A method and system for preventing misoperation based on process flow diagram

By constructing a topological model and dynamic mapping relationship of the process flow diagram, the electrical parameters of the controlled physical entity are monitored in real time, which solves the risk of malfunction caused by changes in physical topology in flexible manufacturing, realizes safe control and predictive maintenance of the actuator, and improves the robustness and safety of the control system.

CN123067384APending Publication Date: 2026-11-10山东港源管道物流有限公司
View PDF 1 Cites 0 Cited by

Patent Information

Application Number
CN202611234384.2
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2026-08-14
Publication Date
2026-11-10

AI Technical Summary

Technical Problem

In flexible manufacturing or complex chemical production scenarios, existing technologies cannot perceive changes in path accessibility caused by dynamic changes in physical topology in real time, and lack the ability to predictively lock out sub-healthy states and hidden connectivity risks of actuators, leading to the risk of control system malfunction.

Method used

By analyzing the process flow diagram to construct an initial topology model, the underlying electrical parameters of the controlled physical entity are obtained in real time. The response damping value is calculated using the dynamic mapping relationship. Combined with the operation priority weight and the admission judgment threshold, the control logic is dynamically adjusted to ensure that the path is open or the operation command is intercepted, thereby realizing real-time monitoring and safety assurance of the actuator.

Benefits of technology

It achieves deep integration of process flow diagram and physical topology, senses and responds to changes in physical topology in flexible manufacturing, eliminates the risk of malfunction, enhances the robustness and safety of the control system, avoids instruction overflow or logic glitches caused by mechanical lag, and has sensor drift recognition capability and predictive maintenance function.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN123067384A_ABST
    Figure CN123067384A_ABST
Patent Text Reader

Abstract

This invention relates to the field of industrial control technology and discloses a method and system for preventing misoperation based on a process flow diagram. The method includes: parsing the process flow diagram to construct an initial topology model and obtaining underlying electrical parameters containing the current component of the actuator motor and the potential gradient of the travel sensor; converting these parameters into response damping values ​​using a dynamic mapping relationship, the amplitude of which is adjusted by the travel completion degree and potential gradient fluctuations; calculating a weighted algebraic sum of the operation priority weight and the response damping value; and determining logical connectivity and executing instruction interception by comparing this weighted algebraic sum with an admission threshold. This invention utilizes underlying electrical parameters to fit the physical evolution trajectory of the actuator in real time, eliminating the spatiotemporal misalignment between control semantics and physical inertia, ensuring that instruction execution is based on a defined physical state, and enhancing the stability of the control system under dynamic changing conditions.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] This invention relates to a method and system for preventing misoperation based on a process flow diagram, belonging to the field of industrial control technology. Background Technology

[0002] When distributed control systems or programmable logic controllers execute process logic, they typically use pre-defined hard-wired interlock relationships or fixed logic scripts to build anti-misoperation mechanisms. To ensure that the execution of underlying instructions conforms to predetermined safety specifications, the stability of the control system depends heavily on the consistency between the instruction sequence and the physical topology. Specifically, the system can only accurately identify and intercept unauthorized operation requests when the logical path defined in the process flow diagram is aligned with the physical connection status of the underlying actuators in real time. However, in flexible manufacturing or complex chemical production scenarios, the physical pipeline and equipment connection relationships often change dynamically with manual valve switching, temporary maintenance bypasses, or process path adjustments. In this case, static interlock logic will cause severe decoupling between control semantics and physical topology, resulting in the control system being unable to perceive the logical on / off changes represented by the process flow diagram. Ultimately, this causes the control program to lose its constraint effect in the face of dynamic physical boundaries, thus generating a serious risk of maloperation.

[0003] However, error prevention solutions relying solely on software logic also have significant shortcomings in practical applications. For example, Chinese invention patent CN107436567A discloses an error prevention system and method for industrial processes. This solution generates electronic operation confirmation cards through a background management system and issues unlocking permissions by combining computer keys and real-time parameters collected by DCS. Although this technology achieves electronic supervision of operation tickets, its core logic is essentially still static rule matching, that is, determining access based on preset operation conditions. The logic loopholes are compensated for by manual review or by adding redundant sensors, which leads to increased system maintenance costs. Furthermore, it cannot cope with the non-instantaneous evolution characteristics of actuator actions in the time domain, and the control program generates instruction overflow or false interception during non-steady-state transition periods. Specifically, the existing technology has the following shortcomings: the control logic and the visualized process flow diagram are semantically isolated; it cannot perceive changes in path reachability caused by dynamic changes in physical topology in real time; and it lacks predictive interlocking capabilities for sub-healthy states of actuators and hidden connectivity risks.

[0004] Therefore, how to achieve deep integration of process semantics and control commands, and generate adaptive defense boundaries based on physical topology constraints, has become the technical problem to be solved by this invention. Summary of the Invention

[0005] To address the problems mentioned in the background section, the technical solution of this invention is as follows: A method for preventing misoperation based on a process flow diagram, comprising the following steps: Step S101: Analyze the process flow diagram, extract the process nodes and the connection relationships between each process node, and construct an initial topology model representing the physical equipment relationship. Step S102: Real-time acquisition of the underlying electrical parameters of the controlled physical entity, including the output current component of the actuator and the potential change gradient of the travel sensor. Step S103: Using a preset dynamic mapping relationship, the harmonic characteristics in the output current component and the time-domain deviation of the potential change gradient are converted into the response damping value of the corresponding topological path. The amplitude of the response damping value is inversely correlated with the travel completion degree of the controlled physical entity and positively correlated with the variance of the potential change gradient. This is used to correct the mechanical response lag of the controlled physical entity during the state transition process. Step S104: Receive the operation command for the target process node and extract the operation priority weight of the operation command; Step S105: Calculate the weighted algebraic sum of the operation priority weight and the response damping value corresponding to the target process node; Step S106: Compare the weighted algebraic sum with the preset admission threshold. In step S107, if the weighted algebraic sum exceeds the admission threshold, the path corresponding to the target process node is determined to be open, and the drive path of the controlled physical entity is released; if the weighted algebraic sum does not exceed the admission threshold, the path corresponding to the target process node is determined to be closed, and the operation command is intercepted to block the drive output of the corresponding controlled physical entity.

