A simulation method for lossless removal and replacement of a single water pump in a multi-pump parallel system

Abstract

This invention discloses a simulation method for non-destructive removal and replacement of a single water pump in a multi-pump parallel system, relating to the field of pump group maintenance simulation technology. The method includes: collecting real-time data from multiple pumps in parallel and performing digital twin synchronous calibration to obtain a pre-removal twin operating baseline; extracting the target water pump's control logic and dynamic adjustment state before power failure from the pre-removal twin operating baseline; and solidifying the pre-shutdown control holding information into a continuously retrievable pre-power failure control state image package; using the pre-power failure control state image package to drive a shadow control loop for virtual probing and inverting the pressure and flow disturbance propagation process caused by the target water pump's withdrawal to form a disturbance peak-shaving isolation window; using the disturbance peak-shaving isolation window to drive the digital twin model for single-pump removal and replacement drills, and simultaneously verifying load-bearing stability and control state continuity to form a single-pump removal and replacement simulation record. This invention achieves the effect of improving on-site removal safety and the accuracy of optimal replacement operation scheme access.

Description

Technical Field

[0001] This invention relates to the field of pump set maintenance simulation technology, and in particular to a simulation method for non-destructive removal and replacement of a single water pump in a multi-pump parallel system. Background Technology

[0002] As water supply, drainage, industrial circulating water, and HVAC systems evolve towards continuous operation and intelligent maintenance, multi-pump parallel systems are widely used for flow regulation and operational redundancy. Existing technologies typically optimize pump unit operation through variable frequency control, pressure closed-loop, flow balancing, and digital twin monitoring, and can provide online display or simulation analysis of pump unit status, valve status, and control signals. However, their focus is mainly on operation monitoring, fault diagnosis, and energy efficiency management, and simulation guidance for the entire process of removing and replacing a single pump under conditions of system non-stop operation or low disturbance is still insufficient.

[0003] The existing maintenance methods for multi-pump parallel systems have two main shortcomings: First, when a single pump is taken out of service, existing methods rely heavily on fixed shutdown sequences or manual experience, making it difficult to anticipate the propagation process of pressure and flow disturbances. This can easily lead to sudden load changes and fluctuations in mains pipe pressure at the moment of power failure. Second, existing simulation maintenance schemes typically only verify individual operations such as shutdown, valve closure, and pressure relief. They lack a unified verification of the control status before power failure, the load taken over by the remaining pumps, and the removal of residual energy. This results in the risk of misoperation and loss of status during on-site removal and replacement. Summary of the Invention

[0004] In view of the aforementioned existing problems, the present invention is proposed.

[0005] Therefore, this invention provides a simulation method for non-destructive removal and replacement of a single pump in a multi-pump parallel system, which solves the problems of unpredictable disturbance propagation before removal and replacement of a single pump and difficulty in uniformly verifying the safety status of removal.

[0006] To solve the above-mentioned technical problems, the present invention provides the following technical solution: This invention provides a simulation method for non-destructive removal and replacement of a single water pump in a multi-pump parallel system. The method includes: acquiring real-time data from multiple pumps in parallel and performing digital twin synchronous calibration to obtain a pre-removal twin operating baseline; extracting the target water pump's control logic and dynamic adjustment state before power failure from the pre-removal twin operating baseline; and solidifying the pre-shutdown control holding information into a continuously retrievable pre-power failure control state image package; driving a shadow control loop through the pre-power failure control state image package to perform virtual probing and inverting the pressure and flow disturbance propagation process caused by the target water pump's withdrawal, forming a disturbance peak-shifting isolation window; and using disturbance peak-shifting isolation... A window-driven digital twin model was used to conduct a single-pump dismantling and replacement drill, simultaneously verifying load bearing stability and control state continuity, thus generating a single-pump dismantling and replacement simulation record. Based on the single-pump dismantling and replacement simulation record, a target pump dismantling energy island was constructed, and residual energy nodes in the target branch were checked for barrierization. Nodes that did not meet the dismantling access conditions were converted into executable constraints, forming a dismantling safety constraint set. Based on the dismantling safety constraint set and the single-pump dismantling and replacement simulation record, operation paths with excessive fluctuations and unconverged residual energy were screened out, and the shutdown isolation and reconnection grid sequence was optimized to generate the optimal replacement operation plan.

[0007] As a preferred embodiment of the simulation method for non-destructive removal and replacement of a single water pump in a multi-pump parallel system described in this invention, the real-time data of the parallel multi-pump system includes hydraulic operation data, water pump operation data, valve status data, electrical operation data, control loop data, and safety maintenance data.

[0008] As a preferred embodiment of the simulation method for non-destructive removal and replacement of a single water pump in a multi-pump parallel system described in this invention, the specific steps for obtaining the pre-removal twin operating baseline are as follows: Collect real-time data from multiple pumps operating in parallel, configure the collection time and device identifier for the real-time data of multiple pumps operating in parallel, and obtain the original operating data frame; The original running data frames are time-aligned and anomaly-removed, and topology mapping is performed according to pipeline connection relationships, electrical power supply relationships, and control signal relationships to aggregate the hydraulic state and electrical control state at the same time into topology synchronization state data. Based on the topology synchronization state data, the pressure and flow response, electrical response, and control hold-up states in the twin operating space are calibrated, and the system operating boundary before the target water pump is removed is locked to obtain the twin operating benchmark before removal.

[0009] As a preferred embodiment of the simulation method for non-destructive removal and replacement of a single water pump in a multi-pump parallel system described in this invention, the specific steps of extracting the target water pump's control logic and dynamic adjustment state before power failure from the pre-disassembly twin operating baseline, and solidifying the pre-shutdown control holding information into a continuously callable pre-power failure control state image package are as follows: Based on the pre-dismantling twin operation benchmark, locate the target water pump power failure action related objects, and extract the interlock judgment and adjustment trigger relationship from the control points corresponding to each related object according to the correlation strength to form the control logic chain before power failure; Based on the control logic chain before power failure, read the continuous adjustment feedback within the time window before shutdown, and identify the control state that needs to be maintained according to the direction and magnitude of change, thus forming the control holding information before shutdown; The control state of the target water pump before power failure is encapsulated using the pre-shutdown control hold information, and a verification identifier and playback sequence are configured to form a mirror package of the control state before power failure.

[0010] As a preferred embodiment of the simulation method for non-destructive removal and replacement of a single water pump in a multi-pump parallel system described in this invention, the specific steps of performing virtual probing by driving a shadow control loop through a control state mirror packet before power failure are as follows: Perform integrity verification on the control state image package before power failure, load the control logic and hold state according to the playback order, and generate a virtual execution sequence; A dual-channel shadow control loop is established based on the virtual execution sequence. The control flow that runs according to the control logic playback order in the virtual execution sequence is used as the reference playback flow, and the control flow that adds virtual probing actions according to the same playback order is used as the virtual probing flow. The pressure boundary, flow boundary, and control hold boundary of the virtual test flow are verified around the dual-channel shadow control loop, and a safe virtual test action set is obtained by screening.

