Process planning method for multi-objective requirements

Abstract

The present application relates to the technical field of test scheduling, and particularly relates to a process flow planning method for multi-object demand, comprising the following steps: extracting test demand and test equipment information based on an expert rule base, and constructing a standardized test demand model; constructing a test equipment resource data representation model, including a hierarchical dependency model of main equipment resources and a digital coding model of process valves and pipelines; constructing a mapping model of test demand and test equipment, and matching between test demand and equipment resource combination templates; constructing a multi-tester process system resource configuration conflict matrix, and real-time detecting resource conflicts of equipment, valves and pipelines; generating a process parameter sample set based on Latin hypercube sampling, and planning and optimizing the process flow; through the above manner, global quantitative representation is carried out, so as to real-time calculate conflict intensity, conflict position and resolution priority, thereby achieving the effect of supporting safe scheduling of large-scale concurrent tests.

Description

Technical Field

[0001] This invention relates to the field of test scheduling technology, and in particular to a process flow planning method oriented towards the needs of multiple objects. Background Technology

[0002] Current large-scale aerodynamic test support systems are generally designed to conduct test tasks in parallel with multiple test objects, multiple test categories, and multiple test instruments. The test requirements are characterized by high frequency, strong constraints, and tight parameter coupling. The process flow planning directly determines the utilization rate of test resources, test safety, and execution efficiency.

[0003] In existing technologies and engineering practices, resource conflicts are difficult to automatically identify and resolve in scenarios involving multiple testing equipment operating concurrently, and process flows cannot be automatically planned. Resource allocation conflicts among multiple testing equipment rely on manual investigation, lacking quantification and automatic resolution mechanisms. When multiple testing equipment simultaneously use gas, share the main gas supply line and key equipment, resource competition exhibits characteristics such as spatiotemporal coupling, strong concealment, and wide propagation. Examples include: the same unit / valve being occupied by multiple test tasks, pipeline flow exceeding limits, pressure fluctuations exceeding the safety envelope, and mutually exclusive flow directions leading to gas path turbulence. Existing technologies lack a unified resource allocation conflict matrix, making it impossible to quantitatively represent equipment occupancy, valve status, pipeline connectivity, and flow direction constraints across the entire domain. Conflict identification relies on manual experience to check each task individually, failing to calculate conflict intensity, conflict location, and resolution priority in real time, thus hindering the safe scheduling of large-scale concurrent tests. Summary of the Invention

[0004] The purpose of this invention is to provide a process planning method for multiple object requirements, which aims to solve the technical problems in the prior art, such as the inability to fully quantify the representation of equipment occupancy, valve status, pipeline connectivity, and flow direction constraints, the reliance on manual experience for task-by-task conflict identification, the inability to calculate conflict intensity, conflict location and resolution priority in real time, and the difficulty in supporting safe scheduling for large-scale concurrent experiments.

[0005] To achieve the above objectives, the present invention employs a process flow planning method oriented towards the needs of multiple objects, comprising the following steps:

[0006] A standardized test requirement model is constructed by extracting test requirements and test equipment information from an expert rule base.

[0007] Establish a data representation model for experimental equipment resources, including a hierarchical dependency model of main equipment resources and a digital coding model of process valves and pipelines;

[0008] Construct a mapping model between test requirements and test equipment, and match test requirements with equipment resource combination templates;

[0009] Construct a resource allocation conflict matrix for a multi-tester process system to detect resource conflicts among equipment, valves, and pipelines in real time;

[0010] A process parameter sample set is generated based on Latin hypercube sampling, and the process flow is planned and optimized.

[0011] Among them, in the step of extracting test requirements and test equipment information based on the expert rule base and constructing a standardized test requirements model:

[0012] Establish an expert rule base, define the attributes of the test objects, divide the test objects into overall and layout test types, and map and match overall and local core resources;

[0013] Extract the core parameters and time constraints of the experiment, quantify the key aerodynamic parameters of temperature, pressure, and flow rate, and determine the expected start time and duration of the experiment.

[0014] After extracting the core experimental parameters and time constraints, quantifying key aerodynamic parameters such as temperature, pressure, and flow rate, and determining the expected start time and duration of the experiment:

[0015] Based on the range of core parameter combinations, determine the category of derived experiments and complete the standardized data modeling and digital coding of the experimental requirements.

[0016] Among the steps in establishing a data representation model for experimental equipment resources, including a hierarchical dependency model of the main equipment resources and a digital coding model of process valves and pipelines:

[0017] Digital codes are used to identify the in-use, out-of-use, and maintenance status of the main equipment of the compressor unit, heating furnace, and drying unit, forming a unified status identifier.

[0018] A hierarchical dependency model for main equipment is constructed, abstracting resources into basic equipment units and composite equipment systems. The basic units record unique numbers, technical parameters and operating status, while the composite systems dynamically calculate the overall aggregation status based on the bottleneck effect.

[0019] Establish a comprehensive data model for the equipment to quantify its physical operational capabilities, energy consumption characteristics, safe operating range, and lifecycle status.

[0020] This includes establishing a comprehensive data model of the equipment to quantify its physical operational capabilities, energy consumption characteristics, safe operating range, and lifecycle status:

[0021] Digital modeling is performed on process valves and pipelines, with valves set as real nodes and pipeline intersections as virtual nodes. Matrix coding is used to represent valve opening and closing states and pipeline conflict constraints.