[0006] Preferably, generating the response damping value using a preset dynamic mapping relationship includes: step S201, extracting the feedback current signal of the controlled physical entity during the transition from startup to steady state; step S202, calculating the peak factor of the feedback current signal within a preset sampling period, and determining the real-time state change trend of the controlled physical entity by combining the slope of the potential change gradient; step S203, performing nonlinear correction on the response damping value based on the deviation between the real-time state change trend and the preset steady-state benchmark, so that the response damping value increases when the controlled physical entity experiences hysteresis.

[0007] Preferably, after step S102, the method further includes: step S301, calculating the admission reference value of each process node based on the underlying electrical parameters; step S302, associating each topological edge in the initial topology model with a preset opening threshold; step S303, determining whether the admission reference value of the current active node meets the opening threshold corresponding to the target topological edge; and step S304, if not, determining that the operation command is an illegal migration command and intercepting it.

[0008] Preferably, after step S101, the method further includes: step S401, monitoring the maintenance status signal of the controlled physical entity; step S402, when the controlled physical entity is detected to be in maintenance status, adding a corresponding temporary maintenance path in the initial topology model according to the preset maintenance logic library, so as to reconstruct the connectivity of the initial topology model; wherein, the temporary maintenance path is limited to unidirectional logical conduction and is used to isolate the process loops of the maintenance area and the operating area.

[0009] Preferably, the initial topology model construction process includes: step S501, identifying the source node in the initial topology model; step S502, determining the activation weight of each process path in the initial topology model based on the underlying electrical parameters; step S503, using a connectivity algorithm, calculating the logical reachable domain starting from the source node based on the activation weight, and generating a dynamic constraint matrix based on the logical reachable domain.

[0010] Preferably, after intercepting the operation command, the method further includes: step S601, extracting the action time features of the target process node in each execution process in real time, and generating a time-domain fingerprint sequence; step S602, calculating the deviation rate η of the time-domain fingerprint sequence relative to the standard time reference, which is calculated as: η=|T real -T std | / T std , among which, T real T is the current measured motion time of the controlled physical entity when performing an operation on the target process node. std The preset standard action time benchmark is used; in step S603, if the deviation rate η exceeds the preset sub-health risk threshold, the admission weight of the target process node is reduced.

[0011] Preferably, before intercepting the operation command, the method further includes: step S701, controlling the upstream process component in the initial topology model to generate a test disturbance signal; step S702, acquiring the monitoring signal of the downstream process component, and extracting the response features in the monitoring signal that have a specific frequency correlation with the test disturbance signal; step S703, calculating the physical connectivity probability between the upstream process component and the downstream process component based on the response features; and step S704, if the physical connectivity probability is inconsistent with the connectivity determination of the initial topology model, outputting a topology conflict alarm and locking the subsequent commands of the corresponding process node.

[0012] Preferably, the interception operation command includes: step S801, while intercepting the operation command, identifying the fault node that caused the path to be broken; step S802, performing visual highlighting of the fault node, the broken path and the associated interlocking equipment in the process flow diagram.

[0013] Preferably, after constructing the initial topology model in step S101, the method further includes: step S901, obtaining the associated process parameter gradient of the topology edge connecting adjacent process nodes; step S902, calculating the consistency residual between the associated process parameter gradient and the measured state data of the corresponding process node; step S903, when the consistency residual exceeds the preset residual threshold, correcting the connectivity weight of the corresponding topology edge to the open circuit state to perform reliability compensation on the initial topology model. After releasing the driving path of the controlled physical entity, the method further includes: real-time monitoring of the process parameter change rate after the driving path is released; if the process parameter change rate does not match the preset energy evolution trend of the operation command, then performing emergency blocking and intercepting all subsequent operation commands.

[0014] A misoperation prevention control system based on a process flow diagram, the system comprising: The topology modeling unit is used to parse the process flow diagram to extract process nodes and the connection relationships between process nodes, and to build an initial topology model that represents the physical equipment relationship. The parameter acquisition unit is used to acquire the underlying electrical parameters of the controlled physical entity in real time, including the output current component of the actuator and the potential change gradient of the travel sensor. The response damping calculation unit is used to convert the time-domain deviation of the harmonic characteristics and potential change gradient in the output current component into the response damping value of the corresponding topological path using a preset dynamic mapping relationship. The instruction parsing unit is used to receive operation instructions for the target process node and extract the operation priority weights of the operation instructions; The logic arbitration unit is used to calculate the weighted algebraic sum of the operation priority weight and the response damping value corresponding to the target process node, and compare the weighted algebraic sum with the preset admission judgment threshold. The drive control unit is used to release the drive path of the controlled physical entity or intercept operation commands to block the drive output of the corresponding controlled physical entity based on the comparison result of the logic arbitration unit.

[0015] Compared with the prior art, the beneficial effects of the present invention are: 1. In the process flow diagram, the physical relationships of the process flow diagram are extracted to construct a directed topology model, which maps the physical entity state data in real time and generates a dynamic constraint matrix that evolves with the physical boundary. This enables the control system to perceive the dynamic connection and disconnection of the process path, solving the logic vacuum caused by changes in physical topology in flexible manufacturing scenarios with varying operating conditions. In addition, this invention introduces a process node logic potential energy gradient verification mechanism, which calculates the potential energy scalar based on the real-time parameters of active nodes, compares it with the topology edge admission potential energy threshold to determine the path conduction status, and couples the geometric connectivity of the path with the maturity of the process state. This ensures that the issuance of action commands is based on the prior physical state meeting the standards, eliminating the risk of system surge or medium backflow caused by actions triggered when physical parameters have not reached the stable range.

[0016] 2. This invention establishes a physical execution trajectory logic impedance model, extracts the travel feedback signal and execution speed gradient during the non-steady-state transition period of the actuator, calculates the dynamic logic impedance value of the topological edge, suppresses or releases the execution intensity gradient of the operation command, solves the problem of fuzzy logic judgment in the intermediate state of the actuator's travel, realizes asynchronous alignment between the command flow and the physical travel flow, enhances robustness to continuous command sequences, and avoids command overflow or logic glitches caused by mechanical lag; it adopts a multi-source physical criterion consistency residual self-compensation and micro-excitation response detection mechanism, uses the energy conservation criterion to calculate the node flow or pressure residual, and combines the response characteristics generated by the upstream node's micro-amplitude modulation signal in the downstream to invert the implicit connectivity state of the physical layer.