[0011] As a preferred embodiment of the simulation method for non-destructive removal and replacement of a single water pump in a multi-pump parallel system described in this invention, the specific steps for forming a disturbance peak-shifting isolation window are as follows: Within the set of safe virtual probing actions, frequency reduction probing, compensation probing, and slow closing probing are performed sequentially, and the differences in disturbance response relative to the reference playback stream are recorded to generate a probing response event stack. Extract the pressure response differences and flow response differences corresponding to different probing actions from the probing response event stack, and split them into perturbation atom libraries according to the source of the actions; By comparing the arrival time and decay characteristics of each perturbation atom at different measurement points using the perturbation atom library, the perturbation propagation path diagram of the target water pump exiting the perturbation is derived. Based on the perturbation propagation path diagram, the response overlap and attenuation buffer relationship of the perturbation atoms at the same measurement point are determined, and a phase conflict diagram is formed; The disturbance isolation degree of each candidate peak-shifting window is calculated based on the phase conflict diagram, and the candidate peak-shifting window with the highest disturbance isolation degree that satisfies the pressure boundary and flow boundary is selected as the disturbance peak-shifting isolation window.

[0012] A preferred embodiment of the simulation method for non-destructive removal and replacement of a single water pump in a multi-pump parallel system according to the present invention, wherein the specific steps for forming a single pump removal and replacement simulation record are as follows: Extract the execution sequence of each key action from the disturbance peak isolation window, and connect the shutdown isolation and reinstallation grid connection process into the same timing chain to generate a peak-shaving exercise instruction table; Based on the peak-shifting drill instruction list, the pre-dismantling twin operation baseline and the pre-power-out control status image package are invoked, and a rollback-capable transaction drill space is constructed. Using a rollback-capable transaction simulation space to drive a digital twin model, the process of target pump shutdown and new pump connection is simulated according to the off-peak simulation instruction table, forming a replacement simulation trajectory that records load changes and control hold status. The stability of the remaining water pump load and the continuity of the control state are simultaneously verified based on the replacement exercise trajectory, and the replacement exercise trajectory that meets the simulation access criteria is solidified into a single pump replacement simulation record.

[0013] As a preferred embodiment of the simulation method for non-destructive removal and replacement of a single water pump in a multi-pump parallel system described in this invention, the specific steps for constructing the target water pump removal energy island based on the single pump replacement simulation record are as follows: Extract the action sequence and state trajectory of the target water pump from the single pump replacement simulation record to the mechanical dismantling, and delineate the target water pump dismantling boundary; Based on the target water pump removal boundary, trace the connection points within the target water pump branch that may still retain energy and re-input energy segment by segment, and bind the residual state of each connection point with the corresponding isolation action into a traceable multi-energy domain node graph; The state changes of each energy node in the simulation record are read around the multi-energy domain node graph, and the residual energy release process is recorded as an energy debt ledger; The energy debt ledger is used to uniformly determine the degree of residual energy release within the target water pump removal boundary and the risk of re-input outside the boundary, and energy nodes that have not met the removal safety requirements are marked to obtain the closed state of the energy island; Based on the closed state of the energy island, the isolation actions and detection feedback corresponding to the unconverged energy nodes are traced back, and the boundary of the target water pump removal is supplemented and corrected. The boundary information and energy debt ledger are then solidified to form the target water pump removal energy island.

[0014] As a preferred embodiment of the simulation method for non-destructive removal and replacement of a single water pump in a multi-pump parallel system described in this invention, the specific steps for forming the removal safety constraint set are as follows: The residual and isolated states of the energy nodes in the branch are read from the energy island removed from the target water pump, and organized into an energy verification object chain according to the order in which energy enters the branch. Based on the energy verification object chain, the changes in node residual quantity and the effectiveness of isolation actions are continuously compared, and the node states that meet the verification rules are aggregated into an energy barrier state table; The energy convergence of each energy node in the target water pump branch is calculated using the energy barrier status table, and the energy nodes whose energy convergence has not reached the preset convergence threshold are bound to the corresponding isolation action to obtain the list of safety constraints to be removed. Based on the list of safety constraints for dismantling, and combined with the boundary information of the target water pump dismantling energy island, the constraint execution conditions and release conditions are configured to form a set of safety constraints for dismantling.

[0015] As a preferred embodiment of the simulation method for non-destructive removal and replacement of a single water pump in a multi-pump parallel system described in this invention, the specific steps for generating the optimal replacement operation plan are as follows: Extract each candidate operation path and its corresponding state trajectory from the single pump replacement simulation record, and match each candidate operation path with the admission conditions in the dismantling safety constraint set to obtain the replacement path map to be screened. Based on the selection of the replacement path graph, operation paths with excessive fluctuations, state interruptions, and unconverged residual energy are identified, and path nodes and dependent paths that do not meet the removal safety constraint set are deleted to obtain the admission replacement path graph. Based on the access replacement path diagram and the disturbance peak isolation window, the execution sequence of shutdown isolation action and reinstallation grid connection action is rearranged, and the target water pump exit process and system recovery process are connected to obtain a set of candidate replacement paths; Based on the candidate replacement path set, the path is optimized according to the principles of operational stability, control continuity, and energy convergence. The execution order, feedback conditions, and backoff relationships in the optimized path are then solidified into the optimal replacement operation plan.

[0016] The beneficial effects of this invention are as follows: By using the control state mirror package before power failure to drive the shadow control loop for virtual probing and inverting the propagation process of pressure and flow disturbances caused by the withdrawal of the target water pump, the disturbance response caused by frequency reduction, compensation, and slow shutdown actions can be identified in advance before the actual shutdown, thereby improving the stability of single pump withdrawal and the reliability of simulation before replacement; by constructing the target water pump removal energy island through single pump replacement simulation records and uniformly converging and verifying the residual risks in the target water pump branch, the overall judgment of the residual energy release degree and external re-input risk within the target water pump removal boundary can be achieved, thereby improving the safety of on-site removal and the accuracy of the optimal replacement operation plan access. Attached Figure Description

[0017] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the following description of the embodiments will be briefly introduced. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0018] Figure 1 This is a flowchart illustrating a simulation method for the non-destructive removal and replacement of a single water pump in a multi-pump parallel system.

[0019] Figure 2 A flowchart for creating a mirror image of the pre-disassembly twin operating baseline and the control state before power failure.

[0020] Figure 3 A flowchart for forming the shadow control loop and disturbance peak isolation window.

[0021] Figure 4 This is a flowchart for a rollback single-pump replacement drill. Detailed Implementation

[0022] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, the specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings.

[0023] Many specific details are set forth in the following description in order to provide a full understanding of the invention. However, the invention may also be practiced in other ways different from those described herein, and those skilled in the art can make similar extensions without departing from the spirit of the invention. Therefore, the invention is not limited to the specific embodiments disclosed below.

[0024] Secondly, the term "one embodiment" or "embodiment" as used herein refers to a specific feature, structure, or characteristic that may be included in at least one implementation of the present invention. The phrase "in one embodiment" appearing in different places in this specification does not necessarily refer to the same embodiment, nor is it a single or selective embodiment that is mutually exclusive with other embodiments.

[0025] Reference Figures 1-4 As one embodiment of the present invention, this embodiment provides a simulation method for non-destructive removal and replacement of a single water pump in a multi-pump parallel system, including the following steps: S1. Collect real-time data of multiple pumps in parallel and perform digital twin synchronous calibration to obtain the pre-dismantling twin operating benchmark. Extract the control logic and dynamic adjustment status of the target water pump before power failure from the pre-dismantling twin operating benchmark, and solidify the pre-shutdown control holding information into a continuously callable pre-power failure control status image.

[0026] Collect real-time data from multiple pumps connected in parallel, and configure the collection time and device identifier for the real-time data of multiple pumps connected in parallel to obtain the original operating data frame.

