Auxiliary control workshop control system based on DCS

Abstract

The invention relates to the technical field of industrial automation control, and discloses an auxiliary control workshop control system based on a DCS, and the system comprises a process topological graph construction module which is used for mapping a physical workshop into a dynamic graph containing an equipment digital object in real time; the task analysis module is used for converting the advanced task instruction into a quantitative target state set on a graph; the dynamic path planning module is used for automatically searching and generating an optimal atomic operation instruction sequence for reaching a target state based on the real-time state of the map; and the instruction execution and transaction management module and the exception handling and secure rollback module are used for ensuring atomic execution of the instruction sequence and secure rollback during failure. According to the method, a process control problem is converted into a graph theory path planning problem, so that the control system has the capabilities of dynamically avoiding faults and autonomously deciding, and the automation level, the reliability and the safety of the operation of the auxiliary control system are remarkably improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of industrial automation control, in particular to a DCS-based auxiliary control workshop control system. BACKGROUND

[0002] In large industrial production environments such as thermal power plants, auxiliary control workshops (such as chemical water treatment systems, circulating water pump rooms, compressed air stations, etc.) are key components to ensure the stable operation of the main machine and the safety of production. Currently, the automatic control of these workshops generally relies on distributed control systems (DCS) or programmable logic controllers (PLC). In its specific engineering implementation, the control logic is usually constructed using sequential function charts (SFC) or logic programs fixed in function blocks, and the design idea is to pre-write a set of fixed and step-by-step start-stop steps and interlocking conditions for a specific production task (such as "start No. 1 circulating pump group").

[0003] This pre-set control method performs well when dealing with deterministic and ideal working conditions, however, the complexity and dynamics of the actual operating environment of auxiliary control workshops make the inherent drawbacks of this static logic increasingly prominent.

[0004] The core problem is that the control logic is decoupled from the actual process topology of the physical devices. The control program is essentially a rigid sequence of instructions that does not understand the deep physical and logical relationships between devices, such as "mutual backup", "medium flow", or "electrical superior and inferior". Therefore, when unexpected situations such as device failure, valve sticking, instrument failure, or temporary maintenance isolation of a pipeline occur in the production process, the fixed sequential logic cannot make autonomous decisions based on the global state to dynamically avoid problem points. The result is often that the process is forced to be interrupted, triggering generalized alarms, and then highly relying on the on-site experience and tedious manual intervention of the operators to diagnose faults, switch to backup devices, or perform emergency operations. This not only greatly increases the burden and risk of misoperation of the operators, but also brings unpredictable potential risks to the safe and stable operation of the entire system.

[0005] In addition, the rigid nature of this control strategy also leads to the complexity of system operation and maintenance. Any minor changes in the process flow or addition or replacement of devices means tedious reprogramming, repeated testing, and downloading of the underlying control logic, with a long cycle and high risk. At the same time, the system cannot dynamically select a more economical or more beneficial operation path for device maintenance based on real-time energy consumption data or health status assessment, thus having inherent limitations in achieving fine and optimal management. SUMMARY

[0006] In view of the deficiencies of the prior art, the application provides a DCS-based auxiliary control plant control system, in the prior control system, preset control logic is separated from real-time process topology states of physical devices in a plant, so that the system cannot make autonomous decisions and dynamic re-planning when facing dynamic conditions such as device failure, maintenance or temporary change of a process path, and thus problems of low system operation reliability and high dependence on manual intervention are caused.

[0007] To solve the above technical problems, the application provides the following technical solutions.

[0008] The first aspect of the application provides a DCS-based auxiliary control plant control system, comprising:

[0009] one or more processors;

[0010] a memory having instructions executable by the one or more processors stored thereon;

[0011] The instructions are executed by the one or more processors to implement:

[0012] a process topology graph construction module configured to construct a dynamic process topology graph comprising a plurality of device digital objects according to devices in the auxiliary control plant and their associated relationships;

[0013] a task analysis module configured to analyze received high-level task instructions into a target state set of the dynamic process topology graph;

[0014] a dynamic path planning module configured to dynamically plan and generate an operation instruction sequence for achieving the target state set according to real-time states of the dynamic process topology graph and the target state set;

[0015] an instruction execution module configured to execute instructions in the operation instruction sequence in sequence;

[0016] an exception handling module configured to trigger a preset safety rollback mechanism to restore device states of the auxiliary control plant to a preset safety state when the instruction execution module fails to execute instructions.