[0017] 3. By reusing information to achieve cross-audit logic, the system is equipped with the ability to identify sensor reading drift, logic freeze, or pipeline leakage, and corrects the topology model connectivity weights to ensure that the control logic is consistent with the physical reality boundary. This invention combines the action duration fingerprint of the actuator to establish early warning interlocking logic, analyzes the time distribution characteristics during the state transition of physical entities, calculates the rate of change of offset relative to the standard time base, and reduces the access weight of process nodes when the offset meets the risk evolution model. Predictive maintenance is embedded in the real-time control arbitration core, enabling the system to avoid secondary accidents through logic interception before equipment failure, realize the transition from state interlocking to performance interlocking, and ensure the safety of the process sequence. Attached Figure Description

[0018] Figure 1 This is a flowchart of the control logic for integrating dynamic damping mapping and threshold arbitration in this invention. Figure 2 This is a diagram showing the nonlinear evolution of the admission weight as a function of the action deviation rate in this invention. Figure 3 This is a diagram of the architecture of the anti-misoperation control system for multi-dimensional functional modules in this invention. Detailed Implementation

[0019] The following embodiments are intended to explain the present invention and are not intended to limit the scope of protection of the present invention.

[0020] This invention provides a method and system for preventing misoperation based on a process flow diagram, including the output current component of an actuator and the potential change gradient of a travel sensor. The system utilizes a preset dynamic mapping relationship to convert the harmonic characteristics of the output current component and the time-domain deviation of the potential change gradient into a response damping value D for the corresponding topology path. In the quantization process of determining the response damping value, a linear mapping between physical displacement and electrical signal range is first established. The potential feedback signal from the travel sensor is acquired in real time and compared with a pre-calibrated total potential span of the entire travel. The travel completion degree is then quantified by the proportion of the real-time potential to the total potential span. Based on this travel completion degree, the system performs a composite calculation for the damping value. Specifically, the variance of the potential change gradient refers to the variance of the real-time acquired travel sensor potential change within a preset sliding sampling window. The value obtained by performing variance statistical calculation on the gradient signal sequence, with the sliding sampling window shifting forward in real time with the control cycle, is used to capture the mechanical evolution stability of the controlled physical entity at the current moment and in its preceding neighborhood. The composite calculation specifically involves multiplying a preset damping gain coefficient by the sliding variance of the potential change gradient, which characterizes mechanical oscillations, and simultaneously multiplying a preset unsteady-state correction factor by the remaining stroke proportion obtained by subtracting the stroke completion percentage from the value 1. Finally, the two product terms are algebraically summed to obtain the response damping value D, which reflects the physical evolution state in real time. This quantization process transforms instantaneous mechanical fluctuations and remaining displacement space into logic impedances with definite physical meanings. This allows for the calculation of the weighted algebraic sum of the operation priority weight w for the operation command targeting the process node and the response damping value D, which is then multiplied by a preset admission threshold T. th A quantitative comparison is performed; if the weighted algebraic sum exceeds the admission threshold T... th If the path corresponding to the target process node is deemed to be active, the driving path of the controlled physical entity is released. If the weighted algebraic sum does not exceed the admission threshold T... th If the path corresponding to the target process node is broken, the operation command is intercepted and the drive output of the corresponding controlled physical entity is blocked. In the process of building the initial topology model, the system identifies the source node in the initial topology model, determines the opening weight of each process path in the initial topology model according to the underlying electrical parameters, uses connectivity algorithms such as depth-first search, calculates the logical reachable domain starting from the source node based on the opening weight, and generates a dynamic constraint matrix based on the logical reachable domain. This matrix is ​​used to limit the legal migration path of the control command.

[0021] To correct for the mechanical response hysteresis of the controlled physical entity during state transition, the system extracts the feedback current signal of the controlled physical entity during the start-up to steady-state transition and calculates the peak factor C of the feedback current signal within a preset sampling period. f Crest factor C fThe ratio of the peak value to the effective value of the feedback current signal is used to characterize the fluctuation of the current signal. The system combines the slope of the potential change gradient to determine the real-time state change trend of the controlled physical entity, and performs nonlinear correction on the response damping value D according to the deviation between the real-time state change trend and the preset steady-state benchmark, so that it increases when the controlled physical entity experiences hysteresis. The amplitude of the response damping value D is inversely correlated with the stroke completion degree of the controlled physical entity and positively correlated with the variance of the potential change gradient. The parameter acquisition unit collects the motor current component at a frequency of 1000Hz, extracts the three lowest harmonic characteristics of the energy proportion using fast Fourier transform, and simultaneously calculates the sliding variance σ of the potential change gradient within a 100ms sliding sampling window. 2 Completion rate C p The physical essence of the dynamic mapping relationship in this invention is to establish a mathematical bridge that aligns the electromechanical energy evolution characteristics of the controlled physical entity across dimensions to the damping constraints of the logic layer. During the mapping process, the system continuously compares the real-time collected electromechanical coupling parameters with a preset steady-state benchmark. This preset steady-state benchmark is defined as the vector space formed by the standard current peak factor and the standard potential gradient slope when the actuator operates normally under standard conditions. Thus, the stability characteristics of the physical entity during state transitions are accurately characterized by the coupling fluctuations of the electromechanical parameters. Based on this mapping logic, the system follows formula D. k =α·σ 2 +β·(1-C p The instantaneous response damping value during the sampling period was calculated. The damping gain coefficient α and the unsteady-state correction factor β were derived from the fitting model of 50 sets of load gradient measurement dynamic tests. The peak factor C of the feedback current signal was monitored in real time. f If C f A deviation of 15% from the steady-state baseline triggers the nonlinear correction function e. ΔCf Increase the damping amplitude to compensate for the ambiguity in logical judgment caused by mechanical response lag. Output the judgment result to the logic arbitration unit to realize the quantitative constraint of the intermediate state of physical stroke and eliminate the spatiotemporal misalignment between control semantics and physical inertia.