[0027] The specific process includes: when collecting real-time data from multiple pumps in parallel, reading hydraulic operation data at pipeline pressure and flow measurement points, reading pump operation data at pump operation monitoring points, reading valve status data at valve position feedback devices, reading electrical operation data at frequency converters, circuit breakers, and electrical detection points, reading control circuit data at PLC control points, and reading safety maintenance data from maintenance permit, isolation confirmation, and pressure relief confirmation records. The data content obtained from each measurement point is bound to the corresponding acquisition time, and the pump, valve, pipeline, electrical equipment, and control point to which each measurement point belongs are written into the equipment identifier. The acquisition time is used to determine the position of different data entering the same time series, and the equipment identifier is used to determine the actual source of different data, so that the real-time data of multiple pumps in parallel has a traceable relationship in both the time dimension and the equipment source dimension, and the original operation data frame is obtained.

[0028] The original operating data frames are time-aligned and anomaly-removed. Topology mapping is then performed according to pipeline connection relationships, electrical power supply relationships, and control signal relationships to aggregate the hydraulic and electrical control states at the same time into topology-synchronized state data.

[0029] The specific process includes: when aligning the original operating data frames, using the acquisition time as a unified index, adjusting pressure readings, flow readings, valve opening feedback, current feedback, frequency feedback, and control commands to the same sampling time; removing anomalies from the adjusted data according to the measurement point range, sampling interval, adjacent change trends, and communication status, deleting out-of-bounds values, sudden jump values, duplicate values, and interrupted values; locating the hydraulic position according to the pipeline connection relationship, the power supply position according to the electrical power supply relationship, and the control point according to the control signal relationship for the retained data, mapping the hydraulic state and electrical control state at the same time to the target pump, parallel branch, valve, and control point, and aggregating them into topology synchronization state data.

[0030] Based on the topology synchronization state data, the pressure and flow response, electrical response, and control hold-up states in the twin operating space are calibrated, and the system operating boundary before the target water pump is removed is locked to obtain the twin operating benchmark before removal.

[0031] The specific process includes, when calibrating the twin operating space based on the topology synchronization state data, writing the pressure feedback, flow feedback, electrical feedback and control feedback corresponding to the same moment into the twin operating space, and comparing them item by item with the pressure, flow response, electrical response and control holding state in the twin operating space; when the comparison deviation exceeds the allowable range, correcting the corresponding state according to the pipeline connection relationship, electrical power supply relationship and control signal relationship, until the pressure range, flow range, electrical response range and control holding range of the target water pump and parallel branches before dismantling are stable and consistent, thus obtaining the twin operating benchmark before dismantling.

[0032] It should be noted that the allowable range is a state deviation range pre-set based on the operating boundary, equipment rated parameters, sensor range and accuracy, and on-site safety and maintenance requirements before the target water pump is removed.

[0033] Based on the pre-dismantling twin operation benchmark, the target water pump power failure action associated objects are located, and the interlock judgment and adjustment trigger relationship is extracted from the control points corresponding to each associated object according to the association strength, forming the control logic chain before power failure.

[0034] The specific process includes reading the branch, power supply path, inverter status, circuit breaker status, inlet valve status, outlet valve status, remaining pump compensation status, and PLC control points of the target water pump before power failure based on the pre-dismantling twin operation benchmark. The equipment that will participate in the target water pump's frequency reduction, power failure, valve locking, and load compensation after the shutdown command is issued is identified as the associated objects of the target water pump's power failure action. The association strength is determined based on the degree of influence of each associated object on pressure stability, flow maintenance, electrical isolation, and control maintenance when the target water pump stops. The sequential relationship between start permission, shutdown interlock, frequency reduction trigger, valve locking, and compensation adjustment is extracted from the control points corresponding to each associated object according to the association strength, forming the control logic chain before power failure.

[0035] Based on the control logic chain before power failure, the continuous adjustment feedback within the time window before shutdown is read, and the control state that needs to be maintained is identified according to the direction and magnitude of change, thus forming the control holding information before shutdown.

[0036] The specific process includes determining the trigger positions and sequence of each control action before the target water pump stops based on the control logic chain before the power outage; continuously reading the frequency feedback of the inverter, the output feedback of the PLC, the valve position feedback, the pressure regulation feedback, the flow regulation feedback, and the remaining water pump compensation feedback at the acquisition time within the time window before the stop; determining the increase or decrease trend of feedback between adjacent acquisition times as the direction of change; determining the degree of feedback deviation from the stable operating range as the magnitude of change; and organizing the interlock permission state, regulation output state, valve holding state, and compensation acceptance state that still need to continue after the stop into the control holding information before the stop.

[0037] The control state of the target water pump before power failure is encapsulated using the pre-shutdown control hold information, and a verification identifier and playback sequence are configured to form a mirror package of the control state before power failure.

[0038] The specific process includes: when encapsulating the control state of the target water pump before power failure using the control hold information before shutdown, organizing the interlocking permission state, regulation output state, valve hold state, and compensation acceptance state according to the trigger sequence of the control logic chain before power failure, and writing the acquisition time, equipment identifier, feedback value, hold value, and consistency verification content corresponding to each control state into the encapsulated content; the verification identifier is generated based on the written content and is used to verify the integrity of the content later; the playback order is configured according to the control trigger sequence and is used for subsequent sequential reading and reproduction, forming a mirror package of the control state before power failure.

[0039] It should be noted that the control state image package before power failure is a verifiable and replayable data set of the target water pump's control logic, dynamic adjustment state, and holding state before power failure, used to support the continuous operation of the shadow control loop.

[0040] S2. Virtual probing is performed by driving the shadow control loop through the control state mirror package before power failure, and the propagation process of pressure and flow disturbance caused by the exit of the target water pump is inverted to form a disturbance peak isolation window.

[0041] The integrity of the control state image package before the power failure is verified, and the control logic and the held state are loaded in the playback order to generate a virtual execution sequence.

[0042] The specific process includes: when performing integrity verification on the control state image package before power failure, reading the verification identifier, acquisition time, device identifier, control logic chain before power failure, dynamic adjustment status, and control hold information before shutdown, and checking each item of the packaged content for missing items, out-of-order, duplicate writing, and inconsistent feedback; after the verification is passed, reading the trigger sequence relationship in the control logic chain before power failure according to the playback order, loading the interlock judgment, shutdown trigger, adjustment output, valve hold, and compensation acceptance status into the virtual execution environment in sequence, while keeping the device identifier and acquisition time corresponding to each control status unchanged, and arranging the loaded control states in a continuously readable order to generate a virtual execution sequence.

[0043] A dual-channel shadow control loop is established based on the virtual execution sequence. The control flow that runs in the control logic playback order in the virtual execution sequence is used as the base playback flow, and the control flow that adds virtual probing actions in the same playback order is used as the virtual probing flow.

[0044] The specific process includes establishing a dual-channel shadow control loop based on the virtual execution sequence. The interlock judgment, adjustment trigger, valve holding and compensation acceptance sequence already arranged in the virtual execution sequence are divided into two synchronously running control flows. One control flow does not change the control logic playback sequence and serves as the reference playback flow. The other control flow adds virtual trial actions corresponding to the target water pump frequency reduction, the remaining water pump compensation, the outlet valve slow closing and the control holding delay in the same playback sequence and serves as the virtual trial flow. Neither control flow issues real control commands to the field equipment. They are only used to compare the changes in pressure, flow, load and control holding before and after the actions are added, thus forming a dual-channel shadow control loop.

[0045] The pressure boundary, flow boundary, and control hold boundary of the virtual test flow are verified around the dual-channel shadow control loop, and a safe virtual test action set is obtained by screening.