[0022] Among them, in the steps of constructing a mapping model between test requirements and test equipment, and matching test requirements with equipment resource combination templates:

[0023] Collect historical test data, convert equipment parameters and environmental indicators into unified feature vectors, and complete data preprocessing and standardization;

[0024] Cluster the experimental requirements data and optimize the initial cluster center selection method and similarity calculation method;

[0025] The clustering results are divided into typical test condition clusters, and each cluster is bound to a corresponding optimal equipment resource combination template.

[0026] After dividing the clustering results into typical test condition clusters and binding the corresponding optimal equipment resource combination template to each cluster:

[0027] Once new test requirements are input, the similarity with each operating condition cluster is calculated, the requirements are assigned to the optimal cluster, and the corresponding equipment resource template is called.

[0028] Among the steps involved in constructing a resource allocation conflict matrix for a multi-tester process system and real-time monitoring of resource conflicts among equipment, valves, and pipelines:

[0029] The test pipeline system is converted into a topology diagram composed of nodes and pipes, and the key resource sets of equipment, valves, and pipes are divided.

[0030] Construct a multidimensional heterogeneous resource conflict matrix and use different values ​​to identify resource non-conflict, self-occupation, strong conflict and soft competition relationships;

[0031] Perform flow path pre-simulation to complete comprehensive conflict detection of physical mutual exclusion, parameter boundaries, and medium flow direction.

[0032] After performing flow path pre-simulation and completing comprehensive conflict detection of physical mutual exclusion, parameter boundaries, and medium flow direction:

[0033] The conflict matrix is ​​updated in real time, and the matrix elements are dynamically corrected according to changes in equipment operating status, forming a dynamic evolution and feedback mechanism.

[0034] Among them, in the step of generating a process parameter sample set based on Latin hypercube sampling and planning and optimizing the process flow:

[0035] Define the process parameter space and select resource allocation weight, execution time period, and flow adjustment range as key sampling dimensions;

[0036] Generate process parameter samples covering the entire domain, and retain feasible samples through dual filtering using rules and conflict matrices;

[0037] Feasibility assessments were conducted on candidate process routes to select the globally optimal process flow with short response time and balanced energy consumption.

[0038] This invention provides a process planning method oriented towards multiple object requirements. It extracts test requirements and test equipment information from an expert rule base to construct a standardized test requirement model; establishes a test equipment resource data representation model, including a hierarchical dependency model of main equipment resources and a digital coding model of process valves and pipelines; constructs a mapping model between test requirements and test equipment to match test requirements with equipment resource combination templates; constructs a resource configuration conflict matrix for a multi-tester process system to detect resource conflicts of equipment, valves, and pipelines in real time; generates a process parameter sample set based on Latin hypercube sampling and performs process planning and optimization; and through the above methods, performs global quantization representation to calculate conflict intensity, conflict location, and resolution priority in real time, thereby achieving the effect of safe scheduling to support large-scale concurrent tests. Attached Figure Description

[0039] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0040] Figure 1 This is a flowchart of the process planning method for multi-object needs according to the present invention.

[0041] Figure 2 This is a flowchart of steps S100 of the present invention.

[0042] Figure 3 This is a flowchart of steps S200 of the present invention.

[0043] Figure 4 This is a flowchart of steps S300 of the present invention.

[0044] Figure 5 This is a flowchart of steps S400 of the present invention.

[0045] Figure 6 This is a flowchart of steps S500 of the present invention. Detailed Implementation

[0046] Exemplary embodiments will now be described in detail, examples of which are illustrated in the accompanying drawings. When the following description relates to the drawings, unless otherwise indicated, the same numbers in different drawings represent the same or similar elements. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with this application.

[0047] The terminology used in this application is for the purpose of describing particular embodiments only and is not intended to be limiting of the application. The singular forms “a,” “the,” and “the” used in this application and the appended claims are also intended to include the plural forms unless the context clearly indicates otherwise. It should also be understood that the term “and / or” as used herein refers to and includes any or all possible combinations of one or more of the associated listed items.

[0048] Please see Figures 1-6 This invention provides a process flow planning method oriented towards the needs of multiple objects, comprising the following steps:

[0049] S100: Extract test requirements and test equipment information based on the expert rule base to construct a standardized test requirements model.

[0050] In this implementation, a standardized test requirement model is constructed by extracting test requirements and test equipment information from an expert rule base. The specific process is as follows:

[0051] S101: Establish an expert rule base, define the attributes of the test objects, divide the test objects into whole and local types, and map and match the whole and local core resources;

[0052] S102: Extract the core parameters and time constraints of the test, quantify the key aerodynamic parameters of temperature, pressure, and flow rate, and determine the expected start time and duration of the test;

[0053] S103: Determine the category of derived tests based on the range of core parameter combinations, and complete the standardized data modeling and digital coding of test requirements.

[0054] In the aforementioned process, the expert rule base was developed by experts in the testing field based on years of experience, covering common test requirement expression standards, key information judgment criteria, and test equipment status identification rules. For the extraction of specific test equipment parameters, the rule base specifies standard formats for parameter names, units, and value ranges, ensuring that the extracted test requirement information is accurate and conforms to industry standards. To optimize the test system scheduling and maximize resource utilization, the primary task is to extract and integrate test requirement and test plan data based on the expert rule base. After screening and correction, and according to the different status expression rules of test requirements and test equipment, digital coding work is carried out to systematically and standardizedly model the complex and diverse test requirements.

[0055] This study explores the construction of a multi-dimensional experimental requirement data model, transforming descriptive experimental tasks into structured data that can be analyzed and processed by algorithms. The core of this model lies in abstracting each experimental requirement into a unified data object, which comprehensively encapsulates all the key attributes of the experimental task. The key implementation steps are as follows:

[0056] Test object attribute definition: The test objects are clearly divided into types such as "whole" and "part", and based on this classification, they are directly mapped to the core resources such as "whole" or "part" on which they depend, thereby establishing a preliminary relationship between needs and resources.