[0017] In one specific embodiment, each device digital object in the dynamic process topology graph corresponds to a physical device or a logical entity, and encapsulates static attributes, dynamic attributes and a finite state machine defining all possible states and state transition conditions of the object; edges in the dynamic process topology graph represent physical or logical association relationships between device digital objects.

[0018] Preferably, the dynamic path planning module generates the sequence of operation instructions by running a graph search algorithm on one or more processors, and during the search process, the dynamic path planning module is configured to automatically avoid equipment digital objects marked as "fault" or "maintenance isolation" status in the dynamic process topology graph.

[0019] In one embodiment, the evaluation function employed by the graph search algorithm during the search process is:

[0020] f(G n ) = g(G n ) + h(G n );

[0021] where f(G n ) is the estimated total cost from the initial state to the goal state; g(G n ) is the actual cumulative cost of reaching the current intermediate graph state G n from the initial state; and h(G n ) is the estimated cost of reaching the goal state set from the current intermediate graph state G n .

[0022] In one embodiment, the instruction execution module, before executing the sequence of operation instructions, is configured to first create a transaction log for the current task; and when executing each instruction in the sequence, records the instruction content and the state snapshot of the target equipment digital object before execution in the transaction log.

[0023] Further, the safety rollback mechanism triggered by the exception handling module includes at least one of the following modes:

[0024] Reverse state rollback: when the instruction execution fails, the exception handling module reads the transaction log and executes the inverse operation of the successfully executed operation in reverse to restore the equipment state to the state before the task started;

[0025] Safety baseline rollback: when the reverse state rollback cannot be executed or the task risk level is high, the exception handling module executes a globally predefined minimum safety instruction sequence to forcibly restore the associated subsystem to a known minimum safety stable state.

[0026] In one embodiment, when performing the reverse state rollback, the exception handling module is configured to, for each instruction Op i successfully executed before the failed instruction Op k recorded in the transaction log, execute its predefined inverse operation The process can be described as:

[0027] ​

[0028]

[0029] where i is an integer index representing the serial number of the instruction that failed to execute in the entire sequence of operation instructions P = <Op1, Op2,...,OpN>; k is a loop variable; Opk represents the kth atomic operation instruction in the sequence of operation instructions P; Execute(·) is an execution function, indicating that the operation inside the parentheses is sent to the underlying execution mechanism of the DCS for actual execution. m k k

[0030] Preferably, the system further comprises a virtual execution module; the virtual execution module is configured to: before the instruction execution module sends the sequence of operation instructions to the physical device, first perform a rehearsal on the copy of the dynamic process topology map to verify its logical correctness and pre-analyze its possible cascading effects.

[0031] In a specific embodiment, the task parsing module is configured to parse the high-level task instruction into the target state set, wherein the target state set is composed of one or more target predicates, each target predicate being a constraint on the expected state of a device digital object or edge in the dynamic process topology map.

[0032] The second aspect of the present application provides a DCS-based auxiliary control plant control method, comprising the following steps:

[0033] constructing a dynamic process topology map, the dynamic process topology map containing a plurality of device digital objects and reflecting the devices of the auxiliary control plant and their associated relationships;

[0034] parsing the received high-level task instruction into a target state set of the dynamic process topology map;

[0035] dynamically planning and generating a sequence of operation instructions according to the real-time state of the dynamic process topology map and the target state set;

[0036] sequentially executing the instructions in the sequence of operation instructions;

[0037] when an execution instruction fails, triggering a pre-set safety rollback mechanism to restore the device state of the auxiliary control plant to a pre-set safety state.