[0022] In the instruction arbitration phase, the system receives operation instructions for the target process node. Specifically, the operation instructions are transmitted in the form of structured data packets on the control bus. The instruction message includes the target node address, function code, and initial weight values ​​representing process safety attributes. The initial weight values ​​are static parameters pre-calibrated based on the process specifications, reflecting the inherent impact of different levels of action on system safety. After extracting these initial weights, the instruction parsing unit retrieves the time-domain fingerprint sequence corresponding to the target process node from memory and introduces real-time sub-health assessment feedback of the physical entity, deeply fitting the logic layer instruction requirements with the physical layer execution capabilities. The system uses the deviation rate between the current measured action time and the standard action time benchmark to assess the performance degradation degree of the physical entity, and uses this deviation rate as a correction factor to perform nonlinear reduction of the initial weights, thereby extracting the operation priority weight w. The weighted algebraic sum of the operation priority weight w and the response damping value D is calculated, and the result is compared with the admission judgment threshold T. th To perform the comparison, in specific numerical applications, an admission threshold T is set. th If the current extracted operation priority weight w is 0.80, the response damping value D is 0.10, and the weighting coefficients are all 1.0, then the weighted algebraic sum is 0.90, the system determines the path is open and releases the drive path; if the variance of the potential change gradient increases due to the lag in the actuator's action, causing the response damping value D to rise accordingly, and the weighted algebraic sum is lower than 0.75, the system intercepts the operation command and identifies the fault node causing the path to be broken while intercepting the operation command, and performs visual highlighting of the fault node and the broken path in the process flow diagram; for maintenance conditions, the system The system monitors the maintenance status signals of controlled physical entities. When a specific controlled physical entity is detected to be under maintenance, the system adds a temporary maintenance path to the initial topology model according to a preset maintenance logic library to reconstruct the connectivity of the initial topology model. The temporary maintenance path is limited to unidirectional logical conduction to isolate the process loops between the maintenance area and the operating area. The system improves the reliability of the topology model through a consistency residual verification mechanism. The system obtains the associated process parameter gradients of the topology edges connecting adjacent process nodes and calculates the consistency residuals between the associated process parameter gradients and the measured status data of the corresponding process nodes. ;in The absolute value of the deviation between the measured state data and the calculated value based on the evolution law of physical quantities; when the consistency residual... When the residual threshold is exceeded, the system corrects the connectivity weight of the corresponding topology edge to an open circuit state to perform reliability compensation on the initial topology model. After the system releases the drive path, it continuously monitors the rate of change of process parameters. If the rate of change of process parameters is found to be inconsistent with the preset energy evolution trend of the operation command, an emergency interlock is executed and subsequent operation commands are intercepted. For the health status assessment of the actuator, the system extracts the action time characteristics of the target process node in each execution process, generates a time-domain fingerprint sequence, and calculates the deviation rate η relative to the standard time reference. Deviation rate η = |T real -T std | / T std Where η is the deviation rate, and T real T is the current measured motion time of the controlled physical entity when performing an operation on the target process node. std The system uses a preset standard action time benchmark; if the deviation rate η exceeds the preset sub-health risk threshold, the system reduces the admission weight of the target process node, thereby realizing the conversion from state-locked to performance-locked.

[0023] To detect potential hidden connectivity risks within physical pipelines, the system controls upstream process components in the initial topology model to generate test disturbance signals before intercepting operational commands. These test disturbance signals are micro-amplitude modulated excitation signals S. inc The system acquires monitoring signals from downstream process components and extracts the signals from the monitoring signals that correspond to the micro-amplitude modulation excitation signal S. inc Response characteristics R with specific frequency correlation inc Based on response features R inc Calculate the physical connectivity probability between upstream and downstream process components; if the physical connectivity probability is inconsistent with the connectivity determination of the initial topology model, the system outputs a topology conflict alarm and locks the corresponding process node, driving the control unit to generate a micro-modulation excitation signal S with a frequency range of 10Hz to 200Hz and an amplitude range of 2.0%. inc The cross-correlation algorithm was used to analyze downstream monitoring signals and extract response features R. incThe power spectrum components of the excitation source have the same frequency and coherence. The physical connectivity probability calculation logic involves calculating the maximum value of the normalized cross-correlation coefficient between the response signal and the excitation signal, comparing it with a preset physical boundary connectivity threshold. If the calculated value falls within the range of [0.85, 1.0], physical connectivity is determined. The determination result is XORed with the initial topology model static logic. If the two are inconsistent, the connectivity weight of the corresponding topology edge is cleared by the dynamic constraint matrix, and a conflict alarm is triggered. The process utilizes the physical principle of energy conservation to achieve logical reconstruction and cross-auditing of physical authenticity, avoiding control risks caused by the failure of a single feedback source in the perception layer. This invention also provides a A method and system for preventing misoperation based on a process flow diagram are disclosed. The system includes a topology modeling unit, a parameter acquisition unit, a response damping calculation unit, an instruction parsing unit, a logic arbitration unit, and a drive control unit. The topology modeling unit parses the process flow diagram to construct an initial topology model; the parameter acquisition unit acquires the underlying electrical parameters of the controlled physical entity; the response damping calculation unit converts the underlying electrical parameters into response damping values ​​D using a dynamic mapping relationship; the instruction parsing unit extracts the operation priority weight w; and the logic arbitration unit executes a weighted algebraic sum and an admission threshold T. th The drive control unit is used to perform path release or command interception based on the comparison result.

[0024] Example 1: In a continuous chemical production scenario involving the transport of high-viscosity fluids, the electric actuators in the physical pipelines exhibit mechanical response lag due to prolonged exposure to a corrosive environment. When the system issues an operation command to activate the electric actuator, the feedback signal of the actuator motor deviates asymmetrically from the evolution trajectory of the physical stroke. The system analyzes the process flow diagram and extracts process nodes and their connections, constructing an initial topology model characterizing the relationships between physical devices. The electric actuator is a specific process node in the initial topology model. The parameter acquisition unit acquires the underlying electrical parameters of the electric actuator in real time, including the output current component of the actuator motor and the potential change gradient of the stroke sensor. The response damping calculation unit extracts the harmonic characteristics in the output current component and calculates the response damping value D of the corresponding topology path based on the time-domain deviation of the potential change gradient. During the transition from startup to steady state of the electric actuator, the system calculates the peak factor C of the feedback current signal within a preset sampling period. f Peak factor C fThe ratio of the peak value to the effective value of the feedback current signal is used to determine the real-time state change trend of the controlled physical entity. The system calculates the response damping value D, and the amplitude of the response damping value D is inversely correlated with the stroke completion degree of the controlled physical entity and positively correlated with the variance of the potential change gradient. When the system receives an operation command for the downstream pumping equipment, the command parsing unit extracts the operation priority weight w of the operation command, and the logic arbitration unit calculates the weighted algebraic sum of the operation priority weight w and the response damping value D, and compares it with the preset admission judgment threshold T. th Compare them.