[0057] Core parameter attribute definition: The core aerodynamic parameters involved in the experiment, namely temperature (T), pressure (P) and flow rate (Q), are precisely quantified.

[0058] Key time constraint attribute definition: A clear time constraint dimension is introduced to precisely define the user's desired experiment execution time window. This window is defined by an "expected start time" and a "duration," and is the primary basis for the scheduling algorithm to arrange the timeline and determine feasibility.

[0059] Derived Test Category Attribute Definition: To facilitate the subsequent scheduling algorithm's classification and batch processing of similar tasks, a "test category" is introduced as a derived attribute. This attribute is not directly specified by the user, but is automatically generated based on a set of preset rules and the combination range of the three core parameters submitted by the test object: temperature, pressure, and flow rate. When the test requirement parameter is: temperature greater than the high-temperature threshold, the system can automatically classify it as a "high-temperature test." Based on the rule definition, the system can also determine whether the input exceeds the system's design capability indicators and provide feedback to the user. This mapping relationship from underlying parameters to upper-level categories achieves the standardization of requirements and simplifies complex parameter combinations into a limited, clearly defined test mode. Table 1 shows an example of a partial test category determination rule matrix for the overall system.

[0060]

[0061] Table 1. Examples of Partial Test Category Judgment Rule Matrix Oriented to the Whole

[0062] By integrating basic information, experimental object attributes, core parameters, key time constraints, and derived experimental categories into a unified data structure, a complete and standardized experimental requirement model is constructed. This model provides a solid data foundation for subsequent optimal experimental path planning, resource conflict detection, and dynamic scheduling adjustments, and is a fundamental prerequisite for achieving intelligent and efficient operation of the entire experimental system. Table 2 illustrates this data structure.

[0063]

[0064] Table 2. Standardized Modeling of Experimental Requirements

[0065] For the front-end air supply and air handling system of the test apparatus, based on the analysis of the test process, the test requirements are extracted according to three factors: flow rate, temperature, and pressure. Taking high-altitude chamber testing as an example, the test types are mainly divided into five categories: low-pressure direct air supply test, medium-pressure direct air supply test, dry air supply test, negative temperature air supply test, and high temperature air supply test. The test requirements and test equipment information configuration rules will be described in detail below in conjunction with the five types of test attributes.

[0066] Low-pressure direct gas supply test:

[0067] 1. Traffic is used to configure core test equipment resources;

[0068] 2. Temperature-based indicators for different types of direct gas supply tests;

[0069] 3. Pressure Differentiation Indicators for Low-Pressure Test Types. Table 3 shows the requirements and equipment configuration rules for low-pressure direct gas supply tests from atmospheric sources.

[0070]

[0071] Table 3. Rules for the Configuration of Test Requirements and Equipment for Low-Pressure Direct Gas Supply from Atmospheric Sources

[0072] Medium-pressure direct gas supply test:

[0073] 1. Traffic is used to configure core test equipment resources;

[0074] 2. Temperature-based indicators for different types of direct gas supply tests;

[0075] 3. Pressure Differentiation Indicators for Medium-Pressure Test Types. Table 4 shows the requirements and equipment configuration rules for medium-pressure direct gas supply tests from atmospheric sources.

[0076]

[0077] Table 4. Rules for the Configuration of Test Requirements and Equipment for Medium-Pressure Direct Gas Supply from Atmospheric Sources

[0078] Drying gas supply test

[0079] 1. Traffic is used to configure core test equipment resources;

[0080] 2. Temperature-based indicators for different types of drying gas supply tests;

[0081] 3. Pressure varies depending on the equipment configuration type; 25-40 atmospheres of air intake are mixed with drying air. Table 5 shows the drying air supply test requirements and equipment configuration rules.

[0082]

[0083] Table 5. Rules for Drying Gas Supply Test Requirements and Test Equipment Configuration

[0084] Negative temperature gas supply test

[0085] 1. Traffic is used to configure core test equipment resources;

[0086] 2. Temperature differentiation indicators for negative temperature gas supply test types;

[0087] 3. Pressure is differentiated based on equipment configuration type. Table 6 shows the configuration rules for negative temperature gas supply test requirements and test equipment information.

[0088]

[0089] Table 6. Rules for Configuration of Test Equipment and Requirements for Negative Temperature Gas Supply Test

[0090] High-temperature gas supply test

[0091] 1. Traffic is used to configure core test equipment resources;

[0092] 2. Temperature differentiation indicators for negative temperature gas supply test types;

[0093] 3. Pressure is differentiated based on equipment configuration type. Table 7 shows the configuration rules for high-temperature gas supply test requirements and test equipment information.

[0094]

[0095] Table 7. Rules for High-Temperature Gas Supply Test Requirements and Equipment Configuration

[0096] S200: Establish a data representation model for experimental equipment resources, including a hierarchical dependency model of main equipment resources and a digital coding model of process valves and pipelines.

[0097] In this embodiment, a data representation model for experimental equipment resources is established, including a hierarchical dependency model of the main equipment resources and a digital coding model of process valves and pipelines. The specific process is as follows:

[0098] S201: Digital coding of the main equipment of compressor units, heating furnaces, and drying units to form a unified status identifier for their in-use, out-of-use, and maintenance status.

