Pipe control system simulation platform and method based on real environment isomorphic simulation

Abstract

This invention relates to the field of industrial control technology and discloses a simulation platform and method for a pipeline control system based on isomorphic simulation of a real environment. The method includes: acquiring model parameter data, control engineering data, and point configuration data; establishing a pipeline physical model and a virtual programmable logic controller; establishing a point mapping table based on the point configuration data and generating point semantic fingerprints for the mapped points; performing collaborative simulation scheduling, writing the state variables of the pipeline physical model into the input image area according to the point mapping table, while keeping the pump unit control variables and valve control variables corresponding to the output image area unchanged; dividing the process into multiple sub-cycles and updating the pipeline physical model according to the sub-cycles; and performing shadow consistency verification, extracting the real point semantic fingerprint and comparing it with the point semantic fingerprint, outputting a zero-modification deployment package when consistency is satisfied. This application achieves verifiable consistency with zero-modification deployment, reducing deployment risk and shortening the joint debugging cycle.

Description

Technical Field

[0001] This invention relates to the field of industrial control technology, and more specifically, to a simulation platform and method for pipeline control systems based on isomorphic simulation of real environments. Background Technology

[0002] In pipeline transportation scenarios such as oil and gas, chemical, and water utilities, pipeline networks are typically controlled in a closed loop by pump units, valves, sensor points, and programmable logic controllers (PLCs). Control engineering often operates on a scan cycle basis, achieving timing consistency between sampling, logical operations, and execution output through input and output image areas. As pipeline network scales and operating conditions fluctuate, control logic iteration, interlocking and protection strategy updates, and sensor point expansion and migration have become commonplace. Relying solely on field trials or static simulations can easily expose hidden problems such as timing inconsistencies and sensor point semantic inconsistencies after deployment, leading to false interlocks, malfunctions, or control deviations. Therefore, a simulation method is needed that isomorphic to the real environment at the control cycle and sensor point semantic levels. This method should enable repeatable and traceable verification without altering the on-site engineering, and perform consistency gating before deployment to reduce deployment risks.

[0003] In existing technologies, for example, patent document CN113189930A proposes to replace the field system with offline simulation testing to verify the correctness of configuration and program (e.g., to build an offline test environment with simulation controller and upper-level monitoring software). However, its core is still biased towards offline testing that can run the function. It lacks sub-cycle update constraints for the "deterministic collaborative scheduling of scanning cycle boundaries" in the control field, which is oriented towards the physical process of the pipeline. It also lacks unified fingerprint verification and shadow consistency gating for the mapping points in terms of semantics such as data type, scaling relationship, quality bit rules, and refresh cycle. Therefore, when the point is expanded, the protocol link changes, or the refresh cycle drifts, there may still be control deviations caused by semantic inconsistencies and timing inconsistencies after offline verification is passed but after field deployment.

[0004] Therefore, it is necessary to design a simulation platform and method for pipeline control systems based on isomorphic simulation of real environment to solve the problems existing in the current technology. Summary of the Invention

[0005] In view of this, the present invention proposes a simulation platform and method for pipeline control systems based on isomorphic simulation of real environment, which aims to solve the problem in the current technology that the simulation passes but the online deployment is deviated, or there are false interlocks or false actions due to the inconsistency between scanning cycle and point semantics between offline simulation and field deployment of pipeline control systems.

[0006] This invention proposes a simulation method for pipeline control systems based on isomorphic simulation of real-world environments, comprising: Acquire model parameter data, control engineering data, and point configuration data. The control engineering data includes the scanning cycle and the update sequence of the input and output image areas. A pipeline physical model is established based on the model parameter data, and a virtual programmable logic controller is established based on the control engineering data. The virtual programmable logic controller latches the input image area according to the scan cycle and executes control logic to obtain the output image area. A point mapping table is established based on the point configuration data, and a point semantic fingerprint is generated for the mapped points in the point mapping table. The point semantic fingerprint is generated by the point semantic description of data type, scaling relationship, quality bit rule and refresh cycle. Cooperative simulation scheduling is performed using the scanning cycle as a unified time reference. At the boundary of each scanning cycle, the state variables of the pipeline physical model are written into the input image area according to the point mapping table, driving the virtual programmable logic controller to update the output image area. The pump unit control variables and valve control variables corresponding to the output image area remain unchanged within the scanning cycle. At the same time, the scanning cycle is divided into multiple sub-cycles and the pipeline physical model is updated according to the sub-cycles. Before zero-modification deployment, a shadow consistency check is performed. The real location semantic fingerprint of the real environment is extracted, and the real location semantic fingerprint is compared with the location semantic fingerprint. The scanning period and the refresh period are checked. If the consistency is satisfied, the zero-modification deployment package is output; otherwise, the generation of the zero-modification deployment package is blocked.

[0007] Furthermore, acquiring the model parameter data, the control engineering data, and the point configuration data includes: reading the pipeline asset configuration to obtain the pipe segment connection relationship, pipe segment geometric parameters, medium physical property parameters, and pump unit characteristic curves and valve characteristic curves to form the model parameter data; parsing the actual programmable logic controller engineering configuration to obtain the scan cycle and the update sequence of the input image area and the output image area, and determining the scan cycle by comparing it with the engineering configuration through the read-only diagnostic interface to form the control engineering data; and parsing the actual programmable logic controller point table and monitoring point table to obtain the data type, scaling relationship, quality bit rules, and refresh cycle to form the point configuration data.

[0008] Furthermore, establishing the physical model of the pipeline includes: The pipeline topology is divided into multiple computing units, and pressure, flow rate, and temperature are set as state variables for each computing unit. The state variables are updated according to the sub-cycle in each scan cycle, and the pump unit control variables and valve control variables in the output image area are read as boundary constraints at the boundary of the scan cycle and remain unchanged in the scan cycle. When establishing the virtual programmable logic controller, the following steps are included: latching the input image area at the beginning of each scan cycle and holding it until the end of the scan cycle, submitting the output image area at the end of the scan cycle, and updating the timer and counter according to the scan cycle.

[0009] Furthermore, when establishing a point mapping table based on the point configuration data, the process includes: The state variables of the pipeline physical model are bound to the input points in the point configuration data, and the pump unit control variables and the valve control variables are bound to the output points in the point configuration data. When there are duplicate bindings or missing bindings, the generation of the mapping relationship of the corresponding mapping points is prohibited.

[0010] Furthermore, generating point semantic fingerprints for each mapping point includes: The data type, scaling relationship, quality bit rule, and refresh cycle are extracted from the mapped points, and normalized and encoded in a fixed field order to obtain the point semantic description. The point semantic description is then summarized to generate the point semantic fingerprint.

[0011] Furthermore, when performing the collaborative simulation scheduling with the scan cycle as a unified time reference, the process includes: writing the state variables of the pipeline physical model into the input image area at the boundary of each scan cycle; driving the virtual programmable logic controller to complete one scan cycle and submitting the output image area; keeping the pump unit control variable and the valve control variable unchanged within the scan cycle, and updating the pipeline physical model according to the sub-cycle with the pump unit control variable and the valve control variable as boundary constraints.