[0046] The specific process includes: when verifying the virtual test flow around the dual-channel shadow control loop, reading the pressure change, flow rate change, and control hold feedback corresponding to each virtual test action in the virtual test flow, comparing the pressure change with the pressure boundary, the flow rate change with the flow rate boundary, and the control hold feedback with the control hold boundary; when the virtual test action does not break the boundary during the target pump frequency reduction, residual pump compensation, outlet valve slow closing, and control hold delay, the virtual test action is retained; when any boundary is broken, the virtual test action is removed and the source of the boundary anomaly is recorded, thus obtaining a set of safe virtual test actions.

[0047] It should be noted that the pressure boundary, flow boundary, and control holding boundary are jointly determined by the system operating boundary, equipment rated parameters, historical stable operating range, and on-site safety and maintenance requirements in the pre-dismantling twin operating baseline.

[0048] Within the set of secure virtual probing actions, frequency reduction probing, compensation probing, and slow-closing probing are performed sequentially, and the differences in disturbance response relative to the reference playback stream are recorded to generate a probing response event stack.

[0049] The specific process includes, when conducting a test within the set of safe virtual test actions, first performing a frequency reduction test according to the control sequence before the target water pump exits, then performing a residual water pump compensation test, and then performing an outlet valve slow-closing test. After each test action, the pressure change, flow change, residual water pump load change, and control hold feedback in the virtual test flow are read. At the same time, the corresponding state of the reference playback flow at the same playback time is read. The state deviation between the two control flows is taken as the disturbance response difference and recorded in sequence according to the test action type, test action time, test action amplitude, and disturbance response difference to generate a test response event stack.

[0050] It should be noted that the trial response event stack is used to save the differences in disturbance response corresponding to each trial action in chronological order, which facilitates the inversion of the pressure and flow disturbance propagation process and the formation of a disturbance peak isolation window.

[0051] Extract the pressure response differences and flow response differences corresponding to different probing actions from the probing response event stack, and split them into perturbation atom libraries according to the source of the action.

[0052] The specific process includes reading the action source, timing, amplitude, and disturbance response difference of each trial action from the trial response event stack. The pressure and flow response differences corresponding to the frequency reduction trial are included in the frequency reduction disturbance content, the pressure and flow response differences corresponding to the compensation trial are included in the compensation disturbance content, and the pressure and flow response differences corresponding to the slow closure trial are included in the slow closure disturbance content. The response start point, peak time, decay period, and affected measurement points are extracted according to the action source to form a disturbance atom library that can characterize the disturbance features of a single trial action.

[0053] By comparing the arrival time and decay characteristics of each perturbation atom at different measurement points using a perturbation atom library, the perturbation propagation path of the target water pump exiting the perturbation can be derived.

[0054] The specific process includes: when using the perturbation atom library to invert the perturbation propagation path, reading the action source, affected measurement point, response start point, peak time, decay period, and response amplitude corresponding to each perturbation atom, and arranging the measurement points of the target water pump branch, outlet main pipe, remaining water pump branch, and end load side in sequence according to the pipeline connection relationship; comparing the response start points of adjacent measurement points sequentially, determining the measurement points with earlier pressure and flow changes as the pre-propagation positions of the perturbation, and determining the measurement points with later changes of the same trend as the post-propagation positions of the perturbation; and then combining the difference in peak time and the length of decay period to determine the strength of the perturbation propagation, connecting the measurement points with continuous arrival relationship, similar change trends, and decreasing response amplitude as propagation branches, and summarizing to form the perturbation propagation path diagram of the target water pump exiting the perturbation.

[0055] It should be noted that the disturbance propagation path diagram is used to represent the transmission sequence and impact location of pressure and flow disturbances when the target water pump exits. Its purpose is to provide a basis for judging disturbance overlap and generating disturbance peak isolation windows.

[0056] Based on the perturbation propagation path diagram, the response overlap and attenuation buffer relationship of perturbed atoms at the same measurement point are determined, and a phase conflict diagram is formed.

[0057] The specific process includes, when forming a phase conflict diagram using the perturbation propagation path diagram, reading the response start point, peak time, decay period and response amplitude of each perturbation atom at the same measurement point, and comparing the degree of overlap of different perturbation atom in time; when the peak periods of two perturbation atom overlap each other and produce pressure and flow changes in the same direction, the two perturbation atom are marked as having a response overlap relationship; when the latter perturbation atom enters the decay period of the former perturbation atom and can weaken the peak overlap, the two perturbation atom are marked as having a decay buffer relationship, and then the decay buffer relationship is written into the phase conflict diagram.

[0058] It should be noted that the phase conflict diagram is used to represent the response overlap and attenuation buffer relationship of different perturbation atoms at the same measurement point. Its function is to identify which actions are likely to cause pressure and flow superposition impacts, and to provide a basis for staggering the action sequence and forming a perturbation peak isolation window.

[0059] The disturbance isolation degree of each candidate peak-shifting window is calculated based on the phase conflict diagram, and the candidate peak-shifting window with the highest disturbance isolation degree that satisfies the pressure boundary and flow boundary is selected as the disturbance peak-shifting isolation window.

[0060] The specific process includes, when calculating the disturbance isolation degree of each candidate peak-shaving window based on the phase conflict diagram, reading the start time, response overlap duration, superimposed response amplitude, maximum deviation of main pipe pressure and maximum deviation of flow for each candidate peak-shaving window corresponding to the frequency reduction test, compensation test and slow closing test, one by one, and comparing the maximum deviation of main pipe pressure with the pressure boundary, comparing the maximum deviation of flow with the flow boundary, and substituting the response overlap duration and superimposed response amplitude into the disturbance isolation degree calculation process for sorting, retaining the candidate peak-shaving window that has not broken through the pressure boundary and flow boundary and has the highest disturbance isolation degree, thus forming the disturbance peak-shaving isolation window.

[0061] The expression for calculating the disturbance isolation degree is: ; in, Indicates the first The disturbance isolation degree of each candidate staggered window; Indicates the sequence number of the candidate staggered peak window; Represents the natural exponential function; Indicates the first The maximum deviation of the main pipe pressure relative to the reference pressure within each candidate peak-shifting window; Indicates the allowable pressure deviation; Indicates the first The maximum deviation of system flow relative to baseline flow within each candidate staggered peak window; Indicates the allowable flow deviation; Indicates the first The number of perturbation atom combinations that cause response overlap within a candidate staggered window; Indicates the sequence number of the perturbed atom combination; Indicates the first Within the candidate staggered peak window The duration of the response overlap of a combination of perturbed atoms; Indicates the first The total duration of each candidate off-peak window; Indicates the first Within the candidate staggered peak window The superposition response amplitude of a combination of perturbation atoms during the overlapping time period; This indicates that the superimposed response magnitude is allowed.

[0062] It should be noted that, It is determined by the stable pressure range of the main pipe, the rated pressure range of the equipment, and the on-site safety maintenance requirements in the pre-dismantling twin operating benchmark. It is determined by the system flow stability range, water load demand, and remaining pump capacity in the pre-dismantling twin operating benchmark. It is determined by the historical response amplitude of each probe action in the perturbation atom library, the overlap relationship in the phase conflict diagram, and the pressure-flow boundary.

[0063] S3. Use the disturbance peak isolation window to drive the digital twin model to conduct a single pump removal and replacement exercise, and simultaneously verify the load bearing stability and control state continuity to form a single pump removal and replacement simulation record.

[0064] Extract the execution sequence of each key action from the disturbance peak isolation window, and connect the shutdown isolation and reinstallation grid connection processes into the same timing chain to generate a peak-shaving exercise instruction table.