[0099] S202: Construct a hierarchical dependency model for the main equipment, abstracting resources into basic equipment units and composite equipment systems; among them, basic units record unique numbers, technical parameters and operating status, and composite systems dynamically calculate the overall aggregation status based on the bottleneck effect;

[0100] S203: Establish a full-dimensional data model for equipment to quantify the equipment's physical operation capabilities, energy consumption characteristics, safe operating range, and life cycle status;

[0101] S204: Digitally model process valves and pipelines, setting valves as real nodes and pipeline intersections as virtual nodes, and using matrix coding to represent valve opening and closing states and pipeline conflict constraints.

[0102] In the above process, based on the different functional classifications of the test equipment, the current status of the compressor unit, turbine, air wave machine, heating furnace, and other equipment—whether in use, out of use, or undergoing commissioning or maintenance—is coded separately. These different statuses are coded as 1, 0, and -1, respectively, as shown in Table 8. Subsequently, the digitally coded test requirements and test equipment are stored in a structured manner to facilitate subsequent querying, statistics, and analysis, providing a solid data foundation for the formulation of test plans.

[0103]

[0104] Table 8. Examples of digital codes for test equipment status

[0105] To achieve refined management and automated scheduling of experimental system resources, a hierarchical dependency model is constructed to systematically and digitally model all equipment resources. The core idea of ​​this method is to deconstruct the complex experimental system into logically clear and computable objects. This not only characterizes the independent attributes of individual devices but also profoundly reveals the structural relationships between the whole and its parts, as well as the crucial state dependencies, providing a solid and reliable data foundation for intelligent scheduling algorithms. Within this model framework, all physical resources are abstracted into two core types of objects: basic equipment units and composite equipment systems.

[0106] Basic equipment units are the smallest indivisible functional entities that constitute a system, i.e., "leaf nodes" in the model hierarchy tree, such as a single valve or an independent heating furnace. Each basic unit is assigned a globally unique equipment number and its inherent technical specifications (such as rated power, pressure rating, flow range, etc.) and real-time operating status (such as normal, faulty, occupied, under maintenance) are accurately modeled. These basic units form the foundation of the entire resource model, and their detailed parameters and dynamic states are the fundamental basis for the upper-level system to conduct capability assessments and availability judgments.

[0107] Composite systems: As branches in the model hierarchy tree, they represent a logical collection of multiple basic units or other subsystems designed to achieve a specific function, such as a complete atmospheric source system or a natural circulating water system. The key to modeling composite systems lies in clearly defining three core attributes:

[0108] First is the list of subsidiary systems, which explicitly lists the unique numbers of all sub-resources directly belonging to this system, clearly presenting the equipment composition of the system.

[0109] Secondly, there are System Specs, which represent the calibrated or calculated effective performance envelope that the system as a whole provides to the outside world, taking into account the performance bottlenecks and losses of all internal auxiliary devices.

[0110] Finally, there is the aggregate state, which is not inherent but dynamically calculated based on the real-time state of all units in its subordinate systems and following the principle of the weakest link effect. Any abnormal state of any critical sub-component will cause the entire composite system to be judged as unusable.

[0111] This hierarchical modeling approach, moving from units to systems and from the bottom layer to the top layer, transforms a static and complex network of physical devices into a dynamic data model with a clear structure, rigorous logic, and real-time state projection capabilities. This model provides the scheduling engine with a multi-resolution view, enabling it to quickly match system-level aggregation capabilities at a macro level, and also drill down to the device component level as needed to verify the current state of any basic component. This provides core technical support for achieving high-precision, high-reliability automated resource scheduling and optimization.

[0112] To address the challenges of diverse equipment types, heterogeneous data formats, and varied operating conditions in experimental systems, this paper proposes a general unified coding and modeling method for equipment resources. This method is based on experimental requirements analysis and feature extraction results using an expert rule base, combined with a hierarchical dependency model of equipment resources in the experimental system. The constructed equipment data model adopts an attribute-value semi-structured data paradigm to digitally map all elements of physical equipment entities.

[0113] The device data model specifically includes the following four key dimensions of feature abstraction:

[0114] Quantitative Characterization of Physical Operation Capacity and Energy Consumption Characteristics: To address the issue of accurately matching experimental tasks with equipment capabilities, the model parameterizes the core operational indicators of the equipment, constructing the output capacity boundary using flow limit and export pressure limit. The rated power and water consumption rate fields describe the equipment's resource consumption characteristics. For example, the model explicitly defines the rated power as 16000kW and the water consumption as 1182m³. 3 / h, this data will serve as a key input variable for the energy consumption cost function in the multi-objective optimization scheduling algorithm.

[0115] Definition of the safety envelope for operating conditions: Considering the test temperature requirements and safety requirements, the model introduces a thermodynamic constraint mechanism. Through the labels of the lower and upper limits of the outlet temperature (OutletTemperatureLow and OutletTemperatureHigh), the temperature range for normal equipment operation is clearly defined. This serves as a quantitative basis for resource feasibility verification, and also as a decision threshold for anomaly detection and alarm mechanisms in the process flow, ensuring that temperature indicators in key areas of the test are under control.

[0116] Discrete-space mapping of equipment lifecycle states: A two-layer state model, encompassing operational and health states, is constructed to address the dynamic changes in equipment states throughout the testing cycle. The operational layer uses binary coding to describe the real-time occupancy of the equipment, where "0" represents idle / disabled and "1" represents in use / running, supporting concurrent control and conflict detection of testing resources. The health layer uses ternary discrete coding to finely characterize the equipment's operational attributes. "0" represents normal availability, "-1" indicates fault shutdown, and "1" indicates maintenance. This coding method enables the system to quickly identify available nodes in the resource pool and provides status markers for fault tracing and maintenance scheduling.