[0012] Furthermore, before writing the state quantity into the input image area, the method includes: applying a communication delay and communication jitter to the state quantity and generating an arrival time; when the arrival time is later than the boundary of the next scan cycle, marking the corresponding mapped point as an invalid quality bit according to the quality bit rule, and maintaining the latch value of the previous scan cycle.

[0013] Furthermore, when entering the shadow consistency check, it includes: The system reads the real programmable logic controller point table and monitoring point table in read-only mode, generates a real point semantic description and a real point semantic fingerprint according to the same field order as the point semantic description; compares the real point semantic fingerprint with the point semantic fingerprint point by point, and when there is an inconsistency between the mapped point marked as an interlocking point or a protection point by the point configuration data, it blocks the generation of the zero-modification deployment package and outputs a difference list.

[0014] Furthermore, verifying the scan period and the refresh period includes: The runtime scan cycle of the actual programmable logic controller is read through the read-only diagnostic interface and compared with the scan cycle in the control engineering data; the arrival time series of the mapped points are collected, and the median interval between adjacent arrival times is calculated as the actual refresh cycle within an observation window covering no less than three refresh cycles, and compared with the refresh cycle in the point semantic description; when the runtime scan cycle or the actual refresh cycle is inconsistent, the corresponding mapped point is marked as an invalid quality bit according to the quality bit rule, and the generation of the zero-modification deployment package is blocked.

[0015] Compared with existing technologies, the advantages of this invention are as follows: By using the scanning cycle of the control engineering as the time base, the update of the pipeline physical process is strongly coupled with the discrete scanning semantics of the programmable logic controller. At the boundary of each scanning cycle, the state variables of the pipeline physical model are deterministically written into the input image area using a point mapping table. The input image area is latched according to the actual update timing, and the control logic is executed to generate the output image area. At the same time, the control variables of the pump unit and valve are kept unchanged within the scanning cycle, and the pipeline physical model is updated using sub-cycles. Thus, the actual closed-loop timing is reproduced on the simulation side, reducing the impact of timing issues. This solution addresses the risks of control deviations and malfunctions caused by misalignment. It upgrades point alignment from "address alignment" to "semantic alignment," generating a point semantic fingerprint for each mapped point. Before deployment with zero modifications, it extracts the real point semantic fingerprint through shadow consistency verification and compares it item by item. This allows hidden problems such as point expansion, migration, refresh cycle drift, or quality bit processing differences to be repeatedly and traceably discovered and intercepted before deployment, avoiding costly trial and error. It improves online efficiency, shortens the joint debugging cycle, reduces on-site trial operation costs, and enhances the reliability and security of interlocking and protection strategies.

[0016] On the other hand, this application also provides a pipeline control system simulation platform based on isomorphic simulation of a real environment, for applying the above-mentioned pipeline control system simulation method based on isomorphic simulation of a real environment, including: The acquisition unit is configured to acquire model parameter data, control engineering data, and point configuration data. The control engineering data includes the scanning cycle and the update sequence of the input image area and the output image area. The building unit is configured to build a pipeline physical model based on the model parameter data; and to build a virtual programmable logic controller based on the control engineering data. The virtual programmable logic controller latches the input image area according to the scan cycle and executes control logic to obtain the output image area. The processing unit is configured to establish a point mapping table based on the point configuration data, and generate point semantic fingerprints for the mapped points in the point mapping table. The point semantic fingerprints are generated by point semantic descriptions of data type, scaling relationship, quality bit rules and refresh cycle. The update unit is configured to perform collaborative simulation scheduling with the scan cycle as a unified time reference. At the boundary of each scan cycle, the state variables of the pipeline physical model are written into the input image area according to the point mapping table, driving the virtual programmable logic controller to update the output image area. During the scan cycle, the pump unit control variables and valve control variables corresponding to the output image area are kept unchanged. At the same time, the scan cycle is divided into multiple sub-cycles and the pipeline physical model is updated according to the sub-cycles. The judgment unit is configured to enter shadow consistency verification before zero-modification deployment, extract the real point semantic fingerprint of the real environment, compare the real point semantic fingerprint with the point semantic fingerprint and verify the scanning period and the refresh period. If the consistency is satisfied, the zero-modification deployment package is output; otherwise, the generation of the zero-modification deployment package is blocked.

[0017] It is understandable that the above-mentioned pipeline control system simulation platform and method based on isomorphic simulation of real environment have the same beneficial effects, and will not be elaborated here. Attached Figure Description

[0018] Various other advantages and benefits will become apparent to those skilled in the art upon reading the following detailed description of preferred embodiments. The accompanying drawings are for illustrative purposes only and are not intended to limit the invention. Furthermore, the same reference numerals denote the same parts throughout the drawings. In the drawings: Figure 1 A flowchart of a pipeline control system simulation method based on isomorphic simulation of a real environment provided in an embodiment of the present invention; Figure 2 This is a functional block diagram of a pipeline control system simulation platform based on isomorphic simulation of a real environment, provided for embodiments of the present invention. Detailed Implementation

[0019] 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 to enable a more thorough understanding of the present disclosure and to fully convey the scope of the disclosure to those skilled in the art. It should be noted that, unless otherwise specified, embodiments and features in the embodiments of the present invention can be combined with each other. The present invention will now be described in detail with reference to the accompanying drawings and embodiments.

[0020] In some embodiments of this application, see Figure 1 As shown, this application proposes a simulation method for pipeline control systems based on isomorphic simulation of real environments, including: S100: Acquire model parameter data, control engineering data, and point configuration data. The control engineering data includes the scanning cycle and the update sequence of the input and output image areas.

[0021] S200: Establish a pipeline physical model based on model parameter data, and establish a virtual programmable logic controller based on control engineering data. The virtual programmable logic controller latches the input image area according to the scan cycle and executes the control logic to obtain the output image area.

[0022] S300: Establish a point mapping table based on the point configuration data, and generate point semantic fingerprints for the mapped points in the point mapping table. The point semantic fingerprint is generated by the point semantic description of data type, scaling relationship, quality bit rules and refresh cycle.

[0023] S400: Cooperative simulation scheduling is performed with the scan cycle as a unified time reference. At the boundary of each scan cycle, the state variables of the pipeline physical model are written into the input image area according to the point mapping table, which drives the virtual programmable logic controller to update the output image area. The pump unit control variables and valve control variables corresponding to the output image area remain unchanged within the scan cycle. At the same time, the scan cycle is divided into multiple sub-cycles and the pipeline physical model is updated according to the sub-cycles.

[0024] S500: Before zero-modification deployment, enter shadow consistency verification, extract the real point semantic fingerprint of the real environment, compare the real point semantic fingerprint with the point semantic fingerprint and verify the scanning cycle and refresh cycle. If the consistency is satisfied, output the zero-modification deployment package; otherwise, block the generation of the zero-modification deployment package.

