of fluctuation sections, and calculating the matching degree between the candidate process sequence and the process associated data to obtain a sequence coordination quantity, and replacing the process sequence data in the basic scheduling data to obtain updated scheduling data. According to the method, it can be guaranteed that the pump body can be stably installed in the dynamic execution environment when the procedure goes wrong.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of data processing, in particular to a pump body installation scheduling management method based on process optimization. BACKGROUND

[0002] In the existing pump body installation production management, a scheduling mode based on a manufacturing execution system is generally adopted, and through data management of production tasks, equipment states and personnel scheduling, the whole process tracking of the pump body from component assembly, sealing test to whole machine detection is realized. The system generally takes a fixed process template as the core, generates a scheduling plan according to a preset order, and distributes tasks in combination with equipment load and capacity constraints. Some systems realize process sequencing by introducing a rule engine, such as generating a scheduling result according to process priority or estimated processing time, realizing automatic scheduling and visual monitoring of the assembly line operation.

[0003] However, when a key process is delayed or reworked, such scheduling management systems may not be able to adjust the scheduling order of subsequent processes in real time. For example, in the sealing detection link of the pump body, if the detection result is unqualified and the sealing component needs to be reassembled, the system may not be able to realize real-time scheduling adjustment according to the process state change, which may cause the subsequent workstations to idle due to waiting for materials, and the detection equipment to be idle, resulting in a disconnection between the production plan and the actual execution. SUMMARY

[0004] The purpose of the present application is to provide a pump body installation scheduling management method based on process optimization, aiming to solve the problems mentioned in the background.

[0005] To solve the above technical problems, the technical solution of the present application is as follows: The pump body installation scheduling management method based on process optimization comprises: Obtaining process identification, process sequence data and execution state data of pump body installation to obtain basic scheduling data; According to the basic scheduling data, identifying the precedence relationship and subsequent relationship between each process to obtain process correlation data, and extracting the state change nodes of each process to obtain state expansion data; According to the state expansion data, the distribution of state change nodes in unit time is counted, the concentration degree of process state change in time dimension is calculated, and the state aggregation index is obtained; According to the state aggregation index, the change fluctuation relationship between adjacent processes is described, the state instability degree of the process in the execution process is calculated, and the process fluctuation value is obtained; According to the process fluctuation value, each process is divided into multiple fluctuation sections, and the processes in each fluctuation section are reorganized to obtain a candidate process sequence; According to the candidate process sequence, a matching degree between the candidate process sequence and the process association data is calculated to obtain a sequence coordination quantity; According to the sequence coordination quantity, a process sequence satisfying a preset coordination threshold is selected, and process order data in the basic scheduling data is replaced to obtain updated scheduling data.

[0006] Further, according to the basic scheduling data, a preceding relationship and a subsequent relationship between processes are identified to obtain process association data, and a state change node of each process is extracted to obtain state expansion data, including: According to the basic scheduling data, the process identifier and the process order are divided into a plurality of logical units based on position difference to obtain position grouping data; According to the position grouping data, a last process identifier of each logical unit is taken as a preceding anchor point, and a first process identifier of an adjacent logical unit is taken as a subsequent anchor point to obtain chain mapping data; According to the chain mapping data, a repeatedly appearing anchor point is folded and an associated path is established based on the process identifier to obtain a folded path set; According to the folded path set, the associated path is layered according to path length and woven into a directed sequence structure to obtain the process association data.

[0007] Further, according to the basic scheduling data, a preceding relationship and a subsequent relationship between processes are identified to obtain process association data, and a state change node of each process is extracted to obtain state expansion data, further including: According to the execution state data, the process execution process is cut into a plurality of state intervals along the time axis to obtain a time period division table; According to the time period division table, the state symbols of each state interval are sequentially converted into a symbol sequence, and the adjacent state difference is mapped and identified to obtain an evolution mapping set; According to the evolution mapping set, a time point at which the state conversion frequency exceeds a preset frequency is located to obtain time focus data; According to the time focus data, a mapping relationship between the time point position and the process identifier is established and arranged in time sequence to obtain the state expansion data.

[0008] Further, according to the state expansion data, the distribution of the state change node in unit time is counted, the concentration degree of the process state change in the time dimension is calculated, and the state aggregation index is obtained, including: According to the state expansion data, the node occurrence intensity of the same process in each time section is calculated to obtain a node rate item, the average value of the node rate of the same process in all time sections is calculated to obtain a normalized rate item, and whether the normalized rate is concentrated in a few time sections is calculated to obtain a section dominant item; Based on the normalized rate term, calculate whether the changes in the normalized rate of adjacent time segments are concentrated at a few jump positions to obtain the change compression term; calculate the influence of the prominence of a single time segment on the aggregation determination to obtain the peaking gating term. By integrating and interacting the segment dominance term, change compression term, and peaking gating term, the concentration of process state change nodes in the time dimension is calculated to obtain the state clustering index.

[0009] Furthermore, based on the state clustering index, the relationship between changes and fluctuations between adjacent processes is characterized, the degree of state instability during process execution is calculated, and process fluctuation values ​​are obtained, including: Based on the process sequence data and the state clustering index, the degree of difference in the state clustering index of adjacent processes at the clustering level is calculated to obtain the clustering offset term; the degree of adoptability of adjacent processes in terms of predecessor and successor relationships is calculated to obtain the association correction term. Based on the process sequence data, calculate the degree of synchronous change of state change nodes of adjacent processes within the same time period to obtain the node co-occurrence term; calculate the degree of deviation between the aggregation level and the node synchronization relationship to obtain the coupling expansion term; By fusing the aggregation offset term, association correction term, node co-occurrence term, and coupling expansion term, the degree of state instability of the process during execution is calculated, and the process fluctuation value is obtained.

[0010] Furthermore, based on the process fluctuation values, each process is divided into multiple fluctuation segments, and the processes within each fluctuation segment are reorganized to obtain a candidate process sequence, including: Write the process fluctuation values ​​into the corresponding process identifiers in the process sequence to form a fluctuation entry sequence; Based on the fluctuation entry sequence, the positions where the process fluctuation value changes across the segment identifier are written into the segmentation mark, and the fluctuation entry sequence is split into multiple fluctuation segments using the segmentation mark to obtain the fluctuation segment data. Based on the fluctuation segment data, the process identifiers within each fluctuation segment are generated into an intra-segment order according to the magnitude of the process fluctuation value, and the segment processes are reorganized according to the intra-segment order to obtain the segment reorganization sequence. Based on the segment recombination sequence, the processes of each recombination segment are spliced ​​together according to the order of the fluctuating segments in the fluctuating segment data to obtain the candidate process sequence.

[0011] Furthermore, based on the fluctuation segment data, the process identifiers within each fluctuation segment are generated into an intra-segment order according to the magnitude of the process fluctuation value. The segment processes are then reorganized according to this intra-segment order to obtain a segment reorganization sequence, including: Based on the fluctuation range data, the process identifiers of adjacent positions within the same fluctuation range are paired up, and the corresponding process fluctuation value is attached to each pair of paired entries to obtain a set of paired entries. The process sequence set is obtained by writing the process identifier with a large process fluctuation value in the pairing entry of the pairing entry set into the first position of the segment, and writing the process identifier with a small process fluctuation value into the last position of the segment. According to the segment sequence set, the process identifier at the last position of the segment is connected with the process identifier at the first position of another segment, and when direct connection is not possible, the intra-zone original position is introduced as a connection key to obtain serial registration data. The segment reorganization sequence is obtained by expanding the segment sequence in the serial registration data into an intra-zone order in sequence according to the connection key, and reorganizing the segment process according to the intra-zone order.

[0012] Further, according to the candidate process sequence, the matching degree between the candidate process sequence and the process association data is calculated to obtain a sequence coordination quantity, including: According to the candidate process sequence, the position mapping relationship of each process identifier in the candidate process sequence is calculated to obtain a sequence position item; the hierarchical structure degree of the directed relationship of the preceding relationship and the subsequent relationship in the association link is calculated to obtain an association hierarchical item; According to the sequence position item and the association hierarchical item, the covering strength of each directed relationship pair on the association relationship of the candidate process sequence is calculated to obtain a relationship covering item; the consistency degree between the adjacent splicing of the candidate process sequence and the association link is calculated to obtain an adjacent penetration item; the inhibition degree of the violation degree of the candidate process sequence on the association relationship is calculated to obtain a reverse sequence penalty item; The relationship covering item, the adjacent penetration item and the reverse sequence penalty item are fused to calculate the matching degree between the candidate process sequence and the process association data, and the sequence coordination quantity is obtained.