[0025] Because the electric actuator is in a non-steady-state transition period and exhibits mechanical hysteresis, the variance of its potential change gradient is at a non-steady-state level, causing the calculated response damping value D to increase accordingly. Consequently, the weighted algebraic sum of the operation priority weight w and the response damping value D does not exceed the admission threshold T. th Based on this dynamic constraint relationship, the system determines that the path corresponding to the target process node is in an open circuit state, and the drive control unit intercepts the operation command of the pumping equipment and blocks the drive output; the system extracts the action time features of the electric actuator in each execution process to generate a time-domain fingerprint sequence, and calculates the current measured action time T. real Compared to the preset standard motion time base T std The deviation rate η; deviation rate η = |T real -T std | / T std Where η is the deviation rate, and T real T is the current measured motion time of the controlled physical entity when performing an operation on the target process node. std The system uses a preset standard action time benchmark. When the deviation rate η exceeds the preset sub-health risk threshold, the system reduces the admission weight of the process node. The physical evolution trajectory of the underlying electrical parameter feedback is deeply coupled with the logical connectivity of the process flow diagram. The system uses the harmonic characteristics of the output current component and the potential change gradient of the travel sensor to jointly construct the logical impedance model of the physical execution trajectory, realizing the asynchronous alignment of the command flow and the physical travel flow. While intercepting the operation command, the system identifies the fault node that causes the path to be broken and highlights the fault node in the process flow diagram, ensuring that every action of the physical entity is based on a determined preceding physical state.

[0026] Example 2: The effectiveness of the method of this invention was verified on a physical experimental platform for high-precision proportioning and conveying. Industrial field conditions with mechanical friction fluctuations and electromagnetic interference were simulated. Data acquisition was based on a current sensor with a measurement accuracy of 0.01mA and a displacement transmitter with a linearity better than 0.1%. The experimenters actively superimposed Gaussian white noise with a signal-to-noise ratio of 20dB onto the potential feedback signal of the stroke sensor and introduced a 50Hz power frequency interference harmonic with an amplitude of 5.0% of the reference signal. The core parameter involved in the experiment, the admission threshold T, was also tested. th The setting takes into account the safety level of the process system and the requirements for production continuity. When the system is in the commissioning and trial operation phase, in order to improve the certainty of the physical state determination results, the access determination threshold T is set. th The upper limit of the value range is 0.75; the system analyzes the process flow diagram of the experimental platform to construct the initial topology model and generate the dynamic constraint matrix. The experiment is designed with three control groups. The sample group of this invention adopts the underlying electrical parameter fitting and response damping correction logic, the control group A adopts the static hard-wired interlocking logic, and the control group B adopts the control logic based on the fixed delay. The setting of the sampling period balances the real-time performance of data acquisition and the processing load of the control unit. When the spectrum width of the monitored feedback current signal increases, the sampling frequency is increased. In this experiment, the sampling period is set to 10.0ms.

[0027] After the experiment was started, the load pressure of the electric actuator was changed to simulate different levels of execution lag, and the judgment results of the three sample groups on the interception command were monitored in real time. Table 1 records the key intermediate characteristic values ​​and final judgment states of each sample group under different execution lag intensities. When the execution lag intensity increases, the response damping value D of the sample group of this invention shows a nonlinear upward trend. When the lag intensity is at a high level, the original potential feedback signal fluctuates due to noise. The response damping calculation unit calculates the peak factor C of the feedback current signal. f Identify non-steady-state transition trends; where C f The peak factor is the ratio of the peak value to the effective value of the feedback current signal; the weighted algebraic sum calculated at this time is 0.65, which is lower than the admission threshold T. th The drive control unit intercepts the operation command; control group A releases the drive due to the inability to recognize the intermediate state of the physical stroke, resulting in the risk of malfunction; control group B triggers interception under extremely high lag, but its fixed delay logic causes the system response speed to decrease under low lag conditions.

[0028] Table 1: Comparison of Execution Lag and Command Interception Performance ; To verify the reasonableness of the key parameter range, the test exceeded the admission threshold T. th Under the operating conditions within the design scope, when the admission judgment threshold T is... thWhen set to 0.95, the system experiences false interceptions due to electrical noise interference during normal execution, causing the operation rate to drop to 64.8%; when the admission threshold T... th When set to 0.30, the system's sensitivity to mechanical lag decreases, making it unable to intercept commands from actuators in a sub-optimal state; experimental results show that the defined access threshold T... th The system maintains quantitative monitoring of the execution trajectory while suppressing environmental disturbances; the system synchronously calculates consistency residuals during the experiment. And perform dynamic compensation, whereby, The consistency residual is the absolute value of the deviation between the measured state data of the displacement sensor and the predicted travel value calculated based on the motor output current component. When human interference on the test platform causes the sensor to generate false limit signals, the consistency residual... When the residual value increases to 0.15 and exceeds the preset residual threshold of 0.08, the system corrects the weight of the corresponding topological edge to an open circuit state and locks subsequent instructions through a self-compensation operator. This process utilizes the physical principle of energy conservation to achieve cross-auditing of logical reconstruction and physical authenticity, avoiding control risks caused by the failure of a single feedback source in the perception layer. Through long-term monitoring of the action time characteristics, a time-domain fingerprint sequence of the electric actuator is generated. As the frictional loss of mechanical components increases, the current measured action time T... real Gradually deviating from the standard motion time reference T std When the deviation rate η increases from 3.2% to 15.6%, the system adjusts the operation priority weight w according to the risk evolution model; deviation rate η = |T real -T std | / T std Where η is the deviation rate, and T real T is the current measured motion time of the controlled physical entity when performing an operation on the target process node. std The preset standard action time benchmark is used; when the deviation rate η exceeds 20.0%, the calculated weighted algebraic sum is lower than the admission threshold T. th And triggers a logical latch.