[0038] The present application provides a DCS-based auxiliary control plant control system, which has the following beneficial effects:

[0039] ​​​​1、The present application maps the real-time working condition of the physical workshop into a dynamic graph through the process topology graph construction module and the dynamic path planning module, and performs real-time path search based on the graph when receiving a task. This enables the system to intelligently identify and automatically avoid these unavailable paths when some equipment in the workshop fails or enters the maintenance state, and autonomously plan a new optimal execution sequence composed of normal equipment, thereby overcoming the defects of the prior art that the control logic is fixed and cannot respond to unexpected situations, and ensuring the continuity of production tasks.

[0040] 2、The present application introduces the transaction processing mechanism in the database field into industrial control through the instruction execution and transaction management module and the exception handling and safety rollback module. For a complete operation sequence, the system manages in a transaction manner, and ensures the traceability of each operation step through the pre-written log. Once a failure occurs during execution, the system can accurately roll back or execute a plan to enter a safety baseline state according to the log, ensuring the atomicity of task execution and effectively avoiding the system being stuck in an uncertain dangerous intermediate state.

[0041] 3、The present application introduces a quantifiable cost function in dynamic path planning, so that the system can not only judge the feasibility when planning a path, but also comprehensively evaluate factors such as energy consumption and equipment wear to find a globally optimal control strategy. At the same time, since the core of the system is based on the process topology graph model rather than fixed code, when the process or equipment changes, maintenance personnel only need to update the graph model data, without the need to make complex modifications and tests to the underlying control program, greatly reducing the difficulty of operation and maintenance and improving the flexibility of the system. BRIEF DESCRIPTION OF DRAWINGS

[0042] Figure 1 is the overall architecture block diagram of the system of the present application;

[0043] Figure 2 is the device digital object structure schematic diagram of the present application;

[0044] Figure 3 is the dynamic path planning flowchart of the present application;

[0045] Figure 4 is the instruction execution and transaction log recording flowchart of the present application;

[0046] Figure 5 is the exception handling and safety rollback flowchart of the present application.

[0047] Among them, 10, process topology graph construction module; 20, task analysis module; 30, dynamic path planning module; 40, instruction execution module; 50, exception handling module. DETAILED DESCRIPTION

[0048] The technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the drawings in the specification of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative work fall within the scope of the present application.

[0049] Referring to the drawings Figure 1 , Figure 1 is a functional module diagram of a DCS-based auxiliary control plant control system according to an embodiment of the present application. The DCS-based auxiliary control plant control system provided by the present application can be deployed on a computing device composed of one or more processors, memories and communication interfaces. When computer program instructions stored in the memory are executed by the processor, the following functional modules are implemented.

[0050] The system of the embodiment can include a process topology graph construction module 10, a task analysis module 20, a dynamic path planning module 30, an instruction execution module 40, and an exception handling module 50.

[0051] The process topology graph construction module 10 is used to map and instantiate the physical devices of the auxiliary control plant and their process association relationship into a dynamic process topology graph running in the memory. The graph is a directed graph, which is formally described as G=(V, E). Wherein, V is the set of all device digital objects, constituting the node set of the graph; E is the set of directed edges representing the process or electrical association relationship between devices. This graph is connected through a real-time data interface with the DCS, and the states of its nodes and edges can be updated in real time.

[0052] The task analysis module 20 is connected with the process topology graph construction module 10. This module receives externally input high-level task instructions, and parses the instructions into a target state set S goal composed of one or more target predicates. The target state set is an accurate constraint on the final desired state of one or more nodes or edges in the graph G.

[0053] The dynamic path planning module 30 is connected with the process topology graph construction module 10 and the task analysis module 20 at its input end. This module obtains the real-time state G current of the dynamic process topology graph and the target state set S goal , and based on this, runs a graph search algorithm to generate an operation instruction sequence P for achieving the target state set. In an embodiment, the evaluation function used by the graph search algorithm is:

[0054] f(G n )=g(G n )+h(G n);

[0055] where f(G n ) is the estimated total cost from the initial state to the target state; g(G n ) is the actual accumulated cost from the initial state to the current intermediate state G n ; h(G n ) is the estimated cost from the current intermediate state G n to the target state set.