[0013] Further, according to the sequence coordination quantity, the process sequence satisfying the preset coordination threshold is selected from the sequence coordination quantity, and the process order data in the basic scheduling data is replaced to obtain updated scheduling data, including: According to the candidate process sequence and the sequence coordination quantity, the preset coordination threshold is compared with the sequence coordination quantity corresponding to the candidate process sequence to obtain threshold comparison data; According to the threshold comparison data, the candidate process sequence satisfying the preset coordination threshold is retained, and the remaining candidate process sequence is excluded to obtain a retained sequence set; According to the retained sequence set, the candidate process sequence in the retained sequence set is judged according to the combination order of the sequence coordination quantity and the sequence position item to obtain a target process sequence; According to the target process sequence, the arrangement order of the corresponding process identifier in the process order data is replaced by the arrangement order of the target process sequence to obtain updated scheduling data.

[0014] Further, according to the reserved sequence set, the candidate process sequences in the reserved sequence set are judged according to the combination order of the sequence coordination quantity and the sequence bit position item to obtain a target process sequence, including: According to the reserved sequence set, a sequence number is written for each candidate process sequence, and the corresponding sequence coordination quantity is connected to obtain coordination connection data; According to the coordination connection data, the sequence bit position item corresponding to each candidate process sequence is extracted as a bit position vector, and the bit position vector is bound with the sequence number to obtain bit position binding data; According to the bit position binding data, the sequence coordination quantity and the bit position vector are written in the same combination field to form a judgment key to obtain judgment key data; According to the judgment key data, the candidate process sequences are sorted according to the judgment key, and the first candidate process sequence in the sorting result is selected as the target process sequence.

[0015] The above-mentioned scheme of the present application at least includes the following beneficial effects: The present application explicitly expresses the process dependency relationship originally implied in the process template, artificial experience or static rules as an enumerable and indexable data structure by identifying the precedence relationship and the subsequent relationship between processes, so that the process association data is no longer dependent on runtime rule judgment or artificial inference, but is directly stored and referenced as a kind of structured relationship data, so that any subsequent processing based on order, matching or judgment can directly read the association data for calculation, so that the dependency relationship between processes has the characteristics of repeated reference, provides a direct basis for subsequent matching calculation of candidate sequences and association relationships, and enables the scheduling adjustment process to be completed without reinterpreting the process logic, realizing stable expression of process dependency relationship.

[0016] The present application extracts the state change node of each process while identifying the process relationship, converts the continuous execution state flow process into a discrete time node set, and the execution state data itself exists in the form of continuous state or event flow, which is difficult to directly perform distributed statistics or cross-process comparison. By extracting the state change node, each state conversion is explicitly recorded as a node with time position and process identification, forming a data basis that can be statistically and rearranged by time section, so that subsequent state changes of different processes can be compared and calculated on a unified time dimension, without directly processing complex original state flow, reducing the complexity of time series processing, and providing data input for subsequent calculation.

[0017] The application calculates the state aggregation index by expanding the state data and counting the distribution of state change nodes in a unit time, and compresses the characteristics from multiple node distribution to a single index. The state expansion data contains a large amount of node information scattered in different time sections, and the scheduling judgment will lead to high processing complexity. By counting the node distribution and forming the state aggregation index, the state change characteristics of each process in the time dimension are mapped to a dimensionless value, which can be directly compared between different processes. The processing result does not depend on specific process parameters, and only reflects the time distribution structure of state change, so that the system can quickly identify the differences in time behavior between different processes in subsequent processing, and provide a unified data scale for further characterization of adjacent process relationships.

[0018] The application realizes the conversion from single process characteristics to adjacent process relationship characteristics by describing the change fluctuation relationship between adjacent processes and calculating the process fluctuation value. The state aggregation index only reflects the time distribution characteristics of a single process, while the process fluctuation value extends this characteristic to the difference description in the order dimension by comparing adjacent process indexes, so that the system can clearly determine which processes have large state change differences at the data level, and provide a direct basis for subsequent division according to fluctuation sections. Through this processing, the scheduling system no longer needs to judge whether the order needs to be adjusted based on artificial rules, but can directly divide the sections according to the numerical result of the fluctuation value, so that the order adjustment is based on calculable data.

[0019] The application divides each process into multiple fluctuation sections by the process fluctuation value, and reorganizes the processes in the section to generate a candidate process sequence. The full sequence adjustment is divided into multiple local section operations, which clearly determines which processes belong to the same adjustment unit in the data structure, avoids affecting the order of unrelated processes during adjustment, and saves the candidate process sequence as the result of section reorganization, so that the original order and the reorganized order can exist in parallel at the data level, providing input for subsequent matching calculation and ruling, so that the order adjustment has a clear range of action and result expression, and is convenient for subsequent comparison, screening and replacement. BRIEF DESCRIPTION OF DRAWINGS

[0020] Figure 1 is the flow chart of the pump body installation scheduling management method based on process optimization provided by the embodiment of the application. DETAILED DESCRIPTION

[0021] Exemplary embodiments of the present disclosure will now be described in more detail with reference to the accompanying drawings. While exemplary embodiments of the present disclosure are shown in the drawings, it should be understood that the present disclosure may be implemented in various forms and should not be limited to the embodiments set forth herein. Rather, these embodiments are provided so that this disclosure will be thorough and complete, and will fully convey the scope of the disclosure to those skilled in the art.

[0022] like Figure 1 As shown, embodiments of the present invention propose a pump installation scheduling and management method based on process optimization, the method comprising: Obtain the process identifier, process sequence data, and execution status data for pump body installation to obtain basic scheduling data; Based on the basic scheduling data, the preceding and subsequent relationships between each process are identified to obtain process association data, and the state change nodes of each process are extracted to obtain state expansion data. Based on the state unfolding data, the distribution of state change nodes within a unit of time is statistically analyzed, and the degree of concentration of process state changes in the time dimension is calculated to obtain the state clustering index. Based on the state clustering index, the relationship between changes and fluctuations between adjacent processes is characterized, the degree of state instability of the process during execution is calculated, and the process fluctuation value is obtained. Based on the process fluctuation value, each process is divided into multiple fluctuation segments, and the processes within each fluctuation segment are reorganized to obtain a candidate process sequence. Based on the candidate process sequence, the degree of matching between the candidate process sequence and the process association data is calculated to obtain the sequence coordination quantity; Based on the sequence coordination quantity, select the process sequence that meets the preset coordination threshold, and replace the process sequence data in the basic scheduling data to obtain the updated scheduling data.

[0023] In this embodiment of the invention, the process identifier, process sequence data, and execution status data of the pump body installation are obtained to obtain basic scheduling data, avoiding the problem of repeated mapping of multi-source data and providing a data foundation for subsequent process relationship identification and status analysis. Based on the basic scheduling data, the preceding and subsequent relationships between each process are identified to obtain process association data, and the status change nodes of each process are extracted to obtain status unfolding data. The dependency relationship between processes is transformed from implicit sequence rules into an explicit association data structure, so that process relationships and execution status are expressed in structured data form, providing direct input for subsequent calculations. Based on the status unfolding data, the distribution of status change nodes per unit time is statistically analyzed, and the concentration of process status changes in the time dimension is calculated to obtain the status clustering index, making the status change characteristics of different processes comparable in the time dimension and providing a data benchmark for characterizing the relationship between adjacent processes.

[0024] According to the state aggregation index, the change fluctuation relationship between adjacent processes is described, the state instability degree of the process in the execution process is calculated, the process fluctuation value is obtained, the time characteristics of a single process are extended to the relationship characteristics between adjacent processes, the scheduling system can identify the process position with significant differences in the sequence, and provide data basis for the subsequent; according to the process fluctuation value, each process is divided into multiple fluctuation sections, and the processes in each fluctuation section are reorganized to obtain a candidate process sequence, the adjustment of the process sequence is limited to the fluctuation section to complete, so that the sequence reorganization has a controllable action range, and the balance between local rearrangement and full sequence consistency is realized; according to the candidate process sequence, the matching degree between the candidate process sequence and the process correlation data is calculated to obtain the sequence coordination quantity, so that the rationality of the candidate process sequence is no longer dependent on simple rule checking, and a comparable evaluation basis is provided for the screening of different candidate sequences; according to the sequence coordination quantity, the process sequence whose sequence coordination quantity satisfies the preset coordination threshold is selected, and the process sequence data in the basic scheduling data is replaced to obtain updated scheduling data, so that the scheduling optimization result is directly reflected to the basic scheduling data, and the subsequent scheduling is based on the updated sequence data, so as to ensure the consistency and sustainable updating ability of the scheduling data.

