Self-adaptive demand measurement model generation method and device based on flexible arrangement
By employing flexible orchestration technology, an adaptive demand measurement model generation method and apparatus are provided, which solves the problems of poor flexibility and weak scalability of existing measurement models, and realizes rapid response and intelligent optimization for complex scenarios.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-12-19
- Publication Date
- 2026-04-17
AI Technical Summary
Existing technologies suffer from poor model flexibility and scalability in complex fields such as emergency rescue and disaster relief, and offshore rescue. They are unable to adapt to uncertain scenarios and lack the ability to handle complex dependencies.
Employing flexible orchestration technology, this invention provides an adaptive demand measurement model generation method and apparatus through a measurement standard normalization module, a calculation formula construction module, a calculation model orchestration engine, and a common parameter management module. It supports user-defined configuration and orchestration, enabling rapid model construction and expansion.
It enables rapid response to unknown scenarios, lowers the technical threshold, improves the ability to handle complex relationships, reduces maintenance and expansion costs, and forms an intelligent self-optimization closed loop.
Smart Images
Figure CN121879743A_ABST
Abstract
Description
Technical Field
[0001] This invention belongs to the field of material data processing technology, specifically relating to a method and apparatus for generating an adaptive demand calculation model based on flexible orchestration. Background Technology
[0002] In complex fields such as emergency rescue and disaster relief, maritime rescue, and typhoon prevention, providing requirements estimation for related tasks is a core task. Traditional requirements estimation usually relies on pre-built fixed models and standard databases. These models show significant limitations when facing complex environments with uncertain task scenarios, uncertain types of participating objects, and uncertain task difficulty and duration.
[0003] The existing technical solutions have the following main drawbacks: 1) Poor flexibility: The calculation models and standards are fixed. Once an "unknown" task or "unknown" calculation standard is encountered that exceeds the preset range, the existing system cannot handle it and requires time-consuming secondary development and modification. 2) Weak scalability: It cannot support users to independently and quickly build new calculation models based on newly emerging task scenarios and standards. 3) Insufficient ability to handle complex dependencies: For complex calculation scenarios with multiple layers of nesting and interdependence between materials, traditional models are often structurally rigid and difficult to arrange and combine in an intuitive way.
[0004] Therefore, there is an urgent need for a technical solution that can adapt to uncertain scenarios, support user-defined configuration and orchestration, and generate new calculation models without secondary development. Summary of the Invention
[0005] (a) Technical problems to be solved The technical problem to be solved by the present invention is how to provide a method and apparatus for generating an adaptive demand measurement model based on flexible orchestration, so as to solve the problems of poor flexibility, weak scalability and insufficient ability to handle complex dependencies in the existing technical solutions.
[0006] (II) Technical Solution To address the aforementioned technical issues, this invention proposes an adaptive demand measurement model generation device based on flexible orchestration. This device includes: a measurement standard normalization module, a calculation formula construction module, a calculation model orchestration engine, a common parameter management module, and an experience coefficient learning and recommendation module. Measurement Standard Normalization Module: This module is used to normalize the input material measurement standards, which have various heterogeneous data formats, into a tree-structured data that can be uniformly recognized within the device. Calculation formula construction module: Provides a graphical or script-based formula editor, supporting user-defined calculation formulas; the calculation formulas are normalized into combinations of variables, constants, and operators; Computational model orchestration engine: Employs a combination of nesting and orchestration techniques to build multi-level computational models, including simple, medium, and complex models. Common Parameter Management Module: Used to define and manage common parameters shared in multiple different calculation formulas; ensures that in a single calculation task, all calculation formulas that reference the common parameter use the same assignment; Experience coefficient learning recommendation module: It is used to analyze the data generated by the execution of historical tasks and provide intelligent initial value recommendations for common parameters in new calculation tasks, thereby realizing the system's self-learning and continuous optimization.
[0007] This invention also provides a method for generating an adaptive demand measurement model based on flexible orchestration, the method comprising the following steps: S1: Measurement Standard Normalization: Receives heterogeneous material measurement standards from external input, parses and converts them into unified tree-structured data according to predefined normalization rules, and uses them as the source of variables for subsequent calculations; S2: Calculation Formula Definition: Through the formula editor, user input is received to create or edit calculation formulas. Formula elements include variables, constants, and operators, and the source of variable references is defined. Variable types support single values, single-level sequences, standard data in tree structures, and references to calculation results from other formulas. Constants use common parameters. S3: Computational Model Orchestration: Build simple, medium, or complex models to meet different needs; S4: Common Parameter Configuration: During model orchestration, common parameters are defined and referenced in all calculation formulas that require these parameters; common parameters are recommended based on historical task data by constructing a weighted algorithm model. S5: Model Execution and Output: Trigger the execution of the orchestrated computational model. The engine schedules the execution of each computational formula according to the dependency relationship and finally outputs the comprehensive calculation result. S6: Model reuse; supports storing complex models as templates, and when performing similar tasks, the system automatically recommends and supports adaptive configurations to quickly respond to task calculation requirements.
[0008] (III) Beneficial Effects This invention proposes a method and apparatus for generating an adaptive demand measurement model based on flexible orchestration. Compared with the prior art, this invention has the following significant advantages: 1. Strong adaptability to various scenarios: By normalizing and atomizing the measurement standards and calculation formulas, this device can quickly assemble measurement models to cope with unknown scenarios, just like "building blocks", which fundamentally solves the problems of rigidity and the need for secondary development in traditional systems.
[0009] 2. High user autonomy: It provides an intuitive formula editor and model arrangement interface, allowing domain experts (non-programmers) to directly participate in model building, which greatly reduces the technical threshold and development cycle, and realizes "what you think is what you get" model building.