[0056] An instruction execution module 40, whose input is connected to the dynamic path planning module 30 and whose output is connected to the underlying execution mechanism of the DCS. This module receives the sequence of operation instructions P and issues them to the physical devices in order. Before execution, this module creates a transaction log L TX for the current task, which records each step of operation and its result.

[0057] An exception handling module 50, which is connected to the instruction execution module 40. This module is triggered when the instruction execution module 40 fails to execute a certain instruction Op i in the sequence P. The exception handling module 50 executes a pre-defined safety rollback mechanism according to the transaction log L TX . In one embodiment, this rollback mechanism includes a reverse state rollback, whose execution process can be described as:

[0058]

[0059] where i is an integer index representing the serial number of the instruction that failed in the entire sequence of operation instructions P = <Op1, Op2,..., Op m >; k is a loop variable; Op k represents the kth atomic operation instruction in the sequence of operation instructions P; represents the pre-defined inverse operation of the operation instruction Op k ; Execute(·) is an execution function, which means issuing the operation inside the parentheses to the underlying execution mechanism of the DCS for actual execution.

[0060] In one preferred embodiment, the system further includes a virtual execution module 60. This module is set between the dynamic path planning module 30 and the instruction execution module 40. It receives the sequence of instructions P generated by the dynamic path planning module 30 and performs a rehearsal on a temporary copy of the graph before issuing it to the instruction execution module 40, in order to check its logic and analyze the impact.

[0061] Refer to the attached drawings Figure 2 - Figure 5, the core function of the process topology map construction module 10 is to instantiate and maintain a dynamic process topology map that accurately reflects the real-time state of the auxiliary control workshop. The construction process first digitizes each process equipment or logical entity in the workshop to form a digital object.

[0062] Each device digital object v is a standardized data structure, which in this embodiment can be formally defined as a five-tuple:

[0063] v = (id, P s , P d , M state , F op ) ;

[0064] Where: id is the unique identifier of the object, used to distinguish in the system, such as a string or integer; P s is the static attribute set of the object, containing information that does not change over time, such as device model, installation location, rated power, design pipe diameter, etc.; P d is the dynamic attribute set of the object, whose value is bound in real time with the corresponding I / O point in the DCS through the communication interface. This set contains telemetry (such as real-time pressure, temperature, current, valve opening feedback) and telematics (such as running / stop state feedback, fault alarm signal); M state is a finite state machine used to describe the behavior of the device, defined as M state = (S, s0, Σ, δ). Where S is the finite set of all predefined states of the device, such as { stopped, running, fault, maintenance isolation}; s0 ∈ S is the initial state of the device, such as "stopped"; Σ is the set of all events that can trigger state transitions, events can be derived from external instructions or internal state changes, such as {start instruction, stop instruction, fault signal}; δ is the state transition function, which defines the next state to which the device should be converted after receiving a certain event in a certain state, for example δ (stopped, start instruction) = running; F op is the set of atomic operation methods that can be called by the object. These methods are specific implementations of state machine events Σ, for example, calling a method to start the device will issue a specific control instruction to start the device to the DCS.

[0065] During system initialization or engineering configuration, technicians instantiate a series of device digital objects in the engineer station software according to the process flow diagram (PQID) of the auxiliary control workshop, and define the connection relationship between them. The set of object instances constitutes the node set V of the map.

[0066] The connection relationship between nodes constitutes the directed edge set E of the map. Each edge e ∈ E is also defined as a standardized data structure e = (v iv j , e , e ). Where v i and v j are the source and target nodes of an edge; T e is the type of the edge, used to represent different physical or logical relationships, such as "media flow", "electrical supply", "control dependency", or "mutual backup"; W e is the weight or dynamic attribute of the edge, which can be a scalar or vector, used to represent, for example, pipe resistance, rated flow, or pressure difference updated by sensors in real time.