[0025] Among them, the process identification, process sequence data and execution state data of the pump body installation are obtained to obtain the basic scheduling data, specifically including: Firstly, data collection is performed on the procedures involved in the pump body installation process, the procedure identifier being obtained by analyzing the pump body installation process flow, each procedure corresponding to a unique procedure identifier for distinguishing different procedures in subsequent scheduling and state analysis, the procedure identifier being read based on a preconfigured process flow file or a scheduling configuration table, and each procedure identifier being associated with and stored in the corresponding pump body installation task to form a procedure identifier set that can reflect the complete pump body installation process. Further, procedure sequence data corresponding to the procedure identifier is obtained, the procedure sequence data being used to describe the arrangement order of each procedure in the current scheduling period, the position order of each procedure identifier in the scheduling plan being recorded by reading the procedure execution plan generated in the scheduling system, and the corresponding sequence position information being assigned to each procedure identifier. The sequence position information reflects the relative position of the procedure in the pump body installation process, enabling the system to identify the adjacent procedure relationship and the preceding and subsequent relationship based on the sequence information in subsequent processing. Meanwhile, execution state data corresponding to each procedure identifier is obtained, the execution state data being used to describe the state change of the procedure in the actual execution process, the running state record of the procedure being read from the execution monitoring module or the state collection module, the change process of the procedure from non-execution, execution to execution completion being collected, and the corresponding time information being recorded for each state change. The execution state data is organized with the procedure identifier as the index, so that each procedure identifier corresponds to a group of state records arranged in time sequence, reflecting the state evolution trajectory of the procedure in the actual execution process. After the data acquisition is completed, the three types of data are uniformly integrated and processed, the procedure sequence data and the execution state data are written into the data field of the corresponding procedure identifier according to the procedure identifier, and the basic scheduling data is formed.

[0026] In a preferred embodiment of the present application, according to the basic scheduling data, the preceding relationship and the subsequent relationship between the procedures are identified, the procedure association data is obtained, and the state change nodes of each procedure are extracted, and the state expansion data is obtained, including: According to the basic scheduling data, the procedure identifier and the procedure sequence are divided into a plurality of logical units based on the position difference, and the position grouping data is obtained; According to the position grouping data, the last procedure identifier of each logical unit is taken as the preceding anchor point, and the first procedure identifier of the adjacent logical unit is taken as the subsequent anchor point, and the chain mapping data is obtained; According to the chain mapping data, the repeatedly appearing anchor points are folded and the associated paths are established based on the procedure identifier, and the folded path set is obtained; According to the folded path set, the associated paths are layered according to the path length and woven into a directed sequence structure, and the procedure association data is obtained.

[0027] In the embodiment of the present application, according to the basic scheduling data, the process identifiers and the process sequence are divided into several logical units based on the position difference to obtain position grouping data, so that the process sequence has the characteristics of groupability and disassembly on the data structure, and clear data boundaries are provided for subsequent identification of the preceding and subsequent relationship; according to the position grouping data, the last process identifier of each logical unit is taken as the preceding anchor point, and the first process identifier of the adjacent logical unit is taken as the subsequent anchor point to obtain chain mapping data, so that the precedence and subsequent constraints between processes can be expressed in the form of clear data pairs, and the dependence on full sequence scanning is reduced; according to the chain mapping data, the repeatedly appearing anchor points are folded and the associated paths are established by process identifiers to obtain a folding path set, so that the preceding and subsequent relationship of the same process in different logical units is concentratedly expressed, the relationship redundancy is reduced, and the traceability of the associated relationship is enhanced; according to the folding path set, the associated paths are layered according to the path length and woven into a directed sequence structure to obtain process association data, so that the preceding and subsequent relationship between processes forms a stable expression form, and provides a clear relationship basis for subsequent calculation.

[0028] According to the basic scheduling data, the process identifiers and the process sequence are divided into several logical units based on the position difference to obtain position grouping data, which specifically includes: First, the process sequence data is read from the basic scheduling data, and the corresponding sequence position mark is allocated to each process identifier according to the arrangement of the process in the scheduling sequence; then, the displacement difference value between adjacent processes is calculated according to the sequence position marks of the adjacent processes in the process sequence, and the displacement difference value is taken as the basis for judging whether the processes belong to the same logical unit; when the displacement difference value between adjacent processes is within a preset continuous range, the corresponding process identifier is classified into the same logical unit, and when the displacement difference value exceeds the continuous range, a separation of the logical unit is formed at the position; by traversing all the process sequence data, a plurality of logical units that are distinguished from each other are formed in turn, and each logical unit is allocated a unique unit identifier, so as to generate position information records containing the logical unit identifier and the internal process identifier set of the logical unit as the position grouping data.

[0029] According to the position grouping data, the last process identifier of each logical unit is taken as the preceding anchor point, and the first process identifier of the adjacent logical unit is taken as the subsequent anchor point to obtain chain mapping data, which specifically includes: The system determines the process identifier of the last sequential position for each logical unit according to the process sequential position marker inside the logical unit, and records the process identifier as the preceding anchor point corresponding to the logical unit; then searches for the next logical unit adjacent to the logical unit in the process sequence, and determines the process identifier of the first sequential position in the adjacent logical unit, and records it as the subsequent anchor point; on this basis, the preceding anchor point and the subsequent anchor point are associated in pairs, and the identification information of the source logical unit and the target logical unit is attached to the association relationship, forming a chain mapping record; by sequentially performing the above processing on all logical units in the position grouping data, a chain mapping composed of multiple preceding anchor points and subsequent anchor points is obtained, forming chain mapping data describing the process connection relationship between logical units.

[0030] Among them, according to the chain mapping data, the repeated anchor points are folded and the association path is established with the process identifier to obtain a folded path set, specifically including: The system uniformly scans the preceding anchor points and subsequent anchor points involved in the chain mapping data, and identifies the same process identifier appearing multiple times therein; for multiple chain mapping records pointing to the same process identifier, the corresponding preceding and subsequent association relationship is merged and processed, and the repeated mapping pointing is removed, and only the unique process identifier is retained as a path node; on this basis, taking the process identifier as a node, the merged preceding and subsequent mapping relationship is concatenated according to the sequence of the logical units, and an association path composed of multiple process identifiers is gradually constructed; when a process identifier simultaneously serves as a connection node of multiple chain mappings, the corresponding multiple paths are merged at the process identifier to form a path structure containing a branch relationship; by performing the above folding and concatenation operations on all chain mapping data, a set of association paths with process identifiers as nodes and logical unit connection relationships as path directions, i.e. the folded path set, is finally obtained.

[0031] In a preferred embodiment of the present application, according to the basic scheduling data, the preceding relationship and the subsequent relationship between the processes are identified to obtain process association data, and the state change nodes of each process are extracted to obtain state expansion data, further comprising: According to the execution state data, the process execution process is cut into multiple state intervals along the time axis to obtain a time interval division table; According to the time interval division table, the state markers of each state interval are sequentially converted into a symbol sequence, and the mapping identification of the adjacent state difference is performed to obtain an evolution mapping set; According to the evolution mapping set, the time points with state conversion frequency exceeding a preset frequency are located to obtain time focus data; According to the time focus data, the mapping relationship between the time point position and the process identifier is established and arranged in time sequence to obtain the state expansion data.

[0032] In the embodiments of the present application, according to the execution state data, the process execution process is cut into multiple state intervals along the time axis to obtain a time interval division table, thereby providing accurate time positioning basis for subsequent, so that the state changes between different processes can be compared on a unified time scale; according to the time interval division table, the state symbols of each state interval are sequentially converted into a symbol sequence, and the mapping identification of the adjacent state difference is performed to obtain an evolution mapping set, thereby compressing the data size and enabling the evolution logic between states to be uniformly analyzed among multiple processes; according to the evolution mapping set, the time points at which the state conversion frequency exceeds the preset frequency are located to obtain time focus data, thereby accurately reflecting the concentrated moments of state changes of the process in the execution process and providing a sparse and information-complete data structure for subsequent; according to the time focus data, the mapping relationship between the time point position and the process identification is established and arranged in time sequence to obtain state expansion data, thereby realizing the alignment of the time dimension across processes and enabling the overall distribution and cross-influence of the process state changes to be analyzed from a global perspective.