[0065] The specific process includes reading the start time, duration, allowable offset, and measurement point response time corresponding to the frequency reduction test, compensation test, and slow closure test from the disturbance peak isolation window, and arranging the frequency reduction exit, compensation takeover, outlet valve slow closure, electrical isolation, and pressure relief confirmation that need to be performed in the target water pump shutdown isolation according to the window sequence. The parameter restoration, low-frequency trial operation, and grid connection restoration that need to be performed in the new pump reinstallation and grid connection are connected to the subsequent time period, so that the target water pump exit action and the new pump connection action form a continuous and rehearsable same time sequence chain, and generating a peak-shaving exercise instruction table.

[0066] Based on the peak-shifting drill instruction table, the pre-dismantling twin operation baseline and the pre-power-out control status image package are invoked, and a rollback-capable transaction drill space is constructed.

[0067] The specific process includes: when calling the pre-dismantling twin operating baseline and the pre-power-out control state mirror package according to the staggered peak exercise instruction table, reading the pressure and flow response, electrical response, and control holding status of the target water pump before its withdrawal according to the action sequence in the staggered peak exercise instruction table, and reading the control logic, holding status, verification flag, and playback sequence in the pre-power-out control state mirror package; when constructing a rollback-capable transaction exercise space, copying the pre-action state corresponding to each exercise action as the exercise starting point, writing the pressure change, flow change, electrical change, and control holding change after the action is executed into the post-action state, and configuring verification records and rollback flags for the pre-action state and post-action state. If any exercise action is abnormal, it can be restored to the corresponding pre-action state according to the rollback flag, forming a rollback-capable transaction exercise space.

[0068] The digital twin model is driven by a rollback-capable transaction simulation space to simulate the process of target pump shutdown and new pump connection according to the staggered peak simulation instruction table, forming a replacement simulation trajectory that records load changes and control hold status.

[0069] The specific process includes, when using the rollback-capable transaction simulation space to drive the digital twin model, simulating the target water pump frequency reduction and shutdown, the remaining water pump compensation and takeover, the outlet valve slow closing, electrical isolation, pressure relief confirmation, new pump parameter restoration, low-frequency trial operation, and reinstallation and grid connection process item by item according to the action sequence in the staggered peak exercise instruction table; after each simulation action is completed, the rollback-capable transaction simulation space simultaneously saves the state before and after the action, and records the load changes, pressure and flow changes, valve action feedback, and control holding status of the remaining water pump; when the simulation action deviates, the rollback-capable transaction simulation space restores to the state before the corresponding action according to the rollback mark, and after continuous simulation, a replacement simulation trajectory is formed that records the load changes and control holding status.

[0070] The training process of the digital twin model is as follows: historical pressure data, flow data, valve opening data, electrical feedback data, and control command data of multiple pumps in parallel under different operating conditions are collected. The collected data is then time-aligned, anomaly-removed, and labeled with operating conditions before being input into the digital twin model. The pressure and flow response, electrical response, and control hold-up status output by the digital twin model are compared with the actual measured status on site. Based on the comparison deviation, the pump performance parameters, pipeline resistance parameters, valve action parameters, and control response parameters are continuously corrected until the digital twin model can stably reproduce the changes in on-site operation under various conditions, including normal operation of the target pump, frequency reduction and shutdown, compensation by the remaining pump, and slow valve closure. This yields a digital twin model that can be used for dismantling and replacement simulation.

[0071] The stability of the remaining water pump load and the continuity of the control state are simultaneously verified based on the replacement exercise trajectory, and the replacement exercise trajectory that meets the simulation access criteria is solidified into a single pump replacement simulation record.

[0072] The specific process includes synchronously reading the remaining pump load changes, pressure and flow changes, valve action feedback, and control holding status corresponding to each exercise action based on the replacement exercise trajectory. The remaining pump load changes are compared with the load boundary in the pre-removal twin operating benchmark, and the control holding status is compared with the holding status in the control status mirror package before power failure. The pre-action and post-action states are combined to determine whether the remaining pump smoothly takes over the load and whether the control status is continuously maintained after the target pump exits. When the load taking over status and control holding status in the replacement exercise trajectory both meet the simulation access criteria, the replacement exercise trajectory, along with the corresponding action sequence, feedback status, and verification conclusions, are solidified into a single pump replacement simulation record.

[0073] It should be noted that the simulation admission criterion is the verification condition for determining whether the dismantling and replacement exercise trajectory can be solidified into a single pump dismantling and replacement simulation record. It is derived from the load boundary in the pre-dismantling twin operating benchmark, the holding state in the control state mirror package before power failure, and the allowable fluctuation range in the disturbance peak isolation window.

[0074] S4. Construct the target pump removal energy island based on the single pump replacement simulation record, and perform barrier verification on the residual energy nodes in the target branch. Convert the nodes that do not meet the removal access conditions into executable constraints to form a set of removal safety constraints.

[0075] Extract the action sequence and state trajectory of the target water pump from the single pump replacement simulation record to the mechanical dismantling, and delineate the target water pump dismantling boundary.

[0076] The specific process includes reading the continuous action sequence of the target water pump from the single pump replacement simulation record as it transitions from frequency reduction to the mechanical dismantling confirmation stage, and simultaneously reading the pressure changes, flow changes, valve position feedback, electrical isolation feedback, pressure relief feedback, and control holding status before and after each action. The target water pump body, target water pump branch pipe section, inlet valve connection position, outlet valve connection position, pressure relief and venting position, electrical isolation position, and control confirmation position that may still affect the dismantling safety are included in the boundary range. The relationship between the inside and outside of the boundary is determined based on the energy residue and isolation confirmation status of each position in the action sequence, and the dismantling boundary of the target water pump is delineated.

[0077] It should be noted that the boundary range is determined by the action sequence, state trajectory, energy residue location, and isolation confirmation location in the single pump replacement simulation record.

[0078] Based on the target water pump removal boundary, the connection locations within the target water pump branch that may still retain energy and re-input energy are traced segment by segment, and the residual state of each connection location is bound to the corresponding isolation action as a traceable multi-energy domain node graph.

[0079] The specific process includes, when tracing the target water pump branch segment by segment according to the target water pump dismantling boundary, sequentially reading the pressure holding status, residual liquid retention status, reverse flow entry status, electrical isolation feedback, and control holding feedback along the target water pump body to the inlet valve connection position, outlet valve connection position, pressure relief and venting connection position, electrical isolation connection position, and control confirmation connection position, and determining whether there is still energy retention and the possibility of energy re-input at each connection position before mechanical dismantling; recording the residual state of connection positions with risks, and establishing a binding relationship between the residual state and the corresponding valve isolation, power-off isolation, pressure relief and venting, and control confirmation actions to form a traceable multi-energy domain node diagram.

[0080] It should be noted that the multi-energy domain node diagram is used to represent the relationship between the residual state and isolation action of each connection position within the boundary of the target water pump removal. Its function is to provide a basis for registering the energy debt ledger and judging whether the residual energy has converged.

[0081] The state changes of each energy node in the simulation record are read around the multi-energy domain node graph, and the residual energy release process is recorded as an energy debt ledger.

[0082] The specific process includes reading the pre-release state, isolation action execution state, detection feedback state, and post-release state of each energy node in the single pump replacement simulation record around the multi-energy domain node diagram. The pressure residue, electrical residue, reverse flow entry, and control trigger risk corresponding to the energy node are recorded as release progress according to the action sequence. The unreleased part, released part, detection confirmation content, and corresponding isolation action are registered in the same ledger entry to form an energy debt ledger.