[0067] The process topology map construction module 10 maintains the dynamic nature of the map through a continuously running data synchronization process. This process subscribes to all the I / O points associated with the digital objects of the devices in the DCS real-time database. When the value of any point changes, the module immediately updates the dynamic attribute set P d of the corresponding node v state . The update of the attribute is input as an event to the state machine M state of the node, which may in turn trigger a state transition, thus ensuring that the internal state of every node in the map accurately reflects the real-time working condition of the physical device.

[0068] The function of the task analysis module 20 is to convert the high-level task instructions input by the operator on the human-machine interface (HMI) into precise, quantified goals based on the map that can be understood by the subsequent dynamic path planning module 30.

[0069] The high-level task instructions T cmd input by the operator are process-oriented descriptions, such as selecting "start No. 1 circulating water system" from the menu or clicking "feed B reactor" on the graphical interface. These instructions are unstructured for the computer system. The task analysis module 20 internally contains an analysis function f parse that implements this conversion from unstructured to structured.

[0070] The implementation of the analysis function f parse can be a pre-set mapping table or a rule engine. The mapping table uniquely maps each pre-defined high-level task instruction T cmd to a target state set S goal . The mapping relationship can be represented as:

[0071] S goal = f parse (T cmd );

[0072] The target state set S goal is a set composed of one or more target predicates c, i.e. Sgoal = {c1, c2,..., c k}. Each target predicate is a logical constraint on the final state of some property of some element in the dynamic process topology graph G, i.e. some device digital object node or some process pipeline edge, that should be achieved at the completion of the task. This way, the task objective is ensured to be concrete, quantifiable and verifiable on the graph.

[0073] To further illustrate, take the high-level task instruction T cmd “Start Cycle 1 Water System” as an example, after being processed by the task parsing module 20, the generated target state set S goal may contain, but not limited to, the following target predicates:

[0074] c1: for the device digital object with identifier “PUMP-101A” (i.e. the main pump A of Cycle 1 Water System), its dynamic property set P d contains a running status property whose value should be “Running”.

[0075] c2: for the device digital object with identifier “VALVE-101A-OUT” (i.e. the outlet valve of main pump A), its dynamic property set P d contains a valve position feedback property whose value should be 100, indicating a full open state.

[0076] c3: for the edge object with identifier “PIPE-MAIN-01” (i.e. the main water supply pipeline of Cycle 1 Water System), its dynamic weight property W e contains a flow value that should be greater than a minimum flow threshold F threshold set according to process requirements.

[0077] In this way, the task parsing module 20 decomposes an abstract task objective into a series of concrete, refined constraints on the final state of different elements in the graph. This generated target state set S goal is passed to the dynamic path planning module 30 as the end point and verification condition for its path search.

[0078] The dynamic path planning module 30 is responsible for generating an executable operation instruction sequence after obtaining the graph real-time state G current provided by the process topology graph construction module 10 and the target state set S goal generated by the task parsing module 20.

[0079] In this embodiment, the module uses an A* graph search algorithm to find a path in the state space from G current to a state that satisfies S goalto the goal state. Here, the "state space" is composed of all possible states of the dynamic process topology map, and the transition between states is triggered by performing an atomic operation on a certain device digital object.

[0080] The core of this A* algorithm is its evaluation function, which is used to evaluate the cost of a path from the initial state through an intermediate map state G n to the goal state:

[0081] f(G n ) = g(G n ) + h(G n );

[0082] where g(G n ) is the actual cumulative cost from the initial state G current to the intermediate state G n . This cost is the sum of the cost of all the atomic operations performed on the path. The cost of a single operation C op may be a comprehensive indicator, which can be calculated by the following formula:

[0083] C op = w t · Δt + w e · ΔE + w w · ΔW;

[0084] In this formula, Δt is the estimated execution time of the operation, ΔE is the estimated energy consumption of the operation, ΔW is the device wear factor set according to the device type and the nature of the operation, w t , w e , w w are preset weight coefficients used to adjust the proportion of each cost.