[0029] Example 3: This example combines Figures 1 to 3 This document describes a method and system for preventing misoperation based on a process flow diagram, such as... Figure 1As shown, the process executes the initial topology model step of parsing the process flow diagram to construct the physical equipment relationship. It then enters the parallel processing flow, receives operation instructions, extracts operation priority weights for the target process node, obtains the underlying electrical parameters including the current component of the execution motor and the potential gradient of the travel sensor, and converts the electrical parameters into the response damping value of the corresponding topology path through dynamic mapping transformation. Then, it calculates the weighted algebraic sum of the operation priority weight and the response damping value. The logic arbitration unit determines whether the weighted algebraic sum is greater than the admission judgment threshold. If the judgment result is yes, it exceeds the threshold, the system determines that the path is open and releases the drive path of the controlled physical entity. If the judgment result is no, it does not exceed the threshold, the system determines that the path is closed, and then intercepts the operation instructions and blocks the drive output.

[0030] like Figure 2 As shown in the figure, a rectangular coordinate system is established with the deviation rate η% as the horizontal axis and the admission weight w as the vertical axis. Five key discrete data points (3.2, 8.5, 15.6, 18.5, and 22.3) are marked on the horizontal axis. The figure includes two legend curves: the solid line represents the real-time change trajectory of the admission weight w, and the dashed line represents the reference benchmark of the sub-health risk threshold η=20%. As the deviation rate η (shown on the horizontal axis) increases, the admission weight w (shown on the solid line) exhibits a non-linear decreasing trend. Especially when the deviation rate η approaches or crosses the sub-health risk threshold, the admission weight w decays to suppress high-risk commands. Figure 3 As shown, the system architecture aims to achieve precise control to prevent industrial misoperation. Its functional modules are distributed in a fishbone pattern along the main axis. The initial topology model construction module includes two sub-items: analyzing the process flow diagram and extracting equipment connection relationships. The dynamic damping conversion module covers response damping value calculation and mechanical response hysteresis correction functions. The underlying electrical parameter sensing module is responsible for extracting motor current components and monitoring potential change gradients. The core logic arbitration module performs priority weight extraction, threshold comparison, and command interception operations. The safety assurance system module integrates consistency residual compensation and action time-domain fingerprint generation mechanisms. All modules work together to support the overall operation of the system.

[0031] Example 4: The conveyor cluster of the industrial control system includes a multi-stage centrifugal pump group and a parallel valve array. It is equipped with a current transformer with a sampling frequency of 1000Hz and a stroke encoder with a measurement accuracy of 0.05%. After collecting the underlying electrical parameters, the system executes a discretization calculation procedure for the response damping value D. In the quantization step of this procedure, the system extracts the displacement potential value fed back by the stroke encoder in real time and maps it to a preset full-range reference interval. By calculating the percentage of the current displacement potential to the total potential amplitude of the full stroke, the stroke completion value is accurately determined. Subsequently, the system performs a weighted algebraic mapping to calculate the instantaneous damping amplitude, specifically including: applying the pre-calibrated damping gain coefficient to the sliding variance of the potential change gradient within the current window. The system performs a multiplication operation, simultaneously multiplying the unsteady-state correction factor with the remaining displacement progress obtained by subtracting the aforementioned stroke completion percentage from 100%. The two multiplication results are then algebraically summed to obtain the response damping value D for that cycle. This quantification method ensures that the damping value rises sensitively with increased mechanical fluctuations or a larger proportion of remaining stroke, thus providing a definite physical quantitative basis for subsequent access arbitration. A sliding sampling window of length L is set for the feedback current signal, with 100 sampling points L and a corresponding window duration of 100ms. Within each sampling cycle k, the system extracts the feedback current signal within the window and calculates the peak factor Cf, simultaneously calculating the sliding variance σ of the stroke sensor potential change gradient. 2 The specific calculation logic of the sliding variance is based on the discrete statistical distribution within the sampling period. The system first obtains the potential change gradient sequence consisting of one hundred sampling points within the current sliding sampling window and calculates the arithmetic mean of the sequence. Then, by comparing the squared deviation of each sampling point in the calculation sequence relative to the arithmetic mean, and then summing all the squared deviations algebraically and taking the mean, the variance quantification result that can dynamically characterize the motion stability of the mechanical mechanism is obtained. In order to ensure continuous monitoring of subtle motion disturbances of physical entities, the sliding sampling window is updated in steps of a single sampling point on the time axis. That is, the sampling window between two adjacent calculation periods maintains a data overlap rate of 99%. This high overlap rate single-step sliding mechanism aims to eliminate logical abrupt changes caused by data discontinuity and ensure that the evolution trajectory of the response damping value during the physical state transition process has extremely high temporal smoothness. The amplitude of the response damping value D is calculated according to the formula Dk=α·σ. 2 +β·(1-Cp) is determined; where Dk is the response damping value of the kth sampling period, α is the damping gain coefficient, and σ 2 Let C be the sliding variance of the potential change gradient, β be the unsteady-state correction factor, and Cp be the percentage of stroke completion. p Specifically refers to the dimensionless normalized parameter characterizing the real-time displacement state of a controlled physical entity, whose numerical range is strictly limited to [missing information]. to Between these points, the actuator is used to correspond to the start point of the action (numerical value). ) to the end of the entire journey (numerical value) The real-time displacement ratio of (1-C); correspondingly, (1-C p Physically representing the remaining stroke ratio of the current actuator, it serves as a dynamic weight adjustment factor, establishing a convergence compensation mechanism: during the action initiation phase, due to C... p Smaller, larger remaining stroke ratio (1-C) p Assign a response damping value D k The higher computational weights utilize logic impedance to suppress the mechanical inertia generated by instantaneous high currents; as the journey nears completion, the remaining proportion automatically decreases, thereby smoothly weakening the damping constraints and ensuring that the physical entity can dock precisely; when the peak factor Cf deviates from the preset steady-state benchmark by 15.0%, the system increases the non-steady-state correction factor β to 1.5 times the original value to correct the logic deviation caused by mechanical response lag.