[0083] The energy debt ledger is used to uniformly determine the degree of residual energy release within the target water pump removal boundary and the risk of re-input outside the boundary, and energy nodes that have not met the removal safety requirements are marked to obtain the closed state of the energy island.

[0084] The specific process includes, when using the energy debt ledger, reading the unreleased portion, released portion, detection and confirmation content, and corresponding isolation actions for each energy node item by item, comparing the unreleased portion with the dismantling safety requirements, comparing the released portion with the completion status of the release actions, and comparing the detection and confirmation content with the execution status of valve isolation, power-off isolation, pressure relief and venting, and control confirmation; at the same time, checking whether there are still risks of pressure back transmission, electrical back transmission, reverse flow entry, and control mis-triggering at the external connection positions along the target water pump dismantling boundary, marking energy nodes that do not meet the dismantling safety requirements and have risks of re-input from outside the boundary, thus obtaining the closed state of the energy island.

[0085] It should be noted that the safety requirements for dismantling are the residual energy limits and isolation confirmation conditions that must be met before the target water pump is allowed to enter the mechanical dismantling process. These requirements are derived from the equipment's rated parameters, on-site maintenance procedures, pre-dismantling twin operating benchmarks, and single-pump replacement simulation records.

[0086] Based on the closed state of the energy island, the isolation actions and detection feedback corresponding to the unconverged energy nodes are traced back, and the boundary of the target water pump removal is supplemented and corrected. The boundary information and energy debt ledger are then solidified to form the target water pump removal energy island.

[0087] The specific process includes: when tracing back the non-converged energy nodes based on the closed state of the energy island, reading the residual state records, isolation action execution records, and detection feedback records of each non-converged energy node in the energy debt ledger to confirm whether the non-convergence is due to inadequate isolation, unconfirmed detection, re-input of risks from outside the boundary, or an incomplete release process; re-incorporating the connection locations with risks into the target pump removal boundary and supplementing the record with the boundary attributes, isolation actions, and detection feedback of the connection locations; when the supplemented and corrected target pump removal boundary covers all non-converged energy nodes, solidifying the boundary information together with the energy debt ledger to form the target pump removal energy island.

[0088] The residual and isolated states of energy nodes within the branch are read from the energy island removed from the target water pump, and then organized into an energy verification object chain according to the order in which energy enters the branch.

[0089] The specific process includes reading the residual state and isolation state of each energy node in the branch from the energy island of the target water pump dismantling, and combining the connection position relationship recorded in the boundary of the target water pump dismantling to determine the order in which energy enters the target water pump branch from the external connection position and is transmitted to the target water pump body. Energy nodes that still have residual states are placed in the pre-verification position, and energy nodes whose isolation state has been confirmed but still need to be continuously observed are placed in the subsequent verification position. At the same time, the isolation action and detection feedback corresponding to each energy node are retained and organized into an energy verification object chain.

[0090] Based on the energy verification object chain, the changes in node residual quantity and the effectiveness of isolation actions are continuously compared, and the node states that meet the verification rules are aggregated into an energy barrier state table.

[0091] The specific process includes: when continuously comparing energy verification object chains, reading the residual amount records and isolation action feedback of energy nodes one by one at adjacent verification times according to the order in which energy enters the target water pump branch; comparing the residual amount changes at the previous verification time with the current verification time; and simultaneously checking whether the isolation action feedback is consistent with the isolation action bound in the energy debt ledger; when the residual amount of an energy node continues to decrease and meets the dismantling safety requirements, and the isolation action feedback is in the confirmation state, the energy node is marked as having passed the verification; when the residual amount of an energy node increases, the isolation feedback is missing, or the isolation feedback is inconsistent, the energy node is marked as having to be verified; and then the verified status and the corresponding detection feedback are aggregated into an energy barrier status table according to the verification order.

[0092] The energy convergence of each energy node in the target water pump branch is calculated using the energy barrier status table, and the energy nodes whose energy convergence has not reached the preset convergence threshold are bound to the corresponding isolation action to obtain the list of safety constraints to be removed.

[0093] The specific process includes: when calculating energy convergence using the energy barrier status table, reading the current value of residual quantity, safety residual threshold, isolation confirmation status, boundary re-input risk quantity, and allowable re-input risk threshold of energy nodes within the target water pump branch according to the verification order; comparing the current value of residual quantity with the safety residual threshold; verifying the isolation confirmation status with the isolation actions bound in the energy debt ledger; comparing the boundary re-input risk quantity with the allowable re-input risk threshold; and then uniformly substituting the comparison content into the energy convergence calculation process; when the energy convergence reaches the convergence threshold, the energy node enters the range of removable constraints; when the energy convergence does not reach the preset convergence threshold, the energy node is reverse-bound to the corresponding power-off isolation, valve isolation, pressure relief and venting, and control confirmation actions to obtain a list of removable safety constraints.

[0094] It should be noted that the convergence threshold is the lower limit of energy convergence for determining whether an energy node has reached a state where it can be removed. It is usually determined by maintenance safety procedures, the closed state of the target water pump removal energy island, historical safe removal records, and on-site risk level. An exemplary value range can be set to 0.80 to 0.95.

[0095] The expression for calculating energy convergence is: ; in, Indicates the energy convergence of the target pump branch; Indicates the number of energy nodes participating in the verification; Indicates the sequence number of the energy node; Indicates the first The residual amount of each energy node; Indicates the first The safety residual threshold for each energy node; Represents a very small positive number; Indicates the first The isolation confirmation status of each energy node is set to 1 when the isolation confirmation is valid and 0 when the isolation confirmation is invalid. Indicates the first The permissible re-input risk threshold for each energy node.

[0096] It should be noted that, This is the maximum residual energy limit that can be retained before mechanical dismantling of an energy node. It is usually determined by the equipment's rated parameters, maintenance safety procedures, sensor detection accuracy, and historical safe dismantling records. An example value can be set to 1% to 5% of the upper limit of the residual energy during normal operation of the corresponding energy node. It is the upper limit of risk of external energy re-entering the boundary that can be withstood before dismantling. It is usually determined by the target water pump dismantling boundary, the reliability of isolation action, on-site maintenance procedures and historical abnormal backflow records. An exemplary value can be set to 1% to 10% of the maximum allowable residual risk of the corresponding energy node.

[0097] Based on the list of safety constraints for dismantling, and combined with the boundary information of the target water pump dismantling energy island, the constraint execution conditions and release conditions are configured to form a set of safety constraints for dismantling.

[0098] The specific process includes: reading each energy node that does not meet the dismantling access conditions and its corresponding isolation action according to the list of dismantling safety constraints; confirming the constraint position and boundary association position by combining the boundary information of the target water pump dismantling energy island; the constraint execution condition is triggered when the residual energy of the energy node does not decrease to the dismantling safety requirements, the isolation action is not confirmed to be effective, the detection feedback is missing, and the risk of re-input from outside the boundary is not eliminated; the constraint release condition is triggered when the residual energy of the energy node meets the dismantling safety requirements, the isolation action feedback is confirmed, the detection feedback is continuously effective, and the risk of re-input from outside the boundary is eliminated; and then organizing the constraint position, constraint execution condition, release condition, and corresponding isolation action into a set of dismantling safety constraints.

[0099] S5. Based on the dismantling safety constraint set and single pump replacement simulation records, the operation paths of interruption due to excessive fluctuation and unconverged residual energy are screened out, and the shutdown isolation and reinstallation grid connection sequence is optimized to generate the optimal replacement operation plan.