[0025] Specifically, in this embodiment, the real environment is the production and operation environment of the pipeline control system, including at least the real programmable logic controller (PLC) and its engineering configuration, the field point table, and the monitoring point table. Access to the real environment in this embodiment is read-only; the simulation side does not write any control variables or configuration data to the real PLC. The scan cycle boundary is the start time of each scan cycle, and the next scan cycle boundary is the start time of the next scan cycle. The monitoring point table is the tag point table in the monitoring system, which, together with the real PLC point table, constitutes the source of point configuration data. The two establish a correspondence through point address identifiers or point labels. The quality bit rule is used to characterize whether the point data is valid in the current scan cycle. When the quality bit rule is set to invalid quality bit, it indicates that the point is not used to update the latch value of the input image area in this scan cycle, and the latch value of the previous scan cycle is used according to the hold strategy.

[0026] Specifically, in step S100, the simulation platform reads pipe segment connection relationships, pipe segment geometric parameters, medium physical property parameters, pump unit characteristic curves, and valve characteristic curves from the pipeline asset configuration to form model parameter data. The pipe segment connection relationships are preferably expressed using a connection relationship table, which includes at least the current pipe segment identifier, upstream node identifier, downstream node identifier, connected equipment type, and connection direction, used to determine the connectivity order of each pipe segment, pump unit, and valve in the pipeline topology. The pipe segment geometric parameters include at least one or more of length, inner diameter, wall thickness, and elevation information. The medium physical property parameters include at least one or more of density, viscosity, compressibility, and specific heat. The pump unit characteristic curves at least include the correspondence between flow rate and head, and preferably also include the correspondence between flow rate and head at different speeds and the correspondence between efficiency, used to determine the boundary effects exerted by the pump unit on adjacent computational units within the scan cycle based on the pump unit control quantities in the output image area. The valve characteristic curve includes at least the relationship between valve opening degree and flow capacity, and preferably also the relationship between flow capacity under different pressure differential conditions, which is used to determine the throttling boundary effect of the valve on adjacent calculation units within the scanning cycle based on the valve control quantity.

[0027] Step S200: The pipeline topology is divided into multiple computational units based on model parameter data, and a pipeline physical model is established using pressure, flow rate, and temperature as state variables of the computational units. In this embodiment, the transmission relationship of state variables between computational units is established through a unit connection table. The unit connection table includes at least the upstream computational unit identifier, downstream computational unit identifier, shared boundary identifier, type of transmitted state variable, transmission direction, and update time order. The simulation platform first determines the adjacency relationship between each computational unit based on the pipe segment connection relationship, and then maps the pump unit, valve, and ordinary pipe segment to the corresponding shared boundary. For adjacent computational units, at least one of the outlet pressure, outlet flow rate, and outlet temperature calculated by one computational unit in the current sub-cycle is used as the inlet boundary quantity of the other computational unit in the next update step. For nodes with branches or merging, the associated computational units are updated sequentially according to the node connectivity relationship, and the transmitted state variables on each shared boundary are refreshed uniformly at the end of each sub-cycle, so that the state variables can be gradually transmitted and evolved under the pipeline topology constraints. Simultaneously, a virtual programmable logic controller is established based on the control engineering data, and its scan execution semantics are configured as follows: at the beginning of each scan cycle, the input image area is latched and held until the end of the scan cycle; the control logic operation is completed in the order of program block execution; and the output image area is submitted all at once at the end of the scan cycle. The timer and counter are cumulatively updated according to the scan cycle to ensure that the update semantics of the simulation side and the real control cycle are consistent.

[0028] Step S300: Based on the point configuration data, a point mapping table is established. The state variables of the pipeline physical model are bound to the input points, and the pump unit control variables and valve control variables are bound to the output points. The mapping relationship is validated; if duplicate bindings or missing bindings occur, no corresponding mapping relationship is generated, and a difference list is output. Subsequently, for each mapped point in the point mapping table, the data type, scaling relationship, quality position rules, and refresh cycle are extracted. Normalized encoding is performed according to a fixed field order to obtain the point semantic description. Normalized encoding includes field value standardization, default value filling, and unit consistency processing, ensuring consistent encoding results for the same semantic meaning across different source point tables. Based on this, a deterministic summary is generated from the point semantic description to obtain the point semantic fingerprint, which serves as the unique representation for subsequent consistency comparisons.

[0029] Step S400: Cooperative simulation scheduling is performed using the scan cycle as a unified time base. At the boundary of each scan cycle, the state variables of the pipeline physical model at the end of the previous scan cycle are written into the input image area according to the point mapping table. The virtual programmable logic controller is triggered to complete one scan cycle and submit the output image area. During this scan cycle, the pump unit control variables and valve control variables corresponding to the output image area remain unchanged, and the pump unit control variables and valve control variables are used as boundary constraints. Among them, the pump unit control variables include at least one or more of the following: pump start / stop status, pump speed setpoint, frequency converter setpoint, and outlet pressure setpoint. The valve control variables include at least one or more of the following: valve open / close status, valve opening setpoint, and valve position command. In this embodiment, the pump speed setpoint and valve opening setpoint are selected as boundary constraints. For example, the pump unit speed setpoint is 2400 r / min, and the valve opening setpoint is 60%, which remain unchanged during the 100ms scan cycle so that the state variables of adjacent computing units are updated according to fixed boundary conditions during this scan cycle. To simulate the field communication and acquisition link, this embodiment applies communication delay and jitter to the state variables before writing them into the input image area and generates arrival times. When the arrival time is later than the boundary of the next scan cycle, the corresponding mapped point is marked as an invalid quality bit according to the quality bit rule, and the latch value of the previous scan cycle is maintained, thereby reproducing the impact of point delay, packet loss or refresh drift on the control closed loop on the simulation side.

[0030] Step S500: Before zero-modification deployment, proceed with shadow consistency verification: Read the real programmable logic controller (PLC) point table and monitoring point table in read-only mode. For each mapped point in the point mapping table, generate a real point semantic description and a real point semantic fingerprint according to the same field order as the point semantic description. Compare the real point semantic fingerprint with the point semantic fingerprint point by point. When there is inconsistency between the mapped points marked as interlocked or protected points by the point configuration data, block the generation of the zero-modification deployment package and output a difference list. The difference list includes at least the mapped point identifier and the inconsistency field category to locate the source of inconsistency in data type, scaling relationship, quality bit rule, or refresh cycle. At the same time, read the runtime scan cycle of the real PLC through the read-only diagnostic interface and compare it with the scan cycle in the control engineering data. Collect the arrival time series of the mapped points, calculate the median interval between adjacent arrival times within an observation window covering no less than 3 refresh cycles as the real refresh cycle to reduce the impact of a few abnormal late arrivals or jitter on the cycle estimation, and compare it with the refresh cycle in the point semantic description. When the runtime scan cycle or the actual refresh cycle is inconsistent, the corresponding mapped point is marked as an invalid quality bit according to the quality bit rule, and the generation of the zero-modification deployment package is blocked. When the consistency gating condition is met, the zero-modification deployment package is output. The zero-modification deployment package includes at least a point mapping table, point semantic fingerprint, scan cycle consistency verification result and refresh cycle consistency verification result, which is used to complete the deployment or joint debugging without changing the on-site engineering configuration.