[0117] Semantic classification and visualization topology mapping: To achieve intuitive presentation of the human-computer interaction interface and rapid indexing of resources, the model integrates semantic identifiers and spatial attributes, and presents them in a concrete form as field labels.

[0118] The category field: assigns a clear ontological category label to the device, which facilitates class attribute reasoning by the rule base;

[0119] The id field serves as a globally unique identifier to enable precise indexing of entities.

[0120] The pos field is the normalized coordinate vector of the storage device in the monitoring topology diagram. It supports dynamic rendering of node positions in the visualization component, realizing an intuitive mapping from physical topology to logical view.

[0121] To enable efficient data manipulation during the status monitoring and updating phase of the experiment, a simple and intuitive coding method is adopted to facilitate rapid identification and processing of valve status information by the computer system. Since each pipeline connects two valves, each control valve in the pipeline is considered a real node, and each pipeline intersection is considered a virtual node. 0 and 1 represent the on / off state of the real nodes, and their flow connection relationship is used as adjacent nodes. A pair of adjacent nodes is considered a pipeline. This method comprehensively captures the status of control valves during the dynamic changes of multiple testers in parallel experiments, and the digital coding comprehensively represents the status of the pipeline. The complex relationships between valves at each control node. Rows and columns correspond one-to-one. This symmetrical structure of individual valves helps to uniformly handle the interrelationships between them, whether in their own state or in situations of connection and conflict with other valves. For elements... :

[0122] when hour, As a valve The valve's own status is identified by binary codes 0 and 1, which clearly indicate the valve's closed and open states, as shown in Table 9.

[0123] when At that time, judge the valve With valve The relationship between them requires in-depth analysis of the pipeline system's topology and experimental process.

[0124] If valve With valves The valves are directly connected and, under specific test conditions, their on / off states are mutually constrained. During a certain test phase, the valves... Valve when open It must be closed. A value of -1 indicates a conflict.

[0125] If, after a comprehensive traversal and analysis of the pipeline network connection diagram, it is determined that there is no direct conflict between the two, then .

[0126] When considering the flow direction of valves, different coding methods are used to distinguish between forward and reverse connections and conflict situations in order to more accurately describe the complex relationships between valves. If the valve... To valve There is a positive conflict, that is, when the medium flows from the valve Flow to valve At this time, the switching states of the two are constrained. .

[0127]

[0128] Table 9. Modeling of Valve and Piping Data for Experimental Plan Nodes

[0129] S300: Construct a mapping model between test requirements and test equipment, and match test requirements with equipment resource combination templates;

[0130] In this embodiment, a mapping model between test requirements and test equipment is constructed, and the test requirements are matched with equipment resource combination templates. The specific process is as follows:

[0131] S301: Collect historical test mission data, convert equipment parameters and environmental indicators into a unified feature vector, and complete data preprocessing and standardization;

[0132] S302: Perform clustering processing on the experimental requirement data, and optimize the initial cluster center selection method and similarity calculation method;

[0133] S303: Divide the clustering results into typical test condition clusters and bind the corresponding optimal equipment resource combination template to each test condition cluster;

[0134] S304: After new test requirements are input, calculate the similarity with each working condition cluster, assign the requirements to the optimal cluster, and call the corresponding equipment resource template.

[0135] In the above process, a wealth of historical data on experimental tasks is collected, such as the objectives of the experimental tasks, the experimental equipment involved, and the experimental environment requirements. This data is preprocessed and converted into feature vector representations suitable for cluster analysis.

[0136] For experimental equipment information, equipment type, equipment parameter range, etc. are converted into numerical features;

[0137] For the test environment requirements, environmental indicators such as pressure, flow rate, and temperature are quantified.

[0138] Based on the information extraction results of the association rules between test categories and equipment resources, the tests are clustered into K typical standard test condition clusters, such as high-temperature gas supply mode and negative-temperature drying gas supply mode. Each cluster center This represents the standard parameter configuration for this type of test.

[0139] For each generated cluster By combining expert knowledge base, the optimal equipment resource combination template is bound to it. .

[0140] When a new experimental requirement x arrives, instead of performing tedious rule-based reasoning, its relationship with the cluster centers is calculated. The Mahalanobis distance is used to assign it to the nearest cluster. .

[0141]

[0142] Call the resource template corresponding to this cluster As the initial recommended option.

[0143] The K-Means clustering algorithm is used to perform cluster analysis on the test task data to explore and construct the mapping relationship between test requirements and test equipment. For different test objects such as local and overall, the mapping relationship can be used to quickly determine the test category and the test equipment resources such as gas supply and air handling systems that need to be used, as well as the main gas supply pipeline required for the test, and further pinpoint the specific test equipment objects.

[0144] S400: Constructs a resource allocation conflict matrix for multi-tester process systems to detect resource conflicts of equipment, valves, and pipelines in real time.

[0145] In this embodiment, a resource allocation conflict matrix for a multi-tester process system is constructed to detect resource conflicts among equipment, valves, and pipelines in real time. The specific process is as follows:

[0146] S401: Convert the test pipeline system into a topology diagram composed of nodes and pipes, and divide the key resource sets of equipment, valves, and pipelines;

[0147] S402: Construct a multidimensional heterogeneous resource conflict matrix, using different values ​​to identify resource non-conflict, self-occupation, strong conflict, and soft competition relationships;

[0148] S403: Perform flow path pre-simulation to complete comprehensive conflict detection of physical mutual exclusion, parameter boundaries, and medium flow direction;

[0149] S404: Real-time update of the conflict matrix, dynamically correcting matrix elements according to changes in equipment operating status, forming a dynamic evolution and feedback mechanism.