[0085] h(G n ) is the estimated cost from the intermediate state G n to the goal state, i.e. the heuristic function. An effective heuristic function is to calculate the number of goal predicates that have not been satisfied in state G n . For example, if the goal state set S goal has 5 predicates, and state G n only satisfies 3 of them, then a basic value of h(G n ) can be 2.

[0086] The execution flow of the dynamic path planning module 30 is as follows:

[0087] Initialize two sets: an open set (Open Set) for storing map states to be examined, and a closed set (Closed Set) for storing map states that have been examined. The initial state Gcurrent Put into open set.

[0088] Loop the following steps until the open set is empty or a solution is found:

[0089] a. Select the state node with the smallest f(G n ) value from the open set as the current node to be expanded.

[0090] b. Determine whether the current node satisfies all predicates in the target state set S goal . If yes, the search is successful, and the operation path is reconstructed by backtracking according to the parent-child relationship between nodes. If not, proceed to the next step.

[0091] c. Expand the current node. Traverse all available atomic operations F op of all device digital objects in the graph. During expansion, the module checks the state of the target device object v. If the current state of the finite state machine v.M state of the object is “fault” or “maintenance isolation”, all operations on the object are considered unavailable and do not participate in subsequent path expansion.

[0092] d. For each available atomic operation, calculate the new graph state (neighbor node) generated after the operation is performed. Calculate the cost from the current node to the neighbor node, and update its g(G n ) value and h(G n ) value.

[0093] e. Add the neighbor node to the open set, and move the current node to the closed set.

[0094] When the search is successful, the module reconstructs the state transition path into an ordered operation instruction sequence P = <Op1, Op2,..., Op m >. The sequence is the final planning result and is passed to the instruction execution module 40.

[0095] The instruction execution module 40 receives the operation instruction sequence P generated by the dynamic path planning module 30, and is responsible for converting it into specific control actions on physical devices, while ensuring the atomicity and recoverability of the operation sequence through a transaction management mechanism.

[0096] Before starting to execute the operation instruction sequence P, the instruction execution module 40 first creates a globally unique transaction ID, and initializes a transaction log L TX based on the ID. The transaction log can be a file or data table structure stored in non-volatile memory to ensure that the system can still recover data after unexpected power failure, etc.

[0097] Subsequently, the module traverses each atomic operation instruction Op in the instruction sequence P in order k For each instruction Op k , the module performs a Write-Ahead Logging operation before issuing the actual control signal to the DCS backend.

[0098] This Write-Ahead Logging operation includes first locating the target device digital object v k of the instruction Op k and taking a snapshot of its current state. This state snapshot contains at least the key variable parameter values in its dynamic attribute set P d and the current state of its finite state machine M state . Then a new log entry is created in the transaction log L TX . This entry contains the transaction ID, the instruction sequence number K, the instruction content Op k (e.g. the target device ID, the method name and parameters of the call), the taken state snapshot, and marks the execution status of this entry as “PENDING”.

[0099] After the log entry is successfully written and persisted, the instruction execution module 40 only then converts Op k into a DCS recognizable protocol format and sends it to the corresponding controller through its communication interface.

[0100] The module then waits for the execution feedback from the DCS. If a feedback indicating successful execution is received within a pre-set timeout, the module updates the status of the corresponding entry in the log L TX to “SUCCESS”. If a failure feedback is received or the timeout occurs, the status is updated to “FAILED” and the execution of the entire instruction sequence is immediately aborted, with the failure event and related information passed to the exception handling module 50.

[0101] If all instructions in the instruction sequence P are successfully executed, the instruction execution module 40 writes a “COMMIT” record at the end of the log L TX , marking the successful closure of this transaction. This transaction log can then be archived.