[0031] Understandably, by using the scan cycle as a unified time reference, the discrete scan semantics of "input image area latching - control logic execution - output image area submission" are deterministically coupled with the continuous sub-cycle updates of the pipeline physical model. This ensures that the closed-loop response on the simulation side is consistent with the real control system in terms of time tick, thereby reducing the risk of timing deviations introduced by relying solely on static simulation or functional integration testing. Simultaneously, point-to-point consistency is elevated from the address layer to the semantic layer of data type, scaling relationship, quality bit rules, and refresh cycle through point-to-point semantic fingerprinting. Furthermore, shadow consistency verification uses semantic differences and tick drift as deployment gating conditions, enabling traceable automatic discovery and blocking before deployment with zero modifications. This reduces implicit control deviations and false interlocking problems caused by point-to-point expansion, migration, or refresh drift, improving online reliability and shortening the integration testing cycle.

[0032] In some embodiments of this application, the acquisition of model parameter data, control engineering data, and point configuration data includes: reading the pipeline asset configuration to obtain pipe segment connection relationships, pipe segment geometric parameters, medium physical property parameters, and pump unit characteristic curves and valve characteristic curves to form model parameter data; parsing the actual programmable logic controller (PLC) engineering configuration to obtain the scan cycle and the update timing of the input and output image areas, and determining the scan cycle by comparing it with the engineering configuration through a read-only diagnostic interface to form control engineering data; and parsing the actual PLC point table and monitoring point table to obtain data types, scaling relationships, quality bit rules, and refresh cycles to form point configuration data.

[0033] In some embodiments of this application, establishing a pipeline physical model includes: dividing the pipeline topology into multiple computational units, and setting pressure, flow rate, and temperature as state variables for each computational unit. The state variables are updated periodically within each scan cycle, and the pump unit control variables and valve control variables in the output image area are read as boundary constraints at the scan cycle boundary and remain unchanged within that scan cycle. Establishing a virtual programmable logic controller includes: latching the input image area at the beginning of each scan cycle and holding it until the end of the scan cycle; submitting the output image area at the end of the scan cycle; and updating the timer and counter periodically.

[0034] Specifically, when establishing the pipeline physical model, the simulation platform constructs the pipeline topology based on the pipe segment connection relationships in the pipeline asset configuration and divides the pipeline topology into multiple computational units. Each computational unit corresponds to a pipe segment or a local subnet formed by the combination of adjacent pipe segments, used to carry the discretized representation of state variables. To ensure that the effect of control variables on the physical process can be reflected in a closed loop in the simulation, this embodiment sets at least pressure, flow rate, and temperature as state variables for each computational unit and establishes the transmission relationship between state variables between computational units, enabling state variables to evolve over time under topological constraints. Sub-cycles are the subdivided time steps of the scan cycle, used to update physical state variables with finer granularity within the same scan cycle. Within each scan cycle, the simulation platform iteratively updates the state variables of each computational unit according to the sub-cycle sequence, and reads the pump unit control variables and valve control variables from the output image area at the boundary of each scan cycle as boundary constraints. The boundary constraints remain unchanged within the scan cycle, ensuring that the continuous update process of the pipeline physical model is consistent with the discrete scan cycle of the control system, avoiding unrealistic numerical oscillations or timing distortions introduced by frequent changes in control variables within the same scan cycle.

[0035] When establishing a virtual programmable logic controller (PLC), the simulation platform configures its scan execution semantics based on control engineering data to ensure consistency with the input / output update mechanism of the real PLC. Specifically, at the beginning of each scan cycle, the virtual PLC latches the input image region and holds the latched value until the end of the scan cycle, thus ensuring that the input quantities used for control logic operations within the current scan cycle are the same set of defined sampled values. At the end of the scan cycle, the virtual PLC performs a one-time commit of the output results calculated by the control logic, updating the output image region to reproduce the commit timing of the real control system. Simultaneously, the virtual PLC cumulatively updates timers and counters according to the scan cycle, ensuring that the interlocking, protection, and state machine logic based on timing and counting has the same time progression and triggering conditions on the simulation side as on the real side. This guarantees that the dynamic behavior of the control logic can be stably reproduced and used for consistency verification.

[0036] In some embodiments of this application, when establishing a point mapping table based on point configuration data, the process includes: binding the state variables of the pipeline physical model to the input points in the point configuration data, and binding the pump unit control variables and valve control variables to the output points in the point configuration data. When there are duplicate bindings or missing bindings, the generation of mapping relationships for the corresponding mapping points is prohibited.

[0037] In some embodiments of this application, generating point semantic fingerprints for each mapped point includes: extracting data type, scaling relationship, quality bit rules and refresh cycle from the mapped point, performing normalized encoding according to a fixed field order to obtain a point semantic description, and generating a summary of the point semantic description to obtain a point semantic fingerprint.

[0038] Specifically, when establishing a point mapping table based on the point configuration data, the simulation platform uses the input and output points recorded in the point configuration data as constraint boundaries to complete the bidirectional binding of physical-side state variables and control-side points. First, the state variables of the pipeline physical model are bound to input points, used to write the state variables into the input image area during collaborative simulation scheduling, thereby providing the virtual programmable logic controller with sampling input consistent with the real points. Second, the pump unit control variables and valve control variables are bound to output points, used to apply the control variables corresponding to the output points to the boundary constraints of the pipeline physical model after the virtual programmable logic controller submits the output image area. To ensure the determinism and verifiability of the mapping relationship, this embodiment performs uniqueness and completeness checks on the input and output points respectively when establishing the point mapping table. When the same state variable or control variable is repeatedly bound to multiple non-equivalent points, it is considered a duplicate binding. When a state variable or control variable that should participate in the closed-loop calculation cannot find a corresponding input or output point, it is considered a missing binding. For cases of duplicate binding or missing binding, no corresponding mapping relationship is generated, and the anomaly point identifier, anomaly type, and associated object are written into the difference list.

[0039] When generating semantic fingerprints for each mapped point, the simulation platform extracts fields directly related to semantic consistency from the point configuration data for each mapped point in the point mapping table. These fields include data type, scaling relationship, quality bit rules, and refresh cycle. The data is then standardized and encoded in a fixed field order to generate a semantic description of the point. Standardized encoding includes at least the following: Standardizing data types by representing Boolean, integer, floating-point, analog, and enumerated quantities with uniform field values, ensuring consistent encoding for data types with the same semantics across different point table sources. Providing a unified expression for scaling relationships, which consist of at least the original unit, engineering unit, scale factor, offset, and rounding strategy, characterizing the conversion from original to engineering quantities. When a point lacks an offset or rounding strategy, the corresponding field is filled with a preset default value. Representing quality bit rules in a regularized manner, which consists of at least the quality bit source, valid judgment conditions, invalid handling strategies, and recovery conditions. Invalid handling strategies include at least one of maintaining the previous latch value, setting a default value, and blocking updates. The refresh cycle is uniformly characterized, consisting of at least a cycle value, time unit, data acquisition source, and allowable fluctuation range, used to represent the arrival time of a point in the monitoring or acquisition link. Preferably, the point semantic description is encoded sequentially in the fixed field order of 'data type—scaling relationship—quality bit rule—refresh cycle', where the scaling relationship can be expressed as 'engineering quantity unit is kilopascal, original quantity unit is count value, scale factor is 0.01, offset is 0, and rounding strategy is rounding to the nearest integer'. The quality bit rule can be expressed as 'quality bit source is the quality identifier of the monitoring point table; when the quality bit is invalid, the latched value of the previous scan cycle is retained; when the quality bit is restored to validity, the current value is written'. The refresh cycle can be expressed as '1000 milliseconds, acquisition source is the monitoring link, and allowable fluctuation range is ±100 milliseconds'. Based on this, deterministic digest generation is performed on the point semantic description to obtain the point semantic fingerprint, where deterministic digest generation means outputting a unique and stable fingerprint result under the same point semantic description input conditions, used to achieve point-by-point comparison and difference location in shadow consistency verification. Preferably, the semantic descriptions of the points are first uniformly characterized and concatenated according to a fixed field order, and then a preset summary operation is performed. This ensures that point semantic descriptions with the same field content and field order generate unique and stable point semantic fingerprints. The allowable fluctuation range is preferably determined based on the historical arrival interval statistics of the monitoring link or acquisition link during stable operation. When a refresh cycle tolerance configuration already exists on site, this tolerance configuration can be directly used as the allowable fluctuation range. Mapped points marked as interlocked points or protection points by the point configuration data are preferably marked based on the interlocked variable list, protection variable list, or key point list confirmed by manual review in the actual programmable logic controller engineering configuration.