[0033] According to the execution state data, the process execution process is cut into multiple state intervals along the time axis to obtain a time interval division table, and the time interval division table comprises: The system first reads the execution state data corresponding to each process from the basic scheduling data, and the execution state data continuously records the state changes of the process in the execution process in time sequence, and each record at least contains the process identification, the state symbol and the corresponding time mark. The system sorts the execution state data of the same process according to time, and takes whether the state symbol changes as the cutting basis to identify the position of the inconsistent state symbol in the adjacent state records on the time axis. For the record segment with continuous same state symbol, the system merges it into the same state interval, and takes the start time of the interval as the interval start mark and the time of the last occurrence of the state symbol as the interval end mark, thereby forming a complete state interval description. For each state interval, the system records the corresponding process identification, state symbol, interval start time and interval end time, and writes multiple state intervals of the same process in the unified data structure in time sequence. Through repeating the above processing process for all processes, the time interval division table containing all state interval information of each process in the execution process is obtained.

[0034] According to the time interval division table, the state symbols of each state interval are sequentially converted into a symbol sequence, and the mapping identification of the adjacent state difference is performed to obtain an evolution mapping set, and the evolution mapping set comprises: The corresponding state interval record of each process is read respectively, and is arranged according to the order of the starting time of the interval. The state symbols appearing in the time period division table are assigned corresponding symbol identifiers, so that each state symbol can be uniquely represented by a symbol. On this basis, the system transcribes the state interval sequence of the same process into a symbol sequence, and compresses the state intervals that appear continuously and have the same symbol during the transcription process, and only retains the symbol representation once, to avoid repeated states interfering with subsequent analysis. Then, the system compares the adjacent compressed symbol sequences, identifies the positions where the symbol changes one by one, and forms a group of state evolution relationships with the previous symbol and the next symbol. For each group of state evolution relationships, the system synchronously records the corresponding process identifier and the position of the state change on the time axis, which can be the end time of the previous state interval or the starting time of the next state interval. The state evolution relationships and their corresponding time information identified in all processes are collected to form an evolution mapping set.

[0035] Among them, according to the evolution mapping set, the time points with state transition frequency exceeding the preset frequency are located to obtain time focus data, specifically including: The system analyzes the time distribution of the state transition records in the evolution mapping set based on the time axis. The system first divides the continuous time axis into multiple adjacent time windows according to a preset time span, and assigns each state transition record in the evolution mapping set to the corresponding time window according to its corresponding time position. Then, the system counts the number of state transition records falling into each time window. When the number of state transitions in a time window exceeds the preset frequency threshold, the system determines that the time window is a state transition intensive area. For the time window determined to be a state transition intensive area, the system extracts the center time point of the window or the position where the number of state transitions reaches a peak as a representative time point, and records the association between the time point and the corresponding process identifier. To avoid excessive concentration of time points caused by repeated marking in adjacent time windows, the system also performs merging processing on the representative time points that are adjacent in time and have a distance less than a preset time distance, and only retains one time point as the time identifier of the intensive area. Through the above processing, the time focus data composed of multiple discrete time points is finally formed.

[0036] In a preferred embodiment of the present application, according to the state expansion data, the distribution of state change nodes in unit time is counted, the concentration degree of process state change in time dimension is calculated, and a state aggregation index is obtained, including: According to the state expansion data, the node occurrence intensity of the same process in each time section is calculated to obtain a node rate item; the average value of the node rate of the same process in all time sections is calculated to obtain a normalized rate item; whether the normalized rate is concentrated in a few time sections is calculated to obtain a section dominant item; According to the normalized rate term, whether the change of the normalized rate of adjacent time sections is concentrated in a few jump positions is calculated to obtain a change compression term; the influence degree of the prominence of a single time section on the aggregation determination is calculated to obtain a peaking gating term; The section dominant term, the change compression term and the peaking gating term are fused and interacted to calculate the concentration degree of the process state change node in the time dimension to obtain a state aggregation index.

[0037] In the embodiment of the present application, according to the state expansion data, the node appearance intensity of the same process in each time section is calculated to obtain a node rate term, which converts the discrete state change in the time dimension into a data term that can be continuously compared, providing basic data conditions for subsequent analysis of the distribution difference between different time sections; the average value of the node rate of the same process in all time sections is calculated to obtain a normalized rate term, avoiding the imbalance of subsequent distribution judgment caused by the difference in the state change frequency of the process itself; whether the normalized rate is concentrated in a few time sections is calculated to obtain a section dominant term, providing a clear distribution feature input for depicting the time distribution form; according to the normalized rate term, whether the change of the normalized rate of adjacent time sections is concentrated in a few jump positions is calculated to obtain a change compression term, which extends the dynamic characteristics of the time distribution from the static concentration degree to the change path level, so that the subsequent fusion can consider both the concentration occurrence and the concentration change of the time behavior characteristics; the influence degree of the prominence of a single time section on the aggregation determination is calculated to obtain a peaking gating term, avoiding the uncontrollable influence of the abnormal value of a single section on the overall concentration judgment; the section dominant term, the change compression term and the peaking gating term are fused and interacted to calculate the concentration degree of the process state change node in the time dimension to obtain a state aggregation index, providing stable, comparable and repeatedly referenced time behavior feature input for subsequent calculation, playing a data hub role in the whole scheduling management method.

[0038] In a preferred embodiment of the present application, according to the state aggregation index, the change fluctuation relationship between adjacent processes is depicted, the state instability degree of the process in the execution process is calculated to obtain a process fluctuation value, including: According to the process order data and the state aggregation index, the order difference degree of the state aggregation indexes of adjacent processes in the aggregation level is calculated to obtain an aggregation offset term; the adoptable degree of adjacent process pairs in the preceding relationship and the subsequent relationship is calculated to obtain a correlation correction term; According to the process order data, the synchronization change degree of the state change nodes of adjacent processes in the same time section is calculated to obtain a node co-occurrence term; the deviation degree between the aggregation level and the node synchronization relationship is calculated to obtain a coupling expansion term; The aggregation offset term, the correlation correction term, the node co-occurrence term and the coupling expansion term are fused to calculate the state instability degree of the process in the execution process to obtain a process fluctuation value.

[0039] In the embodiments of the present application, according to the process sequence data and the state aggregation index, the sequence difference degree of the state aggregation indexes of adjacent processes at the aggregation level is calculated to obtain an aggregation offset term, which reflects the data expression of the difference in the state aggregation degree of adjacent processes in the process sequence structure, and provides the basic sequence difference information for subsequent characterization of the stability between processes; the adoptable degree of adjacent process pairs in the preceding relationship and the subsequent relationship is calculated to obtain a correlation correction term, which avoids making a judgment on the process sequence only according to the numerical change, and ensures the consistency between the fluctuation characterization and the process logic; according to the process sequence data, the synchronization change degree of the state change nodes of adjacent processes in the same time section is calculated to obtain a node co-occurrence term, which reflects the time coupling degree of adjacent processes in the execution process, and characterizes a more complete fluctuation basis; the deviation degree between the aggregation level and the node synchronization relationship is calculated to obtain a coupling expansion term, which identifies the data situation that the state of a process is abnormally aggregated but not synchronized with adjacent processes, and provides the data interaction result for subsequent comprehensive judgment of the execution stability of the process; the aggregation offset term, the correlation correction term, the node co-occurrence term and the coupling expansion term are fused to calculate the state instability degree of the process in the execution process to obtain a process fluctuation value, so that the subsequent section division and sequence reorganization can be directly processed based on the value, and the conversion from multi-dimensional features to single-value decision basis is completed.

[0040] In a preferred embodiment of the present application, according to the process fluctuation value, each process is divided into a plurality of fluctuation sections, and the processes in each fluctuation section are reorganized to obtain a candidate process sequence, which includes: writing the process fluctuation value into the corresponding process identifier according to the process sequence to form a fluctuation item sequence; According to the fluctuation item sequence, the position where the process fluctuation value changes across the section identifier is written into a split mark, and the fluctuation item sequence is split into a plurality of fluctuation sections by the split mark to obtain fluctuation section data; According to the fluctuation section data, the process identifiers in each fluctuation section are generated in the order of the size of the process fluctuation value to generate an intra-zone sequence, and the section processes are reorganized according to the intra-zone sequence to obtain a section reorganization sequence; According to the section reorganization sequence, the reorganized section processes are spliced according to the fluctuation section data to obtain a candidate process sequence.