[0100] The candidate operation paths and corresponding state trajectories are extracted from the single pump replacement simulation records, and each candidate operation path is matched with the admission conditions in the dismantling safety constraint set to obtain the replacement path map to be screened.

[0101] The specific process includes: reading the action nodes and their sequence of actions from the single-pump replacement simulation record, including the target pump's frequency reduction and withdrawal, compensation takeover, outlet valve slow closure, electrical isolation, pressure relief and venting, new pump connection, and reinstallation and grid connection; combining these action nodes into candidate operation paths based on their continuous connection; simultaneously extracting the corresponding pressure and flow rate changes, remaining pump load changes, control holding status, residual energy changes, and action feedback status for each candidate operation path to form the corresponding state trajectory; matching each action node with the dismantling path admission conditions in the dismantling safety constraint set; and writing the action nodes and path connection relationships with constraint judgment criteria into the same screening framework to obtain the replacement path diagram to be screened.

[0102] It should be noted that the dismantling path admission criteria are safety judgment conditions that candidate operation paths must meet before entering the screening process. These criteria are obtained by mapping the execution conditions and release conditions in the dismantling safety constraint set to each action node. The constraint judgment basis is the state trajectory and feedback record required to determine whether the action node meets the admission criteria. This data comes from the correspondence and matching between the single pump replacement simulation record and the dismantling safety constraint set. The replacement path diagram to be screened is used to represent the matching relationship between each candidate operation path and its corresponding state trajectory and the dismantling path admission criteria. This provides a basis for deleting unsafe paths and generating the admission replacement path diagram.

[0103] Based on the selection of the replacement path graph, operation paths with excessive fluctuations, state interruptions, and unconverged residual energy are identified, and path nodes and dependent paths that do not meet the dismantling safety constraint set are deleted to obtain the admission replacement path graph.

[0104] The specific process includes: reading the action nodes and corresponding state trajectories of each candidate operation path based on the candidate replacement path diagram; comparing the pressure and flow changes with the fluctuation limits in the dismantling safety constraint set; comparing the control holding state with the control continuity requirements; and comparing the residual energy changes with the energy convergence requirements. When a certain action node has a situation of exceeding the fluctuation limit, state interruption, or residual energy not converging, the corresponding action node is marked as not satisfying the dismantling safety constraint set. The affected subsequent dependent paths are located along the action sequence relationship. Then, the path nodes and dependent paths that do not satisfy the dismantling safety constraint set are deleted, and the path relationships that can continuously pass the constraint check are retained to obtain the admission replacement path diagram.

[0105] It should be noted that the fluctuation limit is jointly determined by the allowable pressure and flow rate variation range in the disturbance peak isolation window and the system operating boundary in the pre-dismantling twin operating baseline; the energy convergence requirement is jointly determined by the energy debt ledger in the target pump dismantling energy island and the release conditions in the dismantling safety constraint set; the control continuity requirement is jointly determined by the control hold status in the control state mirror package before power failure and the control hold feedback in the dismantling exercise trajectory.

[0106] Based on the access replacement path diagram and the disturbance peak isolation window, the execution sequence of shutdown isolation action and reinstallation grid connection action is rearranged, and the target water pump exit process and system recovery process are connected to obtain a set of candidate replacement paths.

[0107] The specific process includes reading the path nodes, action sequence, and dependencies that have passed the constraint verification based on the access replacement path diagram, and combining the action start time, duration, and allowable offset range recorded in the disturbance peak isolation window, rearranging the actions of target pump frequency reduction and withdrawal, remaining pump compensation and takeover, outlet valve slow closing, electrical isolation, pressure relief and venting, new pump connection, low-frequency trial operation, and reinstallation and grid connection; during the rearrangement, load transfer-related actions are arranged in the period of lower disturbance, pressure relief and venting are arranged after the residual energy meets the energy convergence requirements, and new pump connection and reinstallation and grid connection are arranged after the control maintains a stable state, forming a set of candidate replacement paths that connect the target pump withdrawal process and the operation recovery process.

[0108] Based on the candidate replacement path set, the path is optimized according to the principles of operational stability, control continuity, and energy convergence. The execution order, feedback conditions, and backoff relationships in the optimized path are then solidified into the optimal replacement operation plan.

[0109] The specific process includes, when optimizing the candidate replacement path set, reading the action nodes, execution time, pressure and flow rate changes, remaining pump load changes, control hold status, residual energy status, and abnormal backoff connection relationships of each candidate replacement path, and using whether the pressure and flow rate changes are within the fluctuation limits as the criteria for judging operational stability, whether the control hold status meets the control continuity requirements as the criteria for judging control continuity, and whether the residual energy status meets the energy convergence requirements as the criteria for judging energy convergence; prioritizing candidate replacement paths that simultaneously meet the above criteria and have short operation connection times, and solidifying the target pump shutdown isolation sequence, new pump connection sequence, field feedback conditions, and backoff relationships into the optimal replacement operation plan.

[0110] In summary, this invention achieves the following: First, it uses a virtual probe driven by a mirrored control state package before power failure to drive a shadow control loop. Second, it inverts the propagation process of pressure and flow disturbances caused by the target pump's shutdown. This allows for early identification of disturbance responses caused by frequency reduction, compensation, and slow shutdown actions before actual shutdown, improving the stability of single pump shutdown and the reliability of pre-replacement simulation. Third, it constructs an energy island for target pump removal through single pump replacement simulation records and performs unified convergence verification of residual risks in the target pump branch. This enables a comprehensive assessment of the degree of residual energy release and external re-input risks within the target pump removal boundary, improving on-site removal safety and the accuracy of the optimal replacement operation plan.

[0111] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention, and all such modifications or substitutions should be covered within the scope of the claims of the present invention.

Claims

1. A simulation method for non-destructive removal and replacement of a single water pump in a multi-pump parallel system, characterized in that, include: Real-time data from multiple pumps in parallel is collected and digital twin synchronous calibration is performed to obtain the pre-dismantling twin operating benchmark. The control logic and dynamic adjustment status of the target water pump before power failure are extracted from the pre-dismantling twin operating benchmark, and the control holding information before shutdown is solidified into a continuously callable pre-power failure control status image package. Virtual probing is performed by driving the shadow control loop through the control state mirror package before power failure, and the propagation process of pressure and flow disturbance caused by the exit of the target water pump is inverted to form a disturbance peak isolation window; A single pump removal and replacement drill was conducted using a disturbance-induced peak isolation window-driven digital twin model, and the load bearing stability and control state continuity were verified simultaneously to form a single pump removal and replacement simulation record. Based on the simulation records of single pump replacement, an energy island for target pump removal is constructed, and the residual energy nodes in the target branch are checked for barrierization. Nodes that do not meet the removal access conditions are converted into executable constraints to form a set of removal safety constraints. Based on the dismantling safety constraint set and single pump replacement simulation records, operation paths that are interrupted by fluctuation exceeding limits and those with unconverged residual energy are screened out, and the shutdown isolation and reinstallation grid connection sequence is optimized to generate the optimal replacement operation plan.

2. The simulation method for non-destructive removal and replacement of a single water pump in a multi-pump parallel system as described in claim 1, characterized in that, The real-time data of the parallel multi-pump system includes hydraulic operation data, pump operation data, valve status data, electrical operation data, control circuit data, and safety maintenance data.