[0150] In the above process, the control node valve-pipeline state conflict matrix is ​​constructed based on topology analysis: the topological relationship between the control node valve and the pipeline network determines the logical correctness of the medium flow. Using matrix framework thinking, a node-link association model is established.

[0151] The complex test pipeline system is highly abstracted as a weighted directed graph in graph theory.

[0152] The conflict matrix is ​​G = (V, E, W).

[0153] in, The set represents a collection of nodes, which includes both physical nodes (real nodes representing valves) and virtual nodes (merging and branching points). For each node... Define its state variables These correspond to the off and on states, respectively. This represents the set of edges, which is the physical pipeline connecting two nodes. It is a set of weights for each edge. Its weight vector It includes the maximum flow rate, maximum pressure capacity, and operating temperature range of the pipe section.

[0154] The conflict matrix acts as a bridge connecting the physical layer and the logical scheduling layer. Suppose the system has M concurrent experimental tasks T = {T1, T2, ..., TM}, involving N critical resources. The conflict matrix C has dimensions N×N, and its elements c ij The constraints between resource i and resource j are quantified. Based on the coding standards mentioned in the document, the logic for determining the values ​​of the conflict matrix elements is defined as follows:

[0155]

[0156] After constructing static and dynamic matrix models, a domain-knowledge-guided heuristic traversal algorithm is employed. This algorithm integrates the connectivity detection capability of depth-first search with the hierarchical expansion characteristics of breadth-first search, and incorporates a physics-based pruning strategy. The input to the algorithm is the current experimental plan sequence P. plan and real-time device state vector S curr The output is the conflict set S. etconflict .

[0157] Initialization and state mapping: each step in the test plan is executed;

[0158] Flow path simulation: For each experimental task T k The algorithm starts from the source node and performs a constrained Depth-First Search (DFS) traversal to find a feasible path to the target node of the test chamber. During the traversal, a temporary state mask M is maintained. temp .

[0159] Heuristic collision detection: When the algorithm attempts to access node v next At that time, not only is its static connection checked, but the heuristic function H(v) is also called. next ,S global Dynamic decision-making is performed. The heuristic function H contains the following sub-logic:

[0160] Physical mutual exclusion detection: Query the conflict matrix C. If the current path already has an active node v. curr And c(v) curr ,v next If ) = −1, then it is determined to be a hard conflict.

[0161] Parameter boundary verification: Calculate the cumulative traffic demand ∑Q and pressure setting P for the current path. set If ∑Q>Q max (pipeline) or P set >P limit If (valve) is not found, it is determined to be a parameter overflow conflict.

[0162] Flow consistency check: If task A requires pipeline segment e xy If the flow direction is x→y, but concurrent task B requires the flow direction to be y→x, then a reverse flow conflict is triggered.

[0163] Backtracking and Pruning: Once a conflict is detected, record the conflict type and involved resources, and attempt to backtrack to find an alternative path. If all paths are blocked, report it as infeasible. Utilize the path matrix assignment logic in graph theory to quantitatively fill the elements of the conflict matrix. 1 represents a strong conflict, and 0 represents complete decoupling and parallelizability.

[0164] Parameter sensitivity modeling: In addition to topology, the matrix also incorporates the pipeline's flow limit, maximum pressure, and operating temperature threshold as dynamic constraints. By effectively capturing the nonlinear superposition effect of flow and temperature, the matrix can identify pressure fluctuation conflicts in the main pipe caused by simultaneous gas intake / exhaust from multiple test chambers, thereby ensuring a safety margin for the planned path at the physical level.

[0165] Construction of a Feasibility Analysis Model for Key Test Equipment Resources: For core test equipment resources such as compressor units and drying units, a state conflict matrix is ​​constructed based on state-space description. From the perspective of multi-dimensional state-space representation, a feasibility assessment model for key equipment resources based on matrix superposition is established. This model comprehensively analyzes the operational envelope of key equipment resources such as compressor units, using different operating modes of the equipment as the rows and columns of the matrix. Through in-depth mining of equipment coding information, energy and load conflicts under different state combinations are quantified. In the assessment model, the test plan-test requirement association rules are mapped to the conflict matrix. By calculating the rank, eigenvalues, and modal overlap of the matrix, the capability redundancy of key equipment resources under different test conditions is accurately quantified. This model can capture hidden conflicts that are difficult to detect using traditional analysis methods, providing multi-criteria decision support for the scientific selection of test paths.

[0166] A dynamic evolution and feedback mechanism for the conflict matrix under cyber-physical fusion: When the state of a physical entity shifts, a sliding window algorithm and an incremental matrix update algorithm are used to promptly correct the corresponding element values ​​in the conflict matrix. This dynamic mapping mechanism ensures that the conflict matrix is ​​not only a static blueprint but also a dynamic digital twin of the experimental system. With the continuous accumulation of experimental data, machine learning algorithms are introduced to deeply mine equipment operating patterns. Through retrospective analysis of historical conflict cases, the matrix's dimensional division rules and conflict assessment weights are periodically adaptively adjusted. For example, if the failure rate of a certain type of valve increases under specific pressures, the conflict matrix will automatically increase its conflict coefficient under relevant operating conditions. Through this continuous evolution and optimization, the accuracy and practicality of the conflict matrix will continuously improve with the increase in experimental frequency, significantly enhancing the system's ability to control extreme experimental conditions.

[0167] S500: Generates a sample set of process parameters based on Latin hypercube sampling, and plans and optimizes the process flow.