[0102] The exception handling module 50 is activated upon receiving the instruction execution failure event passed by the instruction execution module 40. The core function of this module is to restore the affected subsystem to a certain, safe state according to pre-set mechanisms.

[0103] In this embodiment, the exception handling module 50 is built-in with two different levels of safety rollback mechanisms: reverse state rollback and safety baseline rollback. The module will choose which mechanism to execute according to the risk level configuration of the task or the specific type of failure.

[0104] The reverse state rollback mode is the default rollback strategy. When this mode is triggered, the exception handling module 50 first locates and locks the corresponding transaction log L TX according to the transaction ID in the failure event.

[0105] Subsequently, the module traverses and executes rollback operations in reverse order according to the logic described in the following formula:

[0106]

[0107] where i is an integer index representing the sequence number of the instruction that failed in the entire sequence of operation instructions P = <Op1, Op2,..., Op m > ; k is a loop variable; Opkrepresents the kth atomic operation instruction in the sequence of operation instructions P; k represents the predefined inverse operation of the operation instruction Op k ; Execute(·) is an execution function, indicating that the operation inside the parentheses is actually executed by the underlying execution mechanism of the DCS.

[0108] The specific execution steps are as follows:

[0109] S1, initialize the loop variable k to i-1.

[0110] S2, find the log entry with sequence number k in the transaction log L TX .

[0111] S3, extract the original instruction content Op k from the entry.

[0112] S4, find the predefined inverse operation k of Op in the system. The inverse operation is defined together with the method set F op that defines the digital object of the device. For example, if the instruction Op k is an operation to start a device, then its corresponding inverse operation is an operation to stop the device. If the instruction Op k is an operation to set a device parameter to a certain value, then its inverse operation ​The parameter can be set to a pre-defined initial value, or the value of the parameter before the operation can be read from the state snapshot recorded in the log entry and restored.

[0113] S5, performing the reverse operation through the interface of the instruction execution module 40

[0114] S6, reducing the value of k by 1 and repeating steps S2-S5 until the value of k is less than 1.

[0115] Through the reverse execution process, the system state will be restored to the original state before the task started.

[0116] The safe baseline rollback mode is a more mandatory rollback strategy. It is triggered in the following cases:

[0117] The risk level of the task is pre-set as "high", in which case any failure directly triggers this mode.

[0118] The execution of a certain reverse operation in the reverse state rollback process also fails.

[0119] The transaction log is damaged or unreadable for some reason.

[0120] When the mode is triggered, the exception handling module 50 no longer relies on the transaction log. It will load one or a set of global, minimum safe instruction sequences associated with the current task from a pre-defined configuration library. The sequence does not aim to restore the original state, but to force the associated device or subsystem to a known, absolutely safe and stable state. For example, for a water supply subsystem, the safe baseline rollback sequence can be defined as: first, perform the stop operation of all associated water pumps; second, perform the closing operation of all associated outlet valves; third, perform the opening operation of all associated emptying valves. The sequence is directly handed over to the instruction execution module 40 for execution.

[0121] Although the embodiments of the present application have been shown and described, it is to be understood that various modifications, substitutions, replacements and changes can be made to these embodiments without departing from the principles and spirit of the present application, and the scope of the present application is defined by the appended claims and their equivalents.​

Claims

1. A DCS-based auxiliary control workshop control system, characterized in that: include: one or more processors; a memory having stored thereon instructions executable by one or more of the processors; When the instructions are executed by one or more processors, they are used to implement: The process topology map construction module is used to construct a dynamic process topology map containing multiple equipment digital objects based on the equipment in the auxiliary control workshop and their association relationships; A task parsing module, configured to parse the received high-level task instructions into a target state set of the dynamic process topology map; A dynamic path planning module, configured to dynamically plan and generate an operation instruction sequence for achieving the target state set based on the real-time state of the dynamic process topology map and the target state set; An instruction execution module, configured to execute instructions in the operation instruction sequence in sequence; The exception handling module is used to trigger a preset safety rollback mechanism when the instruction execution module fails to execute an instruction, so as to restore the equipment status of the auxiliary control workshop to a preset safe status.