[0041] In the embodiment of the present application, the process fluctuation value is written in the corresponding process identification in the process order to form a fluctuation item sequence, which provides a direct basis for identifying the fluctuation change in the subsequent order position, and the fluctuation characteristics can be observed without destroying the original order semantics; according to the fluctuation item sequence, the position where the process fluctuation value changes across the segmented identification is written in the split marker, and the fluctuation item sequence is split into multiple fluctuation segments by the split marker to obtain the fluctuation segment data, avoiding the overall operation of the whole process sequence, and clearly defining the action boundary of the order adjustment; according to the fluctuation segment data, the process identification in each fluctuation segment is generated in the order of the size of the process fluctuation value, and the segment process is reorganized according to the intra-zone order to obtain a segment reorganization sequence, avoiding irrelevant processes from being included in the adjustment range, and ensuring that the rearrangement of the processes in the segment has a clear numerical basis; according to the segment reorganization sequence, the reorganized segment processes are spliced according to the fluctuation segment data to obtain a candidate process sequence, providing clear, complete and verifiable data input for subsequent use.

[0042] In the embodiment of the present application, the process fluctuation value is written in the corresponding process identification in the process order to form a fluctuation item sequence, which provides a direct basis for identifying the fluctuation change in the subsequent order position, and the fluctuation characteristics can be observed without destroying the original order semantics; according to the fluctuation item sequence, the position where the process fluctuation value changes across the segmented identification is written in the split marker, and the fluctuation item sequence is split into multiple fluctuation segments by the split marker to obtain the fluctuation segment data, avoiding the overall operation of the whole process sequence, and clearly defining the action boundary of the order adjustment; according to the fluctuation segment data, the process identification in each fluctuation segment is generated in the order of the size of the process fluctuation value, and the segment process is reorganized according to the intra-zone order to obtain a segment reorganization sequence, avoiding irrelevant processes from being included in the adjustment range, and ensuring that the rearrangement of the processes in the segment has a clear numerical basis; according to the segment reorganization sequence, the reorganized segment processes are spliced according to the fluctuation segment data to obtain a candidate process sequence, providing clear, complete and verifiable data input for subsequent use. The system reads the fluctuation item sequence in the order determined by the process order data, and writes the order position field for each fluctuation item in the reading process, so that each fluctuation item contains at least the process identification, the process fluctuation value and the order position; on this basis, the system performs adjacent traversal on the fluctuation item sequence, and for any adjacent first fluctuation item and second fluctuation item, the corresponding process fluctuation values are extracted from the two, and then the process fluctuation values are mapped to the segmented identification, wherein the segmented identification is used to represent the fluctuation level interval to which the process fluctuation value belongs, and the fluctuation level interval is obtained by dividing the value range of the process fluctuation value according to a preset rule; when the segmented identification of the first fluctuation item is inconsistent with the segmented identification of the second fluctuation item, the system writes the order position of the second fluctuation item in the split marker field, and registers the order position as the segment start position, and at the same time, the position of the last registered split marker is registered as the segment end position, to form a segment boundary record; when the segmented identification of the first fluctuation item is consistent with the segmented identification of the second fluctuation item, the system does not write the split marker in the current traversal, and the second fluctuation item is continued to be included in the item set of the current segment; after the traversal is completed, the system performs split writing operation on the fluctuation item sequence according to the segment boundary record, that is, the continuous fluctuation items between the adjacent two split markers are copied and written in the same fluctuation segment record, and the segment number, the segment start position, the segment end position and the segment containing process identification list are written in the fluctuation segment record, to obtain the fluctuation segment data.

[0043] Specifically, based on the segment recombination sequence, the processes of each recombination segment are spliced ​​together according to the order of the fluctuation segments in the fluctuation segment data to obtain the candidate process sequence, which includes: The system reads the segment number and its corresponding start and end positions from the fluctuating segment data, and generates a segment sequence accordingly. This sequence represents the order of each fluctuating segment in the original process sequence data. The system further retrieves the segment recombination sequence for each segment using the segment number as the key. This recombination sequence is the result of rearranging the process identifiers within the segment according to their internal order, and includes at least the segment number and a list of process identifiers within the segment. During the splicing process, the system processes each segment according to its sequence. For the currently processed segment recombination sequence, boundary writing is performed first, writing the start and end positions of the segment into the boundary fields of the recombined sequence to ensure traceability of the segment's origin after splicing. Subsequently, the system... The list of process identifiers within the segment of the previous segment recombination sequence is appended to the candidate sequence cache. During the appending process, a global position index is regenerated for each process identifier to form a continuous candidate sequence representation. After the appending is completed, the system continues to append the segment recombination sequence of the next fluctuation segment to the candidate sequence cache in the same way until all segments in the previous and next segments have been appended. After all appending is completed, the system performs consistency maintenance on the candidate sequence cache, establishes a correspondence between the global position index of each process identifier in the candidate sequence cache and its source segment number, and writes this correspondence into the mapping field of the candidate process sequence, so that the candidate process sequence simultaneously contains a three-element record of process identifier, candidate position, and source segment. Through the above splicing and maintenance, the candidate process sequence is output.

[0044] In a preferred embodiment of the present invention, based on the fluctuation segment data, the process identifiers within each fluctuation segment are generated into an intra-segment order according to the magnitude of the process fluctuation value, and the segment processes are reorganized according to the intra-segment order to obtain a segment reorganization sequence, including: Based on the fluctuation range data, the process identifiers of adjacent positions within the same fluctuation range are paired up, and the corresponding process fluctuation value is attached to each pair of paired entries to obtain a set of paired entries. By writing the process identifier with the larger process fluctuation value in the paired entries of the paired entry set to the beginning of the segment and writing the process identifier with the smaller process fluctuation value to the end of the segment, a segment sequence set is obtained; Based on the fragment sequence set, the process identifier at the end of one fragment is connected to the process identifier at the beginning of another fragment. When direct connection is not possible, the original position within the area is introduced as the connection key to obtain serialized registration data. By expanding the sequence of fragments in the serially registered data into an intra-regional order according to the linking key, and then reorganizing the segment processes according to the intra-regional order, a segment reorganization sequence is obtained.

[0045] In the embodiments of the present application, according to the fluctuation section data, the process identifiers of adjacent positions in the same fluctuation section are paired two by two, and the corresponding process fluctuation values are hung for each pair of paired entries to obtain a paired entry set, which provides a direct comparison object for subsequent intra-zone reorganization, so that the process adjustment can be carried out around the fluctuation relationship; by writing the process identifier with a large process fluctuation value in the paired entry of the paired entry set into the first position of the segment and writing the process identifier with a small process fluctuation value into the last position of the segment, a segment sequence set is obtained, which avoids one-time sorting of the entire section and gradually constructs the intra-zone sequence through local segments, thereby providing a structured basis for subsequent segment-level splicing; according to the segment sequence set, the process identifier at the last position of a segment is connected with the process identifier at the first position of another segment, and when direct connection is not possible, the original position in the zone is introduced as a connection key to obtain serial registration data, thereby realizing ordered connection between multiple segments, avoiding unordered splicing or conflict splicing between segments, ensuring that the intra-zone sequence generation process has consistent data basis; by sequentially expanding the segment sequence in the serial registration data according to the connection key into the intra-zone sequence, and reorganizing the section process according to the intra-zone sequence, a section reorganization sequence is obtained, which provides stable input for subsequent matching calculation with process-related data.