[0040] Understandably, by establishing a point mapping table with point configuration data as the sole constraint source, and performing blocking checks on duplicate and missing bindings during the generation phase, the data flow and control action paths of the simulation-side input / output image area become deterministic and traceable, avoiding implicit erroneous mappings caused by point expansion, migration, or incomplete configuration. Simultaneously, by constructing point semantic descriptions and generating point semantic fingerprints using data type, scaling relationship, quality bit rules, and refresh cycle, point consistency is elevated from "address matching" to "semantic matching." This enables shadow consistency checks to locate and gate semantic differences and clock drift point by point, significantly reducing the risk of control deviations and false interlocks caused by semantic inconsistencies between offline simulation and field deployment, and improving the reliability and engineering feasibility of zero-modification deployment.

[0041] In some embodiments of this application, when performing collaborative simulation scheduling with the scan cycle as a unified time base, the process includes: writing the state variables of the pipeline physical model into the input image area at the boundary of each scan cycle; driving the virtual programmable logic controller to complete one scan cycle and submitting the output image area; keeping the pump unit control variables and valve control variables unchanged within the scan cycle, and updating the pipeline physical model according to sub-cycles, using the pump unit control variables and valve control variables as boundary constraints.

[0042] In some embodiments of this application, before writing the state quantity into the input image area, the process includes: applying a communication delay and communication jitter to the state quantity and generating an arrival time. When the arrival time is later than the boundary of the next scan cycle, the corresponding mapped point is marked as an invalid quality bit according to the quality bit rule, and the latch value of the previous scan cycle is maintained.

[0043] Specifically, in co-simulation scheduling, this embodiment uses the scan cycle as a unified time base to time-sequence the physical process update and control scan execution, forming a repeatable closed-loop scheduling sequence on the same time axis. The co-simulation scheduler uses the scan cycle boundary as a trigger point. At the beginning of each scan cycle, it acquires the state variables of the pipeline physical model at the end of the previous scan cycle and writes them into the input image area according to the point mapping table, forming the current cycle's sampled input for the virtual programmable logic controller (PPL). Subsequently, the co-simulation scheduler triggers the PPL to complete one scan cycle execution, including latching the input image area, executing control logic, and submitting the output image area at the end of the scan cycle. After the PPL submits the output image area, the co-simulation scheduler reads the pump unit control and valve control variables corresponding to the output image area and keeps them unchanged within the scan cycle, applying them as boundary constraints to the pipeline physical model. Furthermore, to reflect the continuous evolution of the physical process within the same scan cycle, the co-simulation scheduler divides the scan cycle into multiple sub-cycles and updates the state variables of the pipeline physical model sequentially according to the sub-cycles. During each sub-cycle update process, the pump unit control quantity and valve control quantity that remain unchanged within the scan cycle are always used as boundary constraints, thereby ensuring that the control engineering fact that "the control quantity remains unchanged within a scan cycle" is stably reproduced, and the physical model update will not produce unrealistic timing drift due to the control quantity jumping within the cycle.

[0044] Before writing the state variables into the input image area, to reproduce the impact of the real communication link and acquisition link on the data arrival time, this embodiment uses a co-simulation scheduler to apply communication delay and jitter to the state variables and generate arrival times. Communication delay characterizes the transmission time of the state variable from the physical side to the control side, and communication jitter characterizes the fluctuation of transmission time in different cycles. Together, they determine whether the state variable can arrive on time and be latched by the input image area within the current scan cycle. After generating the arrival time, the co-simulation scheduler compares the arrival time with the boundary of the next scan cycle. When the arrival time is later than the boundary of the next scan cycle, it is determined that the state variable has failed to effectively participate in the control input update in this cycle, and the corresponding mapped point is marked as an invalid quality bit according to the quality bit rule. Simultaneously, a hold strategy is implemented for this mapped point, ensuring that the input image area uses the latched value from the previous scan cycle within this scan cycle. Through the above processing, the co-simulation scheduler can explicitly reproduce the impact of point delays, packet loss, or refresh drift on the control logic on the simulation side, and provide traceable quality bit status and timing evidence for subsequent shadow consistency verification.

[0045] In some embodiments of this application, the process of entering shadow consistency verification includes: reading the real programmable logic controller point table and monitoring point table in read-only mode, generating a real point semantic description and a real point semantic fingerprint according to the same field order as the point semantic description. The real point semantic fingerprint is compared point-by-point with the point semantic fingerprint. When there is an inconsistency between the mapped point marked as an interlocking point or protection point by the point configuration data, the generation of the zero-modification deployment package is blocked and a difference list is output. In some embodiments of this application, the process of verifying the scan cycle and refresh cycle includes: The runtime scan cycle of the actual programmable logic controller (PLC) is read through a read-only diagnostic interface and compared with the scan cycle in the control engineering data. Arrival time series of mapped points are collected, and the median interval between adjacent arrival times is calculated as the true refresh cycle within an observation window covering at least three refresh cycles. This median interval is then compared with the refresh cycle in the point's semantic description. When the runtime scan cycle or the true refresh cycle is inconsistent, the corresponding mapped point is marked as an invalid quality bit according to the quality bit rules, and the generation of a zero-modification deployment package is blocked.

[0046] Specifically, during the shadow consistency verification, this embodiment, without modifying the actual environment engineering configuration, accesses the actual programmable logic controller (PLC) point table and monitoring point table in read-only mode, using the point mapping table as the verification object set. For each mapped point in the point mapping table, the simulation platform extracts a set of fields consistent with the point semantic fingerprint from the actual PLC point table and monitoring point table, including at least data type, scaling relationship, quality bit rules, and refresh cycle, and performs normalized encoding according to the same field order as the point semantic description to generate an actual point semantic description. Subsequently, a deterministic digest is generated from the actual point semantic description to obtain the actual point semantic fingerprint. The simulation platform compares the actual point semantic fingerprint with the point semantic fingerprint on the simulation side point by point, and determines the blocking conditions based on the attribute markings of the mapped points in the point configuration data. When there is an inconsistency between the mapped points marked as interlocked points or protection points, the generation of the zero-modification deployment package is directly blocked, and a list of differences is output. The discrepancy list should include at least the identifier of the inconsistent mapping point, the category of the inconsistent field, the source of the inconsistency, and the recommended handling actions. This is to support targeted rectification or remapping on the engineering side and avoid mistakenly bringing high-risk points into zero-modification deployment by using the "only address consistent" approach.