[0032] When the pump unit is under maintenance, the system performs logical reconfiguration. The system obtains the maintenance status signal of the controlled physical entity and calls the virtual topology instructions in the maintenance logic library, superimposing the temporary maintenance path vector V into the adjacency matrix M of the initial topology model. vir The reconstructed adjacency matrix M new Follow the operation rule M new =M∪V vir Wherein, the adjacency matrix M is used to represent the original connection relationship between physical devices, and the temporary maintenance path vector V vir Non-zero elements in the matrix have unidirectional logical conduction properties, used to limit energy migration from maintenance area nodes to operating area nodes; when the logic arbitration unit detects that operating node i is attempting to trigger a transmission command to maintenance node j, the system determines the direction of transmission based on the reconstructed adjacency matrix M. new The system determines the path continuity status. If the connection is subject to unidirectional attribute constraints and is set to a logically disconnected state, the weighted algebraic sum calculated by the logic arbitration unit is set to 0, thereby intercepting the instruction. The system extracts the action time characteristics of the process node and performs dynamic adjustment of the admission weight, extracting the current measured action time T in real time. real And calculate the deviation rate η; deviation rate η = |T real -T std | / T std , among which, T real T represents the current measured action time of the controlled physical entity. stdUsing the standard action time as the reference, η represents the deviation rate. The system determines the weight reduction amount based on the piecewise linear mapping model between the deviation rate η and the admission weight w. When the deviation rate η is greater than 0.1 and less than or equal to 0.2, the admission weight w is reduced according to the linear slope k1. When the deviation rate η is greater than 0.2, the admission weight w is set to 0. Under the condition of increased friction of the actuator seal, the measured value of the deviation rate η is 18.5%. The system reduces the operation priority weight w from 0.90 to 0.62. At this time, the weighted algebraic sum of the operation priority weight w and the response damping value D is 0.58, which is lower than the admission judgment threshold T. th When the value is 0.75, the drive control unit blocks the drive output of the pump unit.

[0033] Example 5: In the calibration condition with adjustable mechanical load, the system executes the benchmark calibration procedure for the response damping value D. A controlled stepping loading device is used to gradually change the running resistance of the controlled physical entity. The harmonic characteristics of the output current component and the gradient variance of the potential change of the stroke sensor are collected in real time under different load levels. A mapping model between the damping gain coefficient α and the fluctuation intensity is obtained through least squares fitting. The system statistically analyzes the bottom-level electrical parameter limits of the actuator before hysteresis or timeout occurs during the controlled experiment, and uses the calculation results as the admission threshold T. th The calibration reference is stored in the logic arbitration unit to complete the access logic setting of the adjustment system during operation. When the system faces the access of new equipment or changes in the field environment, the drive control unit starts the field pre-calibration procedure, drives the controlled physical entity to continuously perform a fixed number of full-stroke actions under no-load conditions, collects the action time distribution characteristics during the execution process and calculates its arithmetic mean, and uses the calculation result as the standard action time reference T. std Used for calculating the deviation rate η; the system synchronously monitors the steady-state fluctuation amplitude of the sensor signal, corrects the residual threshold based on the measured background interference intensity, and performs consistency residual calculation by comparing the fitting error between the measured state data and the initial connectivity weights of the dynamic constraint matrix. The normalization compensation determines the initial process boundary for physical entities with different degrees of mechanical wear.

[0034] The system executes a standardized procedure for the logic impedance coefficient of the actuator nodes. The operator connects the controlled physical entity to a load test bench and records the output current component within a torque range of 0 to 50 N·m. The drive control unit uses a linear regression algorithm to calculate the damping gain coefficient. With the unsteady-state correction factor β; where α is the damping gain coefficient, which is used to quantify the potential change gradient sliding variance σ. 2The contribution weight to the damping amplitude, β is an unsteady-state correction factor, which is used to adjust the influence of the stroke completion deviation on the logic impedance; the system analyzes the amplitude characteristics of the third and fifth harmonics in the current signal by performing discrete Fourier transform, and synchronously collects the gradient sliding variance σ of the stroke sensor potential change under each load step. 2 The calibrated damping gain coefficient α and unsteady-state correction factor β are written into the configuration register of the logic arbitration unit to complete the parameter discretization filling of the response damping value D calculation model. When the system faces changes in the on-site electromagnetic environment, the parameter acquisition unit initiates the environmental noise baseline calibration procedure, continuously monitors the potential feedback signal of the stroke sensor and extracts the background interference intensity V when the actuator is stationary. noise The system is based on the formula =1.2·V noise Dynamically adjust residual threshold; where, The corrected residual threshold, in units of V, V noise The measured potential feedback signal represents the maximum peak-to-peak amplitude within a 5.0s monitoring period, expressed in V. The system synchronously controls upstream process components to generate a micro-modulation excitation signal S with a frequency of 120Hz and an amplitude of 2.0% of the range. inc The response components at corresponding frequencies in the downstream monitoring signals are extracted to calculate the physical connectivity probability, and the calculation results are used to perform consistency residuals on the initial connectivity weights in the dynamic constraint matrix. Normalization process.

[0035] It will be apparent to those skilled in the art that the present invention is not limited to the details of the exemplary embodiments described above, and that the present invention can be implemented in other specific forms without departing from the spirit or essential characteristics of the present invention.

[0036] Finally, 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.