[0168] In this embodiment, a process parameter sample set is generated based on Latin hypercube sampling, and the process flow is planned and optimized. The specific process is as follows:

[0169] S501: Define the process parameter space and select resource allocation weight, execution time period, and flow adjustment range as key sampling dimensions;

[0170] S502: Generate process parameter samples covering the entire domain, and retain feasible samples through dual filtering of rules and conflict matrix;

[0171] S503: Conduct a feasibility assessment of candidate process paths and select the globally optimal process flow with short response time and balanced energy consumption.

[0172] In the above process, the performance of process parameter spatial characterization and sample generation test plan scheduling based on Latin Hypercube Sampling (LHS) is highly dependent on key parameters such as resource allocation ratio, timing arrangement and execution intensity.

[0173] Define the process parameter space S⊂RD. Select the resource allocation weight of the test instrument, the core execution period, and the medium flow rate adjustment range as key dimensions to construct the state vector x:

[0174] ;

[0175] in:

[0176] ω=[ω1,ω2,...,ω m ] T [0,1] mLet ∑mω represent the resource allocation weight vector for m concurrent experimental tasks and satisfy the normalization constraint. i =1.

[0177] τ=[τ start ,τ duration ] T This indicates the start time and duration of the core execution period.

[0178] Q = [qmin, qmax] T This indicates the dynamic adjustment range constraint of the medium flow rate.

[0179] The LHS algorithm is introduced to perform multi-dimensional coverage of the process space. The resource allocation weight ω of the test instrument is selected. [0,1], Core execution period T exec The sampling dimension is input including the medium flow rate adjustment range, etc. The key scheduling parameters of the test system are defined as the dimensions of the sampling space. Assume there are K key parameters, and N samples need to be generated to cover all possible combinations of operating conditions. Define the parameter space Ω⊂R. K The value range of the k-th dimension parameter xk is [L k U k ].

[0180] The steps for generating the LHS algorithm are as follows:

[0181] 1. Interval partitioning: Divide the range of values ​​[0,1] (after normalization) of each dimension k into N equal-probability, non-overlapping sub-intervals. ,i=1,...,N.

[0182] 2. Random Selection: Independently and randomly select a point within each sub-interval. For the i-th interval in the k-th dimension, the selected value ξ... i,k ~U( ).

[0183] 3. Random Permutation: To eliminate correlation between dimensions, the N sampling points in each dimension are randomly permuted. Let π... k If the j-th sample is a random permutation of {1,...,N}, then the value X in the k-th dimension is... j,k for:

[0184] ;

[0185] Space filling and uniformity assurance: The LHS algorithm is used to perform equal-probability partitioning within the cumulative distribution function of each dimension, ensuring that only one sample point is selected in each sub-interval. The process scheme sample set generated in this way can comprehensively cover the boundary conditions of parallel combinations of experimental instruments with relatively low computational cost. The generated process flow sample set is pre-filtered by association rules and conflict matrices to remove physically unexecutable samples, providing high-quality heuristic hotspot regions for subsequent search algorithms.

[0186] Design of a Spatiotemporal Cooperative A* Path Search Algorithm with Dynamic Resource Weights: After obtaining a high-quality initial solution space distribution through LHS, the next step is to plan microscopic executable paths within specific temporal logic. Traditional A* algorithms excel in static map optimization, but often struggle to handle the time-varying characteristics of equipment occupancy when dealing with time-varying process resource scheduling. Therefore, a spatiotemporal cooperative A* algorithm is proposed, which achieves global dynamic optimization of the experimental plan by constructing a three-dimensional spatiotemporal state grid. Experimental equipment resources are used as nodes, and the media connections between devices are used as edges. Directed weighted edges are introduced. This characterizes the directionality of airflow in the medium and labels the operation sequence constraint operators. A deep optimization heuristic evaluation function is used; the improved evaluation function is defined as follows:

[0187] ;

[0188] in, From the initial state to the current process stage The actual cumulative cost; The guided heuristic values ​​obtained from training based on LHS sampled samples are used to preferentially branch towards high-yield regions using a sample guidance algorithm with excellent performance. This is a dynamic conflict cost term, derived in real-time from the conflict matrix constructed above. If the currently planned path conflicts with other experimental tasks on the timeline, This will significantly increase the computational load, forcing the algorithm to avoid obstacles. A spatiotemporal multidimensional path search logic is employed to search for the connection order of test equipment nodes. By introducing timestamp identifiers into the node search, the time-division multiplexing planning problem for multiple testers on the same physical pipeline is solved, thereby achieving the optimal configuration of the serial-parallel process flow.

[0189] Quantitative Feasibility Assessment of Test Plan Scheduling Paths Based on Conflict Matrix Filtering: A complete path selection and convergence analysis process was established using a real-world test scenario as an example. For a certain tester's low-voltage direct supply test requirement, the planning process can be described as candidate rule activation, initially identifying candidate rules based on an association rule base. A candidate initial process path that satisfies the test conditions. Matrix-based conflict reduction is used to transform this... Each path is mapped to a multi-tester resource configuration conflict matrix, and the node valve status conflict matrix is ​​compared with the equipment status conflict matrix in real time. By detecting factors such as pipeline occupancy conflicts between concurrent tests, automatic elimination is achieved. A path that is not executable at physical boundaries or in logical timing. Ultimately reserved. The feasible paths are further refined using a spatiotemporal collaborative A* algorithm. The algorithm selects the globally optimal path with the shortest task response latency and the most balanced energy consumption. This method overcomes the limitations of previous methods relying on manual trial and error, enabling a more comprehensive and accurate assessment of the response capabilities of critical equipment under extreme testing conditions. It provides a scientific and deterministic decision-making basis for planning test routes in complex systems. Through the integrated application of these methods, a perfect combination is achieved, from the broad exploration of the random sampling space to the deep optimization of heuristic search algorithms. This algorithmic system can effectively shield the heterogeneity of heterogeneous resources, adaptively planning efficient, safe, and reliable test processes in dynamically fluctuating resource environments.