[0047] In verifying the scan cycle and refresh cycle, this embodiment performs read-only verification on the control cycle and acquisition cycle respectively and forms gating criteria. For the scan cycle, the simulation platform reads the runtime scan cycle of the real programmable logic controller through the read-only diagnostic interface and compares it with the scan cycle in the control engineering data to confirm that the unified time base used by the simulation side's collaborative simulation scheduling is consistent with the real control cycle. For the refresh cycle, the simulation platform collects the arrival time series of the mapped points and calculates the median interval of adjacent arrival times as the real refresh cycle within an observation window covering at least three refresh cycles. The purpose of using the median interval instead of the mean interval is to reduce the impact of a small number of abnormal delays or jitter on the cycle estimation, making the refresh cycle verification more in line with the robustness requirements under field link fluctuation conditions. The observation window covers at least three refresh cycles to ensure that at least three adjacent arrival intervals are obtained within the observation window, thereby reducing the impact of a single abnormal delay, instantaneous jitter, or occasional packet loss on the real refresh cycle estimation result. When the number of effective arrivals within the observation window is insufficient to form the above-mentioned number of adjacent arrival intervals, the observation window is extended before performing the real refresh cycle estimation. Subsequently, the simulation platform compares the actual refresh cycle with the refresh cycle in the semantic description of the point point point by point. When the runtime scanning cycle or the actual refresh cycle is inconsistent with the corresponding cycle, the corresponding mapped point point is marked as an invalid quality bit according to the quality bit rule, and the generation of the zero-modification deployment package is blocked. This clearly incorporates "beat inconsistency" into the deployment gating, avoiding the risk of the interlocking, protection or control logic triggering sequence being inconsistent with the simulation expectation under the condition of beat drift.

[0048] Understandably, shadow consistency verification elevates pre-deployment verification from "functionally operable" to "semantic and clockwise provably consistent." Point-by-point comparison of point semantic fingerprints enables traceable differentiation of key points across data types, scaling relationships, quality bit rules, and refresh cycles. Interlocking or protected points are prioritized for blocking to reduce security risks. Simultaneously, read-only comparison of runtime scan cycles and robust estimation of the actual refresh cycle based on arrival time series integrate control and acquisition cycles into the same gating framework. This proactively detects and blocks implicit timing deviations caused by scan or refresh cycle drift before zero-modification deployment, significantly improving the consistency and reliability of simulation verification results for on-site deployment behavior and reducing the probability of false interlocks and malfunctions after deployment.

[0049] In one specific embodiment, the simulation platform reads pipeline asset configuration to form model parameter data: pipe segment connection relationships, pipe segment geometric parameters, medium physical property parameters, and pump unit characteristic curves and valve characteristic curves. Example data includes: 6 pipe segments, each with a length ranging from 0.8km to 2.6km, and an outer diameter of 0.40m or 0.35m. The medium density is 780kg / m³, and the specific heat at constant pressure is 2250J / kg / Kelvin. The control engineering data is obtained by parsing the actual programmable logic controller (PLC) engineering configuration: the scan cycle is 100ms, the input image area latch timing is the start latch of the scan cycle, and the output image area commit timing is the end commit of the scan cycle. The runtime scan cycle is read as 100ms through the read-only diagnostic interface and confirmed to match the engineering configuration. Point configuration data is formed by parsing the actual PLC point table and monitoring point table: this data includes at least the data type, scaling relationship, quality bit rules, and refresh cycle. For example: the data type of the pressure point is 32-bit floating point, the original quantity unit is a count value, the engineering quantity unit is kilopascals, the scale factor is 0.01, the offset is 0, the rounding strategy is to retain two decimal places, the refresh cycle is 1000ms, the allowable fluctuation range is ±100ms, the quality bit source is the quality identifier of the monitoring point table, when the quality bit is invalid, it retains the latched value of the previous scan cycle, and when the quality bit is restored to validity, it restores the current value and writes it. The data type of the valve opening point is 16-bit integer, the original quantity unit is a count value, the engineering quantity unit is a percentage, the scale factor is 0.1, the offset is 0, the rounding strategy is to round to the nearest integer, the refresh cycle is 500ms, and the allowable fluctuation range is ±50ms. The above fields together form the point semantic description and are used to generate the point semantic fingerprint.

[0050] The simulation platform divides the pipeline topology into multiple computational units and sets pressure, flow rate, and temperature as state variables for each unit. When each scan cycle is divided into multiple sub-cycles, the sub-cycle value is determined according to the following rules: the sub-cycle should be smaller than the scan cycle, preferably an integer division of the scan cycle, to ensure that the sub-cycle boundary remains aligned with the scan cycle boundary. The selection of the sub-cycle is determined in conjunction with the rate of change of the pipeline physical process, the response beat of the pump unit and valve control action, the communication link jitter range, and the simulation computational overhead. When the sub-cycle value is too large, the update of the physical state variables within the same scan cycle is too coarse, making it difficult to reflect the continuous effect of the control variable within that scan cycle, and easily introducing state variable jumps and timing distortions. When the sub-cycle value is too small, although the update accuracy is improved, the computational burden increases significantly and the improvement on the control semantics at the scan cycle level is limited. In this embodiment, the scan cycle is 100ms, and the sub-cycle is 10ms, so that each scan cycle contains 10 sub-cycles. This allows for sufficiently fine-grained updates to the pipeline physical model within a single scan cycle while ensuring stable alignment between the simulation update step size and the actual control beat. Within each scan cycle, the state variables are updated according to sub-cycles. At the scan cycle boundary, the pump unit control variable and valve control variable in the output image area are read as boundary constraints and remain unchanged within that scan cycle. An example boundary constraint is: pump unit control variable is 2400 r / min, valve control variable is 60% opening, and these constraints remain fixed within the 100ms scan cycle. Simultaneously, a virtual programmable logic controller is established: at the beginning of each scan cycle, the input image area is latched and held until the end of the scan cycle; at the end of the scan cycle, the output image area is committed, and the timer and counter are updated according to the scan cycle.

[0051] The simulation platform establishes a point mapping table based on the point configuration data: pressure, flow, and temperature status variables of the pipeline physical model are bound to input points, and pump unit control variables and valve control variables are bound to output points. When duplicate bindings or missing bindings exist, the corresponding mapping relationship is prevented from being generated and a discrepancy list is output. For each mapped point in the point mapping table, the data type, scaling relationship, quality position rules, and refresh cycle are extracted, and standardized and encoded according to a fixed field order to obtain a point semantic description. A summary of the point semantic description is then generated to obtain a point semantic fingerprint, which is used for consistency comparison before deployment.