[0046] In the embodiments of the present application, according to the fluctuation section data, the process identifiers of adjacent positions in the same fluctuation section are paired two by two, and the corresponding process fluctuation values are hung for each pair of paired entries to obtain a paired entry set, which provides a direct comparison object for subsequent intra-zone reorganization, so that the process adjustment can be carried out around the fluctuation relationship; by writing the process identifier with a large process fluctuation value in the paired entry of the paired entry set into the first position of the segment and writing the process identifier with a small process fluctuation value into the last position of the segment, a segment sequence set is obtained, which avoids one-time sorting of the entire section and gradually constructs the intra-zone sequence through local segments, thereby providing a structured basis for subsequent segment-level splicing; according to the segment sequence set, the process identifier at the last position of a segment is connected with the process identifier at the first position of another segment, and when direct connection is not possible, the original position in the zone is introduced as a connection key to obtain serial registration data, thereby realizing ordered connection between multiple segments, avoiding unordered splicing or conflict splicing between segments, ensuring that the intra-zone sequence generation process has consistent data basis; by sequentially expanding the segment sequence in the serial registration data according to the connection key into the intra-zone sequence, and reorganizing the section process according to the intra-zone sequence, a section reorganization sequence is obtained, which provides stable input for subsequent matching calculation with process-related data. First, the segment first process identifier and the segment last process identifier of each segment in the segment sequence set are read, and the in-zone original positions of the process identifiers in the same fluctuation section are synchronously read; then, the segment last process identifier is taken as the current splice end, the segment first process identifiers of the remaining unprocessed segments are traversed, the process order data recorded in the fluctuation section data is searched, and it is judged whether the in-zone original position corresponding to the current splice end and the in-zone original position of the first process identifier of the to-be-spliced segment satisfy the preset direct splicing condition, the direct splicing condition at least including: the in-zone original positions satisfy the adjacent relationship, or there is no boundary position marked as a split mark between the two; when the direct splicing condition is satisfied, the system writes the current segment number and the to-be-spliced segment number into the splicing registration field, and writes the last process identifier of the current segment and the first process identifier of the to-be-spliced segment into the splicing pair field, and writes the in-zone original position difference corresponding to the splicing pair into the difference field, to form a traceable splicing record; when the direct splicing condition is not satisfied, the system extracts the in-zone original position from the fluctuation section data as a splicing key, the splicing key being composed of the in-zone original position of the last process identifier of the current segment and the in-zone original position of the first process identifier of the to-be-spliced segment, and writes the splicing key into the key field, and writes the key field, the current segment number, and the to-be-spliced segment number into the registration entry, to record that the splicing relationship needs to rely on the in-zone original position to complete the sorting; after completing a splicing registration, the system marks the to-be-spliced segment as an associated segment, and updates the last process identifier of the to-be-spliced segment as a new current splice end, and continues to repeat the above searching, judging and registering process for the remaining unassociated segments, until all the segments in the fluctuation section are registered as splicable links, to form serial registration data including segment numbers, splicing pairs, in-zone original position differences, and splicing keys.

[0047] In the serial registration data, the segment sequence is unfolded into an in-zone order according to the splicing key, and the section processes are reorganized according to the in-zone order, to obtain a section reorganization sequence, specifically including: Firstly, the registration entries in the serial registration data are read, the registration entries are divided into direct connection entries and key connection entries according to whether the key field is contained, and the connection keys in the key connection entries are sorted, the sorting taking the order of the original positions in the area contained by the connection keys as the basis for sorting, so that the fragment external order table is obtained; subsequently, the system writes the fragment number pairs corresponding to the direct connection entries into adjacent positions on the basis of the fragment external order table, so as to ensure that the direct connection relationship is kept continuous in the fragment external order table; after the fragment external order table is obtained, the system reads the process identification sequences of the corresponding fragments in the order of the fragment numbers in the table one by one, and appends the process identification sequences of the fragments to the same unfolding sequence in turn, so as to form the intra-area unfolding sequence; in the appending process, the system performs the de-duplication check on the boundary process identifications between adjacent fragments, and when it is found that the last process identification of the previous fragment is the same as the first process identification of the next fragment or is recorded as the same connection pair in the serial registration data, only one of the process identifications is kept to avoid repeated writing, so that the intra-area order without repetition is obtained; subsequently, the system takes the intra-area order as the reorganization index, and rearranges the section process entries corresponding to the section in the fluctuation section data, that is, the section process entries are extracted and written in order according to the appearance order of the process identifications in the intra-area order, so as to form the section reorganization sequence.

[0048] In a preferred embodiment of the present application, according to the candidate process sequence, the matching degree between the candidate process sequence and the process association data is calculated to obtain the sequence coordination quantity, including: According to the candidate process sequence, the position mapping relationship of each process identification in the candidate process sequence is calculated to obtain the sequence bit item; the hierarchical structure degree of the directed relationship of the preceding relationship and the subsequent relationship in the association link is calculated to obtain the association hierarchical item; According to the sequence bit item and the association hierarchical item, the covering strength of each directed relationship pair on the association relationship of the candidate process sequence pair is calculated to obtain the relationship covering item; the consistency degree between the adjacent splicing of the candidate process sequence and the association link is calculated to obtain the adjacent penetration item; the inhibition degree of the violation degree of the association relationship by the candidate process sequence is calculated to obtain the reverse sequence penalty item; The relationship covering item, the adjacent penetration item and the reverse sequence penalty item are fused to calculate the matching degree between the candidate process sequence and the process association data, and the sequence coordination quantity is obtained.

[0049] In the embodiments of the present application, according to the candidate process sequence, the position mapping relationship of each process identifier in the candidate process sequence is calculated to obtain a sequence bit item, thereby providing a unified bit reference basis for subsequent processes and avoiding repeated traversal of the sequence or relying on implicit order inference in the matching judgment process; the hierarchical structure degree of the directed relationship of the preceding relationship and the subsequent relationship in the associated link is calculated to obtain an associated hierarchical item, thereby being able to distinguish the difference between direct association and indirect association in sequence matching and providing a more fine-grained data basis for subsequent processes and avoiding equating different complexity of the associated relationship in the matching judgment; according to the sequence bit item and the associated hierarchical item, the coverage strength of each directed relationship pair on the associated relationship of the candidate process sequence is calculated to obtain a relationship coverage item, thereby avoiding the problem that the sequence as a whole is considered reasonable only by a single relationship; the consistency degree between the adjacent splicing of the candidate process sequence and the associated link is calculated to obtain an adjacent penetration item, thereby depicting the matching degree between the candidate process sequence and the process association data and identifying whether the adjacent process splicing conforms to the natural connection logic of the process association; the inhibition degree of the violation degree of the associated relationship of the candidate process sequence is calculated to obtain an inverse sequence penalty item, thereby avoiding ignoring the potential conflict risk brought by the inverse sequence only by the coverage or penetration relationship; the relationship coverage item, the adjacent penetration item and the inverse sequence penalty item are fused to calculate the matching degree between the candidate process sequence and the process association data to obtain a sequence coordination quantity, thereby enabling different candidate process sequences to have the characteristics of orderable and filterable under the same scale and providing direct and stable data basis for subsequent processes.

[0050] In a preferred embodiment of the present application, according to the sequence coordination quantity, the process sequence satisfying the preset coordination threshold value is selected, and the process order data in the basic scheduling data is replaced to obtain updated scheduling data, including: According to the candidate process sequence and the sequence coordination quantity, the preset coordination threshold value is compared with the sequence coordination quantity corresponding to the candidate process sequence to obtain threshold comparison data; According to the threshold comparison data, the candidate process sequence satisfying the preset coordination threshold value is retained, and the remaining candidate process sequence is excluded to obtain a retained sequence set; According to the retained sequence set, the candidate process sequence in the retained sequence set is judged according to the combined order of the sequence coordination quantity and the sequence bit item to obtain a target process sequence; According to the target process sequence, the arrangement order of the corresponding process identifier in the process order data is replaced by the arrangement order of the target process sequence to obtain updated scheduling data.

[0051] In the embodiment of the present application, according to the candidate process sequence and the sequence coordination amount, the preset coordination threshold is compared with the sequence coordination amount corresponding to the candidate process sequence to obtain threshold comparison data, avoiding repeated reference to the coordination threshold for judgment in subsequent steps, and providing clear and unified data basis for subsequent screening operation; according to the threshold comparison data, the candidate process sequence whose sequence coordination amount meets the preset coordination threshold is retained, and the remaining candidate process sequence is excluded to obtain a reserved sequence set, realizing the first convergence of the candidate process sequence, and excluding the sequence that does not meet the basic correlation constraint requirement from the subsequent decision-making process, avoiding invalid sequence participating in the ruling; according to the reserved sequence set, the candidate process sequence in the reserved sequence set is ruled according to the combination order of the sequence coordination amount and the sequence position item to obtain the target process sequence, avoiding the case of selecting only a single index, and ensuring that the selected target process sequence has a clear priority basis at the data level; according to the target process sequence, the arrangement order of the corresponding process identifier in the process order data is replaced by the arrangement order of the target process sequence to obtain updated scheduling data, avoiding the problem that the scheduling optimization result only stays at the analysis level and is not actually used, and forming a complete and closed scheduling update link from the data processing point of view.