Claims

1. A method for preventing misoperation based on a process flow diagram, characterized in that, The method includes the following steps: Step S101: Analyze the process flow diagram, extract the process nodes and the connection relationships between each process node, and construct an initial topology model representing the physical equipment relationship. Step S102: Real-time acquisition of the underlying electrical parameters of the controlled physical entity, including the output current component of the actuator and the potential change gradient of the travel sensor. Step S103: Using a preset dynamic mapping relationship, the harmonic characteristics and time-domain deviation of the potential change gradient in the output current component are converted into the response damping value of the corresponding topological path. The amplitude of the response damping value is inversely correlated with the travel completion degree of the controlled physical entity and positively correlated with the sliding variance of the potential change gradient of the travel sensor within a preset sampling period. This is used to correct the mechanical response hysteresis of the controlled physical entity during the state transition process. Step S104: Receive operation instructions for the target process node, parse the operation instructions to obtain initial weights, and determine the operation priority weights by combining the sub-health status evaluation results of the controlled physical entity. Step S105: Calculate the weighted algebraic sum of the operation priority weight and the response damping value corresponding to the target process node; Step S106: Compare the weighted algebraic sum with the preset admission threshold. In step S107, if the weighted algebraic sum exceeds the admission threshold, the path corresponding to the target process node is determined to be open, and the drive path of the controlled physical entity is released; if the weighted algebraic sum does not exceed the admission threshold, the path corresponding to the target process node is determined to be closed, and the operation command is intercepted to block the drive output of the corresponding controlled physical entity.

2. The method for preventing misoperation based on a process flow diagram according to claim 1, characterized in that, The process of generating a response damping value using a preset dynamic mapping relationship includes: step S201, extracting the feedback current signal of the controlled physical entity during the transition from startup to steady state; step S202, calculating the peak factor of the feedback current signal within a preset sampling period, and determining the real-time state change trend of the controlled physical entity by combining the slope of the potential change gradient; step S203, performing nonlinear correction on the response damping value based on the deviation between the real-time state change trend and the preset steady-state benchmark, so that the response damping value increases when the controlled physical entity experiences hysteresis.

3. The method for preventing misoperation based on a process flow diagram according to claim 1, characterized in that, After step S102, the method further includes: step S301, calculating the admission reference value of each process node based on the underlying electrical parameters; step S302, associating each topological edge in the initial topology model with a preset opening threshold; step S303, determining whether the admission reference value of the current active node meets the opening threshold corresponding to the target topological edge; and step S304, if not, determining that the operation command is an illegal migration command and intercepting it.

4. The method for preventing misoperation based on a process flow diagram according to claim 1, characterized in that, Following step S101, the method further includes: step S401, monitoring the maintenance status signal of the controlled physical entity; step S402, when the controlled physical entity is detected to be in maintenance status, adding a corresponding temporary maintenance path in the initial topology model according to the preset maintenance logic library to reconstruct the connectivity of the initial topology model; wherein, the temporary maintenance path is limited to unidirectional logical conduction and is used to isolate the process loops of the maintenance area and the operating area.

5. The method for preventing misoperation based on a process flow diagram according to claim 1, characterized in that, The initial topology model construction process includes: step S501, identifying the source node in the initial topology model; step S502, determining the activation weight of each process path in the initial topology model based on the underlying electrical parameters; step S503, using a connectivity algorithm, calculating the logical reachable domain starting from the source node based on the activation weight, and generating a dynamic constraint matrix based on the logical reachable domain.

6. The method for preventing misoperation based on a process flow diagram according to claim 1, characterized in that, After intercepting the operation command, the process further includes: step S601, extracting the action time features of the target process node in each execution process in real time to generate a time-domain fingerprint sequence; step S602, calculating the deviation rate η of the time-domain fingerprint sequence relative to the standard time reference, which is calculated as: η=|T real -T std | / T std , among which, T real T is the current measured motion time of the controlled physical entity when performing an operation on the target process node. std The preset standard action time benchmark is used; in step S603, if the deviation rate η exceeds the preset sub-health risk threshold, the admission weight of the target process node is reduced.

7. The method for preventing misoperation based on a process flow diagram according to claim 1, characterized in that, Before intercepting the operation command, the process includes: step S701, controlling the upstream process component in the initial topology model to generate a test disturbance signal; step S702, acquiring the monitoring signal of the downstream process component and extracting the response features in the monitoring signal that have a specific frequency correlation with the test disturbance signal; step S703, calculating the physical connectivity probability between the upstream and downstream process components based on the response features; and step S704, if the physical connectivity probability is inconsistent with the connectivity determination of the initial topology model, outputting a topology conflict alarm and locking the subsequent commands of the corresponding process node.

8. The method for preventing misoperation based on a process flow diagram according to claim 1, characterized in that, The interception operation command includes: step S801, while intercepting the operation command, identifying the fault node that caused the path to be broken; step S802, performing visual highlighting on the fault node, the broken path and the associated interlocking equipment in the process flow diagram.

9. The method for preventing misoperation based on a process flow diagram according to claim 1, characterized in that, After constructing the initial topology model in step S101, the method further includes: step S901, obtaining the associated process parameter gradients of the topology edges connecting adjacent process nodes; step S902, calculating the consistency residual between the associated process parameter gradients and the measured state data of the corresponding process nodes; step S903, when the consistency residual exceeds a preset residual threshold, correcting the connectivity weight of the corresponding topology edge to an open circuit state to perform reliability compensation on the initial topology model. After releasing the driving path of the controlled physical entity, the method further includes: real-time monitoring of the process parameter change rate after the driving path is released; if the process parameter change rate does not match the preset energy evolution trend of the operation command, then an emergency interlock is executed and all subsequent operation commands are intercepted.

10. A process flow diagram-based anti-misoperation control system, used to implement the process flow diagram-based anti-misoperation control method of claim 1, characterized in that, The system includes: The topology modeling unit is used to parse the process flow diagram to extract process nodes and the connection relationships between process nodes, and to build an initial topology model that represents the physical equipment relationship. The parameter acquisition unit is used to acquire the underlying electrical parameters of the controlled physical entity in real time, including the output current component of the actuator and the potential change gradient of the travel sensor. The response damping calculation unit is used to convert the time-domain deviation of the harmonic characteristics and potential change gradient in the output current component into the response damping value of the corresponding topological path using a preset dynamic mapping relationship. The instruction parsing unit is used to receive operation instructions for the target process node and extract the operation priority weights of the operation instructions; The logic arbitration unit is used to calculate the weighted algebraic sum of the operation priority weight and the response damping value corresponding to the target process node, and compare the weighted algebraic sum with the preset admission judgment threshold. The drive control unit is used to release the drive path of the controlled physical entity or intercept operation commands to block the drive output of the corresponding controlled physical entity based on the comparison result of the logic arbitration unit.

Citation Information

Patent Citations

  • Industrial flow anti-misoperation system and industrial flow anti-misoperation method

    CN107436567A