[0052] The co-simulation scheduler operates on a unified time base of 100ms scan cycles: at the boundary of each scan cycle, the state variables of the pipeline physical model are written into the input image area according to the point mapping table. The virtual programmable logic controller (PLC) is driven to complete one scan cycle and submit the output image area. Within this scan cycle, the pump unit control variables and valve control variables corresponding to the output image area remain unchanged, and the pipeline physical model is updated every 10ms sub-cycle, using the pump unit control variables and valve control variables as boundary constraints. Before writing to the input image area, the scheduler applies communication delay and jitter to the state variables and generates arrival times. For example, the communication delay is 90ms and the communication jitter is 20ms. If a scan cycle boundary is 2.500s, the next scan cycle boundary is 2.600s. When the calculated arrival time of the state variable is 2.605s, it is determined to be later than the next scan cycle boundary. According to the quality bit rule, the corresponding mapped point is marked as an invalid quality bit, and the latch value from the previous scan cycle is maintained.

[0053] The simulation platform enters shadow consistency verification: It reads the real programmable logic controller (PLC) point table and monitoring point table in read-only mode, generates a real point semantic description and a real point semantic fingerprint according to the same field order as the point semantic description. The real point semantic fingerprint is compared point-by-point with the point semantic fingerprint. When there is an inconsistency between the mapped point marked as an interlocking point or protection point by the point configuration data, the generation of the zero-modification deployment package is blocked, and a difference list is output. Simultaneously, the scan cycle and refresh cycle are verified: the runtime scan cycle is read through the read-only diagnostic interface and compared with the scan cycle in the control engineering data. Arrival time series are collected for the mapped points, and the median interval between adjacent arrival times is calculated as the real refresh cycle within an observation window covering at least three refresh cycles. Example: A pressure point is configured with a refresh cycle of 1000ms, an observation window of 3.2s, and acquisition arrival times of [0.12s, 1.15s, 2.14s, 3.13s], with adjacent intervals of [1.03s, 0.99s, 0.99s]. The median interval after sorting is 0.99s, which is used as the actual refresh cycle and compared with 1000ms. When the runtime scan cycle and the actual refresh cycle are inconsistent, the corresponding mapped point is marked as an invalid quality bit according to the quality bit rule, and the generation of a zero-modification deployment package is blocked. When they are consistent, a zero-modification deployment package is output.

[0054] In summary, this application uses the scan cycle of the control engineering as the time base to strongly couple the update of the pipeline physical process with the discrete scan semantics of the programmable logic controller. At the boundary of each scan cycle, the state variables of the pipeline physical model are deterministically written into the input image area using a point mapping table. The input image area is latched according to the actual update sequence, and the control logic is executed to generate the output image area. Simultaneously, the control variables of the pump unit and valves are kept unchanged within the scan cycle, and the pipeline physical model is updated using sub-cycles. This allows for the reproduction of the actual closed-loop timing on the simulation side, reducing the risk of control deviations and malfunctions caused by timing errors. This scheme elevates the point alignment from "address alignment" to "semantic alignment." A point semantic fingerprint is generated for each mapped point, and before zero-modification deployment, the actual point semantic fingerprint is extracted through shadow consistency verification and compared item by item. This ensures that hidden problems such as point expansion, migration, refresh cycle drift, or quality point processing differences are repeatedly and traceably discovered and intercepted before deployment, avoiding high-cost trial and error. It improved the efficiency of going online, shortened the commissioning cycle, reduced the cost of on-site trial operation, and improved the reliability and security of interlocking and protection strategies.

[0055] Based on another preferred embodiment described above, see [link to preferred embodiment]. Figure 2 As shown, this embodiment provides a pipeline control system simulation platform based on isomorphic simulation of a real environment, used to apply the above-described pipeline control system simulation method based on isomorphic simulation of a real environment, including: The acquisition unit is configured to acquire model parameter data, control engineering data, and point configuration data. The control engineering data includes the scanning cycle and the update sequence of the input and output image areas.

[0056] The building unit is configured to create a physical model of the pipeline based on model parameter data. A virtual programmable logic controller (PLC) is built based on control engineering data. The PLC latches the input image area according to the scan cycle and executes the control logic to obtain the output image area.

[0057] The processing unit is configured to establish a point mapping table based on the point configuration data, and generate point semantic fingerprints for the mapped points in the point mapping table. The point semantic fingerprints are generated by point semantic descriptions of data type, scaling relationship, quality bit rules and refresh cycle.

[0058] The update unit is configured to perform collaborative simulation scheduling with the scan cycle as a unified time base. At the boundary of each scan cycle, the state variables of the pipeline physical model are written into the input image area according to the point mapping table, which drives the virtual programmable logic controller to update the output image area. During the scan cycle, the pump unit control variables and valve control variables corresponding to the output image area remain unchanged. At the same time, the scan cycle is divided into multiple sub-cycles and the pipeline physical model is updated according to the sub-cycles.

[0059] The judgment unit is configured to enter shadow consistency verification before zero-modification deployment, extract the real point semantic fingerprint of the real environment, compare the real point semantic fingerprint with the point semantic fingerprint and verify the scanning cycle and refresh cycle. If the consistency is satisfied, the zero-modification deployment package is output; otherwise, the generation of the zero-modification deployment package is blocked.

[0060] Understandably, by using the scan cycle of the control engineering as the time base, the update of the pipeline physical process is strongly coupled with the discrete scan semantics of the programmable logic controller. At the boundary of each scan cycle, the state variables of the pipeline physical model are deterministically written into the input image area using a point mapping table. The input image area is latched according to the actual update timing, and the control logic is executed to generate the output image area. Simultaneously, the control variables of the pump unit and valves are kept unchanged within the scan cycle, and the pipeline physical model is updated using sub-cycles. This allows for the reproduction of the actual closed-loop timing on the simulation side, reducing the risk of control deviations and malfunctions caused by timing deviations. This solution upgrades point alignment from "address alignment" to "semantic alignment," generating a point semantic fingerprint for each mapped point. Before deployment with zero modifications, the actual point semantic fingerprint is extracted through shadow consistency verification and compared item by item. This allows hidden problems such as point expansion, migration, refresh cycle drift, or quality point processing differences to be repeatedly and traceably discovered and intercepted before deployment, avoiding high-cost trial and error. This improves deployment efficiency, shortens the commissioning cycle, reduces on-site trial operation costs, and enhances the reliability and security of interlocking and protection strategies.

[0061] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit it. Although the present invention has been described in detail with reference to the above embodiments, those skilled in the art should understand that modifications or equivalent substitutions can still be made to the specific implementation of the present invention. Any modifications or equivalent substitutions that do not depart from the spirit and scope of the present invention should be covered within the scope of protection of the claims of the present invention.