2. The DCS-based auxiliary control workshop control system according to claim 1, characterized in that: Each device digital object in the dynamic process topology map corresponds to a physical device or logical entity, and encapsulates the object's static properties, dynamic properties, and a finite state machine that defines all its possible states and state transition conditions; the edges in the dynamic process topology map represent the physical or logical association relationship between device digital objects.

3. The DCS-based auxiliary control workshop control system according to claim 1, characterized in that: The dynamic path planning module generates the sequence of operation instructions by running a graph search algorithm on one or more processors, and during the search process, the dynamic path planning module is configured to automatically avoid equipment digital objects marked as "fault" or "maintenance isolation" status in the dynamic process topology map.

4. The DCS-based auxiliary control workshop control system according to claim 3 is characterized in that: The evaluation function used by the graph search algorithm during the search process is: f(G n )=g(G n )+h(G n ); Among them, f(G n ) is the estimated total cost from the initial state to the target state; g(G n ) is the transition from the initial state to the current intermediate graph state G n The actual cumulative cost of h(G n ) is the state from the current intermediate graph G n The estimated cost to reach the stated set of goal states.

5. The DCS-based auxiliary control workshop control system according to claim 1, characterized in that: Before executing the operation instruction sequence, the instruction execution module is configured to first create a transaction log for this task; when executing each instruction in the sequence, the instruction content and the state snapshot of the target device digital object before execution are recorded in the transaction log.

6. The DCS-based auxiliary control workshop control system according to claim 5, characterized in that: The safe rollback mechanism triggered by the exception handling module includes at least one of the following modes: Reverse state rollback: When the instruction execution fails, the exception handling module reads the transaction log and reversely performs the reverse operation of the successful operation to restore the device state to the state before the task started; Safety baseline rollback: When reverse state rollback cannot be performed or the task risk level is high, the exception handling module executes a globally predefined minimum safety instruction sequence to force the associated subsystem to recover to a known minimum safety and stable state.

7. The DCS-based auxiliary control workshop control system according to claim 6, characterized in that: When executing the reverse state rollback, the exception handling module is configured to: for the failed instruction Op recorded in the transaction log i Each instruction Op that has been successfully executed before k , both perform their predefined inverse operations The process can be described as: Where i is an integer index, representing the number of <Op1,Op2,...,Op m >, the number of the instruction that failed to execute; k is a loop variable; Op k Represents the kth atomic operation instruction in the operation instruction sequence P; Represents the operation instruction Op k Execute(·) is an execution function, which means sending the operation in the brackets to the bottom execution mechanism of the DCS for actual execution.

8. The DCS-based auxiliary control workshop control system according to claim 1, characterized in that: It also includes a virtual execution module; the virtual execution module is configured to: before the instruction execution module sends an operation instruction sequence to the physical device, first rehearse the sequence on a copy of the dynamic process topology map to verify its logical correctness and pre-analyze the possible chain effects.

9. The DCS-based auxiliary control workshop control system according to claim 1, characterized in that: The task parsing module is used to parse the high-level task instructions into the target state set, wherein the target state set is composed of one or more target predicates, each target predicate being a constraint on the desired state of a device digital object or edge in the dynamic process topology map.

10. A DCS-based auxiliary control workshop control method, characterized in that: The method is used to implement a DCS-based auxiliary control workshop control system according to any one of claims 1 to 9, comprising the following steps: Constructing a dynamic process topology map, wherein the dynamic process topology map includes multiple equipment digital objects and reflects the equipment in the auxiliary control workshop and their association relationships; Parsing the received high-level task instructions into a target state set of the dynamic process topology map; Dynamically planning and generating an operation instruction sequence according to the real-time state of the dynamic process topology map and the target state set; Execute the instructions in the operation instruction sequence in sequence; When the execution of an instruction fails, the preset safety rollback mechanism is triggered to restore the equipment status of the auxiliary control workshop to a preset safe state.

Images