3. The simulation method for non-destructive removal and replacement of a single water pump in a multi-pump parallel system as described in claim 1, characterized in that, The specific steps for obtaining the pre-disassembly twin operating baseline are as follows: Collect real-time data from multiple pumps operating in parallel, configure the collection time and device identifier for the real-time data of multiple pumps operating in parallel, and obtain the original operating data frame; The original running data frames are time-aligned and anomaly-removed, and topology mapping is performed according to pipeline connection relationships, electrical power supply relationships, and control signal relationships to aggregate the hydraulic state and electrical control state at the same time into topology synchronization state data. Based on the topology synchronization state data, the pressure and flow response, electrical response, and control hold-up states in the twin operating space are calibrated, and the system operating boundary before the target water pump is removed is locked to obtain the twin operating benchmark before removal.

4. The simulation method for non-destructive removal and replacement of a single water pump in a multi-pump parallel system as described in claim 3, characterized in that, The steps for extracting the target water pump's control logic and dynamic adjustment status before power failure from the pre-dismantling twin operating baseline, and solidifying the pre-shutdown control holding information into a continuously retrievable pre-power failure control status image package, are as follows: Based on the pre-dismantling twin operation benchmark, locate the target water pump power failure action related objects, and extract the interlock judgment and adjustment trigger relationship from the control points corresponding to each related object according to the correlation strength to form the control logic chain before power failure; Based on the control logic chain before power failure, read the continuous adjustment feedback within the time window before shutdown, and identify the control state that needs to be maintained according to the direction and magnitude of change, thus forming the control holding information before shutdown; The control state of the target water pump before power failure is encapsulated using the pre-shutdown control hold information, and a verification identifier and playback sequence are configured to form a mirror package of the control state before power failure.

5. The simulation method for non-destructive removal and replacement of a single water pump in a multi-pump parallel system as described in claim 4, characterized in that, The virtual probing is performed by driving the shadow control loop through the control state mirror packet before power failure. The specific steps are as follows: Perform integrity verification on the control state image package before power failure, load the control logic and hold state according to the playback order, and generate a virtual execution sequence; A dual-channel shadow control loop is established based on the virtual execution sequence. The control flow that runs according to the control logic playback order in the virtual execution sequence is used as the reference playback flow, and the control flow that adds virtual probing actions according to the same playback order is used as the virtual probing flow. The pressure boundary, flow boundary, and control hold boundary of the virtual test flow are verified around the dual-channel shadow control loop, and a safe virtual test action set is obtained by screening.

6. The simulation method for non-destructive removal and replacement of a single water pump in a multi-pump parallel system as described in claim 5, characterized in that, The specific steps for forming the disturbance staggered isolation window are as follows: Within the set of safe virtual probing actions, frequency reduction probing, compensation probing, and slow closing probing are performed sequentially, and the differences in disturbance response relative to the reference playback stream are recorded to generate a probing response event stack. Extract the pressure response differences and flow response differences corresponding to different probing actions from the probing response event stack, and split them into perturbation atom libraries according to the source of the actions; By comparing the arrival time and decay characteristics of each perturbation atom at different measurement points using the perturbation atom library, the perturbation propagation path diagram of the target water pump exiting the perturbation is derived. Based on the perturbation propagation path diagram, the response overlap and attenuation buffer relationship of the perturbation atoms at the same measurement point are determined, and a phase conflict diagram is formed; The disturbance isolation degree of each candidate peak-shifting window is calculated based on the phase conflict diagram, and the candidate peak-shifting window with the highest disturbance isolation degree that satisfies the pressure boundary and flow boundary is selected as the disturbance peak-shifting isolation window.

7. The simulation method for non-destructive removal and replacement of a single water pump in a multi-pump parallel system as described in claim 1, characterized in that, The specific steps for creating a single pump replacement simulation record are as follows: Extract the execution sequence of each key action from the disturbance peak isolation window, and connect the shutdown isolation and reinstallation grid connection process into the same time sequence chain to generate a peak-shaving exercise instruction table; Based on the peak-shifting drill instruction list, the pre-dismantling twin operation baseline and the pre-power-out control status image package are invoked, and a rollback-capable transaction drill space is constructed. Using a rollback-capable transaction simulation space to drive a digital twin model, the process of target pump shutdown and new pump connection is simulated according to the off-peak simulation instruction table, forming a replacement simulation trajectory that records load changes and control hold status. The stability of the remaining water pump load and the continuity of the control state are simultaneously verified based on the replacement exercise trajectory, and the replacement exercise trajectory that meets the simulation access criteria is solidified into a single pump replacement simulation record.

8. The simulation method for non-destructive removal and replacement of a single water pump in a multi-pump parallel system as described in claim 1 or 7, characterized in that, The specific steps for constructing the target water pump removal energy island based on single pump replacement simulation records are as follows: Extract the action sequence and state trajectory of the target water pump from the single pump replacement simulation record to the mechanical dismantling, and delineate the target water pump dismantling boundary; Based on the target water pump removal boundary, trace the connection points within the target water pump branch that may still retain energy and re-input energy segment by segment, and bind the residual state of each connection point with the corresponding isolation action into a traceable multi-energy domain node graph; The state changes of each energy node in the simulation record are read around the multi-energy domain node graph, and the residual energy release process is recorded as an energy debt ledger; The energy debt ledger is used to uniformly determine the degree of residual energy release within the target water pump removal boundary and the risk of re-input outside the boundary, and energy nodes that have not met the removal safety requirements are marked to obtain the closed state of the energy island; Based on the closed state of the energy island, the isolation actions and detection feedback corresponding to the unconverged energy nodes are traced back, and the boundary of the target water pump removal is supplemented and corrected. The boundary information and energy debt ledger are then solidified to form the target water pump removal energy island.

9. The simulation method for non-destructive removal and replacement of a single water pump in a multi-pump parallel system as described in claim 1, characterized in that, The specific steps for forming the demolition safety constraint set are as follows: The residual and isolated states of the energy nodes in the branch are read from the energy island removed from the target water pump, and organized into an energy verification object chain according to the order in which energy enters the branch. Based on the energy verification object chain, the changes in node residual quantity and the effectiveness of isolation actions are continuously compared, and the node states that meet the verification rules are aggregated into an energy barrier state table; The energy convergence of each energy node in the target water pump branch is calculated using the energy barrier status table, and the energy nodes whose energy convergence has not reached the preset convergence threshold are bound to the corresponding isolation action to obtain the list of safety constraints to be removed. Based on the list of safety constraints for dismantling, and combined with the boundary information of the target water pump dismantling energy island, the constraint execution conditions and release conditions are configured to form a set of safety constraints for dismantling.

10. The simulation method for non-destructive removal and replacement of a single water pump in a multi-pump parallel system as described in claim 1, characterized in that, The specific steps for generating the optimal replacement job plan are as follows: Extract each candidate operation path and its corresponding state trajectory from the single pump replacement simulation record, and match each candidate operation path with the admission conditions in the dismantling safety constraint set to obtain the replacement path map to be screened. Based on the selection of the replacement path graph, operation paths with excessive fluctuations, state interruptions, and unconverged residual energy are identified, and path nodes and dependent paths that do not meet the removal safety constraint set are deleted to obtain the admission replacement path graph. Based on the access replacement path diagram and the disturbance peak isolation window, the execution sequence of shutdown isolation action and reinstallation grid connection action is rearranged, and the target water pump exit process and system recovery process are connected to obtain a set of candidate replacement paths; Based on the candidate replacement path set, the path is optimized according to the principles of operational stability, control continuity, and energy convergence. The execution order, feedback conditions, and backoff relationships in the optimized path are then solidified into the optimal replacement operation plan.

Images