[0052] Among them, according to the target process sequence, the arrangement order of the corresponding process identifier in the process order data is replaced by the arrangement order of the target process sequence to obtain updated scheduling data, specifically including: When the target process sequence is determined, the system first reads the currently effective process order data from the basic scheduling data, and parses the process order data into a process identifier list arranged in order, and at the same time, assigns each process identifier in the list a corresponding order index position to form an original order mapping relationship. The order mapping relationship is used to represent the arrangement state of each process in the updated scheduling plan before the update, and provides a basis for comparison for subsequent replacement operations. Then the system reads the process identifier arrangement information in the target process sequence, and performs integrity check on the target process sequence to confirm that the process identifiers contained in the target process sequence are consistent with the set of process identifiers involved in the process order data, and there is no missing or duplicate identifier. After the verification passes, the system takes the target process sequence as a new order reference, and generates a new order index relationship according to the arrangement order of the process identifier in the target process sequence, which is used to represent the updated process arrangement state.

[0053] In the sequential replacement process, the system does not directly overwrite the original process sequence data, but first establishes a correspondence table between the original sequence mapping relationship and the new sequence index relationship, and records the sequence positions of the same process identifier before and after the update. Through the correspondence table, the system can locate the position of each process identifier in the original process sequence data, and replace it with the corresponding sequence index value in the target process sequence, thereby completing the overall update of the process arrangement sequence. After completing the sequence index replacement, the system rewrites the updated sequence index relationship into the process sequence data structure, and performs consistency check on the updated process sequence data to confirm that the number of process identifiers, the sequence continuity and the association with other scheduling fields have not been conflicted. The consistency check is used to ensure that the replacement operation only changes the arrangement sequence of the process identifier, without affecting the process identifier itself and its associated execution state data, process association data and other scheduling information. Finally, the process sequence data that passes the consistency check is written back to the basic scheduling data to form the updated scheduling data, and the process sequence mapping relationship before and after the update is stored as an update record. The unified data source for subsequent scheduling execution, state acquisition and again scheduling calculation is formed, thereby completing the sequence replacement and landing from the target process sequence to the actual scheduling plan at the data processing level.

[0054] In a preferred embodiment of the present application, according to the reserved sequence set, the candidate process sequences in the reserved sequence set are judged according to the combined order of the sequence coordination amount and the sequence bit position item to obtain the target process sequence, comprising: According to the reserved sequence set, write the sequence number for each candidate process sequence and hang the corresponding sequence coordination amount to obtain coordination hanging data; According to the coordination hanging data, extract the sequence bit position item corresponding to each candidate process sequence into a bit position vector, and bind the bit position vector with the sequence number to obtain bit position binding data; According to the bit position binding data, write the sequence coordination amount and the bit position vector into the same combination field and form a judgment key to obtain judgment key data; According to the judgment key data, sort the candidate process sequences according to the judgment key and select the first candidate process sequence in the sorting result as the target process sequence.

[0055] In the embodiment of the present application, according to the reserved sequence set, the sequence number is written for each candidate process sequence, and the corresponding sequence coordination quantity is hung, to obtain coordination hanging data, to realize the unique distinction and numerical reference of different candidate sequences, to enable the corresponding coordination quantity to be quickly retrieved in a key-value access manner, to ensure data consistency and access efficiency; according to the coordination hanging data, the sequence position item corresponding to each candidate process sequence is extracted as a position vector, and the position vector is bound with the sequence number to obtain position binding data, to realize a two-dimensional combination sorting logic, to convert the process order information from a linear list to an index vector that can be queried, to provide a data basis for complex sorting operations; according to the position binding data, the sequence coordination quantity and the position vector are written into the same combination field and form a ruling key to obtain ruling key data, to ensure that candidate sequences of different lengths and structures can still be compared under a unified dimension, to realize the unified expression of multi-attribute data; according to the ruling key data, the candidate process sequences are sorted according to the ruling key, and the first candidate process sequence in the sorting result is selected as the target process sequence, to realize the complete data flow from multi-scheme generation to single-scheme landing, to avoid subjective differences caused by manual judgment and experience selection.

[0056] According to the coordination hanging data, the sequence position item corresponding to each candidate process sequence is extracted as a position vector, and the position vector is bound with the sequence number to obtain position binding data, and specifically includes: First, the candidate process sequence identifier corresponding to each sequence number in the coordination hanging data is read, and a position mapping relationship is generated in the order of appearance from the process identifier list of the candidate process sequence, each process identifier is associated and registered with its position sequence number in the candidate process sequence, to form a position mapping table; on the basis of the position mapping table, the position sequence number is sequentially written into the vector structure field in the reading order of the process identifier, to obtain the position vector, wherein the position vector is used to express the position in the candidate process sequence in a unified data structure; then, the sequence number is taken as the primary key, the position vector field is written into the record row corresponding to the sequence number, and the record row is also hung with the sequence coordination quantity field of the sequence number in the coordination hanging data, to form the position binding data, so that the same sequence number has both the internal structure expression of the candidate process sequence and the external matching representation, to avoid data inconsistency caused by repeatedly generating position mapping in the sorting stage.

[0057] According to the position binding data, the sequence coordination quantity and the position vector are written into the same combination field and form a ruling key to obtain ruling key data, and specifically includes: First, the sequence coordination quantity and position vector corresponding to each sequence number in the position binding data are read. The position vector is lengthened, and its length field is compared with the candidate process sequence length field. If the lengths are inconsistent, the position vector is padded with placeholders or truncated to a consistent length to obtain an aligned position vector, ensuring that the field structure of different candidate process sequences is consistent in the combination field. Based on the aligned position vector, in-order encoding is performed on the position vector. The position number of each position in the position vector is written into the encoding sequence field according to a preset encoding rule. The encoding rule includes writing each position number in a fixed-width format or writing the difference between adjacent position numbers to form differential encoding, so that the encoding sequence field has a comparable lexicographical order feature. Subsequently, the sequence coordination value is written into the main segment of the combined field, and the encoded sequence field is written into the secondary segment of the combined field, so that the combined field forms a key-value structure with the main segment taking precedence and the secondary segment making the decision. Furthermore, in order to avoid the decision being inconsistent due to the occurrence of parallel main segments, a check segment of the sequence number is written into the combined field, so that a deterministic decision can still be achieved through the check segment under the same main segment and secondary segment conditions. Finally, the combined field is recorded as the decision key, and the decision key data is written with the correspondence between the sequence number and the decision key, so that each candidate process sequence corresponds to a sortable, comparable, and reproducible decision key. The decision key binds the matching degree information and internal positional structure information of the candidate process sequence in the same field, providing a single-field entry for subsequent unified sorting.

[0058] Specifically, based on the decision key data, the candidate process sequences are sorted according to the decision key, and the first candidate process sequence in the sorting result is selected as the target process sequence. This includes: First, read the adjudication key data corresponding to each sequence number of the adjudication key, parse the adjudication key into the main section, the slave section and the check section, and sort all candidate process sequences in the first round according to the sorting rule of the main section, so that the candidate process sequence whose sequence coordination quantity corresponding to the main section meets the preset sorting direction is arranged in front; When multiple candidate process sequences with the same main section appear in the first round of sorting, the system performs the second round of sorting on the parallel set, compares the slave section of the adjudication key in dictionary order or compares the order of the difference sequence, so that the candidate process sequence whose bit position structure is more consistent with the preset combination rule is arranged in front, thereby introducing the internal structure difference of the candidate process sequence into the adjudication process and forming parallel elimination; If there is still parallel after the second round of sorting, the third round of sorting is performed on the parallel set, the sequence number corresponding to the check section is sequentially adjudicated, so that the output result has certainty; After completing multiple rounds of sorting, the system looks up the sequence number corresponding to the first sequence number of the sorting result to the bit position binding data to obtain the candidate process sequence identification list corresponding to the sequence number, and writes the candidate process sequence identification list into the target sequence field to form the target process sequence. At the same time, generate the adjudication record field to write the sequence number corresponding to the target process sequence, the adjudication key and the adjudication round information, so that the source and the adjudication basis of the target process sequence can be traced back, and ensure that when the basic scheduling data is replaced in sequence in the future, the target process sequence on which the replacement is based has clear data source and consistent adjudication path.

[0059] The above is the preferred embodiment of the present application. It should be noted that for ordinary skilled persons in the technical field, without departing from the principles of the present application, several improvements and refinements can be made, which should also be considered as the protection scope of the present application.