[0190] Other embodiments of this application will readily occur to those skilled in the art upon consideration of the specification and practice of the disclosure herein. This application is intended to cover any variations, uses, or adaptations of this application that follow the general principles of this application and include common knowledge or customary techniques in the art not disclosed herein.

[0191] It should be understood that this application is not limited to the precise structure described above and shown in the accompanying drawings, and various modifications and changes can be made without departing from its scope.

Claims

1. A process flow planning method oriented towards the needs of multiple objects, characterized in that, Includes the following steps: A standardized test requirement model is constructed by extracting test requirements and test equipment information from an expert rule base. Establish a data representation model for experimental equipment resources, including a hierarchical dependency model of main equipment resources and a digital coding model of process valves and pipelines; Construct a mapping model between test requirements and test equipment, and match test requirements with equipment resource combination templates; Construct a resource allocation conflict matrix for a multi-tester process system to detect resource conflicts among equipment, valves, and pipelines in real time; A process parameter sample set is generated based on Latin hypercube sampling, and the process flow is planned and optimized.

2. The process flow planning method for multiple object requirements as described in claim 1, characterized in that, In the steps of extracting test requirements and test equipment information based on the expert rule base and constructing a standardized test requirements model: Establish an expert rule base, define the attributes of the test objects, divide the test objects into overall and local types, and map and match the overall and local core resources; Extract the core parameters and time constraints of the experiment, quantify the key aerodynamic parameters of temperature, pressure, and flow rate, and determine the expected start time and duration of the experiment.

3. The process flow planning method for multiple object requirements as described in claim 2, characterized in that, After extracting the core experimental parameters and time constraints, quantifying key aerodynamic parameters such as temperature, pressure, and flow rate, and determining the expected start time and duration of the experiment: Based on the range of core parameter combinations, determine the category of derived experiments and complete the standardized data modeling and digital coding of the experimental requirements.

4. The process flow planning method for multiple object requirements as described in claim 1, characterized in that, In the steps of establishing a data representation model for experimental equipment resources, including a hierarchical dependency model of the main equipment resources and a digital coding model of process valves and pipelines: Digital codes are used to identify the in-use, out-of-use, and maintenance status of the main equipment of the compressor unit, heating furnace, and drying unit, forming a unified status identifier. A hierarchical dependency model for main equipment is constructed, abstracting resources into basic equipment units and composite equipment systems. The basic units record unique numbers, technical parameters and operating status, while the composite systems dynamically calculate the overall aggregation status based on the bottleneck effect. Establish a comprehensive data model for the equipment to quantify its physical operational capabilities, energy consumption characteristics, safe operating range, and lifecycle status.

5. The process flow planning method for multiple object requirements as described in claim 4, characterized in that, After establishing a comprehensive data model of the equipment and quantifying its physical operational capabilities, energy consumption characteristics, safe operating range, and lifecycle status: Digital modeling is performed on process valves and pipelines, with valves set as real nodes and pipeline intersections as virtual nodes. Matrix coding is used to represent valve opening and closing states and pipeline conflict constraints.

6. The process flow planning method for multi-object needs as described in claim 1, characterized in that, In the steps of constructing a mapping model between test requirements and test equipment, and matching test requirements with equipment resource combination templates: Collect historical test data, convert equipment parameters and environmental indicators into unified feature vectors, and complete data preprocessing and standardization; Cluster the experimental requirements data and optimize the initial cluster center selection method and similarity calculation method; The clustering results are divided into typical test condition clusters, and each cluster is bound to a corresponding optimal equipment resource combination template.

7. The process flow planning method for multiple object requirements as described in claim 6, characterized in that, After dividing the clustering results into typical test condition clusters and binding the corresponding optimal equipment resource combination template to each cluster: Once new test requirements are input, the similarity with each operating condition cluster is calculated, the requirements are assigned to the optimal cluster, and the corresponding equipment resource template is called.

8. The process flow planning method for multi-object needs as described in claim 1, characterized in that, In the steps of constructing a resource allocation conflict matrix for a multi-testing instrument process system and real-time detecting resource conflicts among equipment, valves, and pipelines: The test pipeline system is converted into a topology diagram composed of nodes and pipes, and the key resource sets of equipment, valves, and pipes are divided. Construct a multidimensional heterogeneous resource conflict matrix and use different values ​​to identify resource non-conflict, self-occupation, strong conflict and soft competition relationships; Perform flow path pre-simulation to complete comprehensive conflict detection of physical mutual exclusion, parameter boundaries, and medium flow direction.

9. The process flow planning method for multiple object requirements as described in claim 8, characterized in that, After performing flow path pre-simulation and completing comprehensive conflict detection steps involving physical mutual exclusion, parameter boundaries, and medium flow direction: The conflict matrix is ​​updated in real time, and the matrix elements are dynamically corrected according to changes in equipment operating status, forming a dynamic evolution and feedback mechanism.

10. The process flow planning method for multiple object requirements as described in claim 1, characterized in that, In the steps of generating a process parameter sample set based on Latin hypercube sampling and planning and optimizing the process flow: Define the process parameter space and select resource allocation weight, execution time period, and flow adjustment range as key sampling dimensions; Generate process parameter samples covering the entire domain, and retain feasible samples through dual filtering using rules and conflict matrices; Feasibility assessments were conducted on candidate process routes to select the globally optimal process flow with short response time and balanced energy consumption.

Images