Claims

1. A simulation method for pipeline control systems based on isomorphic simulation of real environments, characterized in that, include: Acquire model parameter data, control engineering data, and point configuration data. The control engineering data includes the scanning cycle and the update sequence of the input and output image areas. A pipeline physical model is established based on the model parameter data, and a virtual programmable logic controller is established based on the control engineering data. The virtual programmable logic controller latches the input image area according to the scan cycle and executes control logic to obtain the output image area. A point mapping table is established based on the point configuration data, and a point semantic fingerprint is generated for the mapped points in the point mapping table. The point semantic fingerprint is generated by the point semantic description of data type, scaling relationship, quality bit rule and refresh cycle. Cooperative simulation scheduling is performed using the scanning cycle as a unified time reference. At the boundary of each scanning cycle, the state variables of the pipeline physical model are written into the input image area according to the point mapping table, driving the virtual programmable logic controller to update the output image area. The pump unit control variables and valve control variables corresponding to the output image area remain unchanged within the scanning cycle. At the same time, the scanning cycle is divided into multiple sub-cycles and the pipeline physical model is updated according to the sub-cycles. Before zero-modification deployment, a shadow consistency check is performed. The real location semantic fingerprint of the real environment is extracted, and the real location semantic fingerprint is compared with the location semantic fingerprint. The scanning period and the refresh period are checked. If the consistency is satisfied, the zero-modification deployment package is output; otherwise, the generation of the zero-modification deployment package is blocked.

2. The pipeline control system simulation method based on isomorphic simulation of a real environment according to claim 1, characterized in that, When acquiring the model parameter data, the control engineering data, and the point configuration data, the process includes: reading the pipeline asset configuration to obtain the pipe segment connection relationship, pipe segment geometric parameters, medium physical property parameters, and pump unit characteristic curves and valve characteristic curves to form the model parameter data; parsing the actual programmable logic controller (PLC) engineering configuration to obtain the scan cycle and the update sequence of the input image area and the output image area, and determining the scan cycle by comparing it with the engineering configuration through a read-only diagnostic interface to form the control engineering data; and parsing the actual PLC point table and monitoring point table to obtain the data type, scaling relationship, quality bit rules, and refresh cycle to form the point configuration data.

3. The pipeline control system simulation method based on isomorphic simulation of a real environment according to claim 2, characterized in that, When establishing the physical model of the pipeline, the following are included: The pipeline topology is divided into multiple computing units, and pressure, flow rate, and temperature are set as state variables for each computing unit. The state variables are updated according to the sub-cycle in each scan cycle, and the pump unit control variables and valve control variables in the output image area are read as boundary constraints at the boundary of the scan cycle and remain unchanged in the scan cycle. When establishing the virtual programmable logic controller, the following steps are included: latching the input image area at the beginning of each scan cycle and holding it until the end of the scan cycle, submitting the output image area at the end of the scan cycle, and updating the timer and counter according to the scan cycle.

4. The pipeline control system simulation method based on isomorphic simulation of a real environment according to claim 3, characterized in that, When establishing a point mapping table based on the point configuration data, the following is included: The state variables of the pipeline physical model are bound to the input points in the point configuration data, and the pump unit control variables and the valve control variables are bound to the output points in the point configuration data. When there are duplicate bindings or missing bindings, the generation of the mapping relationship of the corresponding mapping points is prohibited.

5. The pipeline control system simulation method based on isomorphic simulation of a real environment according to claim 4, characterized in that, Generating point semantic fingerprints for each mapping point includes: The data type, scaling relationship, quality bit rule, and refresh cycle are extracted from the mapped points, and normalized and encoded in a fixed field order to obtain the point semantic description. The point semantic description is then summarized to generate the point semantic fingerprint.

6. The pipeline control system simulation method based on isomorphic simulation of a real environment according to claim 4, characterized in that, When performing the collaborative simulation scheduling with the scan cycle as a unified time reference, the process includes: writing the state variables of the pipeline physical model into the input image area at the boundary of each scan cycle; driving the virtual programmable logic controller to complete one scan cycle and submit the output image area; keeping the pump unit control variable and the valve control variable unchanged within the scan cycle, and updating the pipeline physical model according to the sub-cycle with the pump unit control variable and the valve control variable as boundary constraints.

7. The pipeline control system simulation method based on isomorphic simulation of a real environment according to claim 6, characterized in that, Before writing the state quantity into the input image area, the process includes: applying a communication delay and communication jitter to the state quantity and generating an arrival time; when the arrival time is later than the boundary of the next scan cycle, marking the corresponding mapped point as an invalid quality bit according to the quality bit rule, and maintaining the latch value of the previous scan cycle.

8. The pipeline control system simulation method based on isomorphic simulation of a real environment according to claim 7, characterized in that, When entering the shadow consistency check, it includes: The system reads the real programmable logic controller point table and monitoring point table in read-only mode, generates a real point semantic description and a real point semantic fingerprint according to the same field order as the point semantic description; compares the real point semantic fingerprint with the point semantic fingerprint point by point, and when there is an inconsistency between the mapped point marked as an interlocking point or a protection point by the point configuration data, it blocks the generation of the zero-modification deployment package and outputs a difference list.

9. The pipeline control system simulation method based on isomorphic simulation of a real environment according to claim 8, characterized in that, Verifying the scan period and the refresh period includes: The runtime scan cycle of the actual programmable logic controller is read through the read-only diagnostic interface and compared with the scan cycle in the control engineering data; the arrival time series of the mapped points are collected, and the median interval between adjacent arrival times is calculated as the actual refresh cycle within an observation window covering no less than three refresh cycles, and compared with the refresh cycle in the point semantic description; when the runtime scan cycle or the actual refresh cycle is inconsistent, the corresponding mapped point is marked as an invalid quality bit according to the quality bit rule, and the generation of the zero-modification deployment package is blocked.

10. A pipeline control system simulation platform based on isomorphic simulation of a real environment, used to apply the pipeline control system simulation method based on isomorphic simulation of a real environment as described in any one of claims 1-9, characterized in that, include: The acquisition unit is configured to acquire model parameter data, control engineering data, and point configuration data. The control engineering data includes the scanning cycle and the update sequence of the input image area and the output image area. The building unit is configured to build a pipeline physical model based on the model parameter data; and to build a virtual programmable logic controller based on the control engineering data. The virtual programmable logic controller latches the input image area according to the scan cycle and executes control logic to obtain the output image area. The processing unit is configured to establish a point mapping table based on the point configuration data, and generate point semantic fingerprints for the mapped points in the point mapping table. The point semantic fingerprints are generated by point semantic descriptions of data type, scaling relationship, quality bit rules and refresh cycle. The update unit is configured to perform collaborative simulation scheduling with the scan cycle as a unified time reference. At the boundary of each scan cycle, the state variables of the pipeline physical model are written into the input image area according to the point mapping table, driving the virtual programmable logic controller to update the output image area. During the scan cycle, the pump unit control variables and valve control variables corresponding to the output image area are kept unchanged. At the same time, the scan cycle is divided into multiple sub-cycles and the pipeline physical model is updated according to the sub-cycles. The judgment unit is configured to enter shadow consistency verification before zero-modification deployment, extract the real point semantic fingerprint of the real environment, compare the real point semantic fingerprint with the point semantic fingerprint and verify the scanning period and the refresh period. If the consistency is satisfied, the zero-modification deployment package is output; otherwise, the generation of the zero-modification deployment package is blocked.

Images