Claims

1. A pump body installation scheduling management method based on process optimization, characterized by, The method comprises: Obtaining process identification, process sequence data and execution state data of the pump body installation, to obtain basic scheduling data; According to the basic scheduling data, the preposition relationship and the subsequent relationship between the processes are identified, the process correlation data is obtained, and the state change nodes of each process are extracted, to obtain state expansion data; According to the state expansion data, the distribution of the state change nodes in a unit of time is counted, the concentration degree of the process state change in the time dimension is calculated, and the state aggregation index is obtained; According to the state aggregation index, the change fluctuation relationship between adjacent processes is described, the state instability degree of the process in the execution process is calculated, and the process fluctuation value is obtained; According to the process fluctuation value, each process is divided into multiple fluctuation sections, and the processes in each fluctuation section are reorganized, to obtain a candidate process sequence; According to the candidate process sequence, the matching degree between the candidate process sequence and the process correlation data is calculated, and the sequence coordination quantity is obtained; According to the sequence coordination quantity, the process sequence satisfying the preset coordination threshold is selected, and the process sequence data in the basic scheduling data is replaced, to obtain updated scheduling data; According to the basic scheduling data, the preposition relationship and the subsequent relationship between the processes are identified, the process correlation data is obtained, and the state change nodes of each process are extracted, to obtain state expansion data, comprising: According to the basic scheduling data, the process identification and the process sequence are divided into a plurality of logical units based on the position difference, to obtain position grouping data; According to the position grouping data, the last process identification of each logical unit is taken as a preposition anchor point, and the first process identification of the adjacent logical unit is taken as a subsequent anchor point, to obtain chain mapping data; According to the chain mapping data, the repeatedly appearing anchor points are folded and the associated paths are established based on the process identification, to obtain a folding path set; According to the folding path set, the associated paths are layered according to the path length and woven into a directed sequence structure, to obtain the process correlation data; According to the execution state data, the process execution process is cut into a plurality of state intervals along the time axis, to obtain a time interval division table; According to the time interval division table, the state symbols of each state interval are sequentially converted into a symbol sequence, and the adjacent state differences are mapped and identified, to obtain an evolution mapping set; According to the evolution mapping set, the time points with a state conversion frequency exceeding a preset frequency are located, to obtain time focus data; According to the time focus data, the time point position and the process identification are mapped and arranged in time sequence, to obtain the state expansion data.

2. The process optimization-based pump body installation schedule management method according to claim 1, characterized by, According to the state expansion data, the distribution of the state change nodes in a unit of time is counted, the concentration degree of the process state change in the time dimension is calculated, and the state aggregation index is obtained, comprising: According to the state expansion data, the node occurrence intensity of the same process in each time section is calculated, to obtain a node rate item; the average value of the node rate of the same process in all time sections is calculated, to obtain a normalized rate item; whether the normalized rate is concentrated in a few time sections is calculated, to obtain a section dominant item; According to the normalized rate term, whether the change of the normalized rate of adjacent time segments is concentrated in a few jump positions is calculated to obtain a change compression term; the influence degree of the prominence of a single time segment on the aggregation determination is calculated to obtain a peak gating term; The segment dominant term, the change compression term and the peak gating term are fused and interacted to calculate the concentration degree of the process state change node in the time dimension to obtain a state aggregation index.

3. The process optimization-based pump body installation schedule management method according to claim 2, characterized by, According to the state aggregation index, the change fluctuation relationship between adjacent processes is described, and the state instability degree of the process in the execution process is calculated to obtain a process fluctuation value, including: According to the process order data and the state aggregation index, the order difference degree of the state aggregation indexes of adjacent processes in the aggregation level is calculated to obtain an aggregation offset term; the adoptable degree of adjacent process pairs in the preceding relationship and the subsequent relationship is calculated to obtain a correlation correction term; According to the process order data, the synchronous change degree of the state change nodes of adjacent processes in the same time segment is calculated to obtain a node co-occurrence term; the deviation degree between the aggregation level and the node synchronization relationship is calculated to obtain a coupling development term; The aggregation offset term, the correlation correction term, the node co-occurrence term and the coupling development term are fused to calculate the state instability degree of the process in the execution process to obtain the process fluctuation value.

4. The process optimization-based pump body installation schedule management method according to claim 3, characterized by, According to the process fluctuation value, each process is divided into a plurality of fluctuation segments, and the processes in each fluctuation segment are reorganized to obtain a candidate process sequence, including: The process fluctuation value is written into the corresponding process identifier according to the process order to form a fluctuation entry sequence; According to the fluctuation entry sequence, the position where the process fluctuation value changes across the segment identifier is written into a segmentation mark, and the fluctuation entry sequence is split into a plurality of fluctuation segments by the segmentation mark to obtain fluctuation segment data; According to the fluctuation segment data, the process identifiers in each fluctuation segment are generated in the order of the size of the process fluctuation value, and the segment processes are reorganized according to the intra-zone order to obtain a segment reorganization sequence; According to the segment reorganization sequence, the reorganized segment processes are spliced according to the fluctuation segment data to obtain the candidate process sequence.

5. The process optimization-based pump body installation schedule management method according to claim 4, characterized by, According to the fluctuation segment data, the process identifiers in each fluctuation segment are generated in the order of the size of the process fluctuation value, and the segment processes are reorganized according to the intra-zone order to obtain a segment reorganization sequence, including: According to the fluctuation segment data, the process identifiers of adjacent positions in the same fluctuation segment are paired two by two, and the corresponding process fluctuation value is hung for each pair of paired entries to obtain a paired entry set; By writing the process identifier with a larger process fluctuation value in the paired entry set into the first position of the segment, and writing the process identifier with a smaller process fluctuation value into the last position of the segment, a segment sequence set is obtained; According to the segment sequence set, the process identifier at the last position is connected with the process identifier at the first position of another segment, and when direct connection is not possible, the original position in the zone is introduced as a connection key to obtain serial registration data; By expanding the segment sequence in the serial registration data into an intra-zone order in sequence according to the connection key, and reorganizing the segment processes according to the intra-zone order, a segment reorganization sequence is obtained.

6. The process optimization-based pump body installation schedule management method according to claim 5, characterized by, According to the candidate process sequence, the matching degree between the candidate process sequence and the process correlation data is calculated to obtain a sequence coordination quantity, including: According to the candidate process sequence, a position mapping relationship of each process identifier in the candidate process sequence is calculated to obtain a sequence bit item; a hierarchical structure degree of a directed relationship of a preceding relationship and a subsequent relationship in the association link is calculated to obtain an association hierarchical item; According to the sequence bit item and the association hierarchical item, a coverage strength of each directed relationship pair on the association relationship of the candidate process sequence pair is calculated to obtain a relationship coverage item; a consistency degree between adjacent splicing of the candidate process sequence and the association link is calculated to obtain an adjacent penetration item; an inhibition degree of a violation degree of the association relationship by the candidate process sequence is calculated to obtain a reverse sequence penalty item; The relationship coverage item, the adjacent penetration item and the reverse sequence penalty item are fused to calculate a matching degree between the candidate process sequence and the process association data, and a sequence coordination quantity is obtained.

7. The process optimization-based pump body installation schedule management method according to claim 6, characterized by, According to the sequence coordination quantity, a process sequence whose sequence coordination quantity meets a preset coordination threshold is selected, and the process order data in the basic scheduling data is replaced to obtain updated scheduling data, including: According to the candidate process sequence and the sequence coordination quantity, a threshold comparison data is obtained by threshold comparison between the preset coordination threshold and the sequence coordination quantity corresponding to the candidate process sequence; According to the threshold comparison data, the candidate process sequence whose sequence coordination quantity meets the preset coordination threshold is retained, and the remaining candidate process sequence is excluded to obtain a retained sequence set; According to the retained sequence set, the candidate process sequence in the retained sequence set is judged according to the combination order of the sequence coordination quantity and the sequence bit item to obtain a target process sequence; According to the target process sequence, the arrangement order of the corresponding process identifier in the process order data is replaced by the arrangement order of the target process sequence to obtain the updated scheduling data.

8. The process optimization-based pump body installation schedule management method according to claim 7, characterized by, According to the retained sequence set, the candidate process sequence in the retained sequence set is judged according to the combination order of the sequence coordination quantity and the sequence bit item to obtain a target process sequence, including: According to the retained sequence set, a sequence number is written for each candidate process sequence and the corresponding sequence coordination quantity is connected to obtain coordination connection data; According to the coordination connection data, the sequence bit item corresponding to each candidate process sequence is extracted as a bit vector, and the bit vector is bound with the sequence number to obtain bit binding data; According to the bit binding data, the sequence coordination quantity and the bit vector are written in the same combination field to form a judgment key to obtain judgment key data; According to the judgment key data, the candidate process sequence is sorted according to the judgment key, and the first candidate process sequence in the sorting result is selected as the target process sequence.

Images