[0010] 3. Excellent ability to handle complex relationships: Based on the proposed dual technology of "nesting + orchestration", it can not only describe the vertical multi-layer composition relationship between materials (nesting) well, but also flexibly handle the horizontal cross-dependency and combination relationship (orchestration), and can represent the extremely complex security logic in reality.
[0011] 4. Low maintenance and expansion costs: Since all models are built based on a unified normalization standard and method, adding new standards or models only requires configuration on the front end, without modifying the core back-end devices, which greatly reduces the long-term maintenance and expansion costs of the system.
[0012] 5. Enhanced intelligence: The system has formed an intelligent closed loop of "data-driven and continuous evolution", which solves the problem of traditional systems becoming more rigid with use, and instead becomes more intelligent with use. Attached Figure Description
[0013] Figure 1 This is a schematic diagram of the module composition of the demand calculation model generation device in an embodiment of the present invention; Figure 2 This is a schematic diagram of the measurement standard normalized to a tree structure data in an embodiment of the present invention; Figure 3 This is a schematic diagram illustrating the construction logic of simple models, medium-sized models (nested), and complex models (orchestration) in embodiments of the present invention; Figure 4 This is a flowchart illustrating the steps of the demand estimation model generation method in an embodiment of the present invention. Detailed Implementation
[0014] To make the objectives, contents, and advantages of the present invention clearer, the specific embodiments of the present invention will be described in further detail below with reference to the accompanying drawings and examples.
[0015] This invention relates to a method and apparatus for generating intelligent demand estimation models when the task scenario, object, and standards are all uncertain.
[0016] This invention aims to overcome the shortcomings of the prior art by providing a method and apparatus for generating demand measurement models for uncertain scenarios based on flexible orchestration technology. This addresses the key challenges of poor flexibility, weak scalability, and inability to adapt to unknown scenarios and standards in existing measurement systems. Specific objectives include: 1. Provide a device and method that support the autonomous construction of measurement models, enabling them to dynamically adapt to the measurement needs of different task scenarios.
[0017] 2. Enables the device to be adapted to various known or unknown measurement standards by users through independent configuration, without requiring secondary modifications to the underlying device.
[0018] 3. The device enables users to flexibly construct various known or unknown complex measurement models through independent configuration and arrangement, without the need for coding development.
[0019] To achieve the above-mentioned objectives, the present invention adopts the following technical solution: In a first aspect, the present invention provides a demand measurement model generation device based on flexible orchestration technology, comprising: The measurement standard normalization module is used to normalize the input material measurement standards, which have various heterogeneous data formats, into a tree-structured data that can be uniformly identified within the device. The core of this module lies in defining a set of core components and their organization rules, mapping measurement standards from different sources and in different formats into structured tree nodes and attributes.
[0020] The calculation formula construction module provides a graphical or script-based formula editor, supporting user-defined calculation formulas. These formulas are normalized into combinations of variables, constants, and operators. Variable types support single values, single-level sequences, standard data types in tree structures, and referencing calculation results from other formulas. Constants use common parameters.
[0021] Computational model orchestration engine: Employs a combination of nesting and orchestration techniques to construct multi-layered computational models. It supports: Simple model building: calls a single calculation formula that does not reference the results of other formulas.
[0022] Medium-level model construction: This method utilizes a multi-level nested calculation formula technique to handle material measurement needs with multiple layers of dependencies. Variables in higher-level formulas can reference the calculation results of lower-level formulas, and the nesting level is unlimited.
[0023] Complex model construction: Through calculation formula arrangement technology, multiple simple and / or medium models are arranged, combined and associated to solve the needs of a single support mission for the calculation of multiple types of materials that are interdependent.
[0024] The Common Parameter Management module is used to define and manage common parameters shared across multiple different calculation formulas. It ensures that all calculation formulas referencing a common parameter use the same assignment within a single calculation task, achieving "one-stop definition, multiple references".
[0025] Experience coefficient learning and recommendation module: This module analyzes data from historical task executions to provide intelligent initial value recommendations for common parameters in new calculation tasks, thereby enabling the system's self-learning and continuous optimization. The module includes: Historical task archive: Used to store a complete record of each task execution, including task feature vectors, the combination of coefficients used, and performance evaluation after task completion; Task Feature Analysis and Similarity Matching Unit: Used to parse the features of new tasks and search for historical tasks with similar features in the historical archive; Performance-weighted recommendation engine: Based on similar historical tasks and their performance, it calculates and recommends the initial values of each experience coefficient in new tasks through a weighted fusion algorithm. Recommendation confidence assessment unit: Used to assess the reliability of recommendation results and provide users with a reference.
[0026] Secondly, the present invention provides a method for generating a demand estimation model based on flexible orchestration technology, applied to the aforementioned device, comprising the following steps, such as... Figure 4 As shown: S1: Measurement Standard Normalization: Receives heterogeneous material measurement standards from external input, parses and converts them into unified tree-structured data according to predefined normalization rules, and uses them as the variable source for subsequent calculations.
[0027] S2: Formula Definition: The formula editor receives user input, allowing users to create or edit formulas. Formula elements include variables, constants, and operators, and the source of variable references can be defined. Variable types support single values, single-level sequences, standard data from tree structures, and referencing calculation results from other formulas. Constants use common parameters.
[0028] S3: Computational Model Orchestration: Build simple, medium, or complex models to meet different needs; S31: For simple needs, construct a simple model and directly associate it with the formulas and variables defined in step S2.
[0029] S32: For hierarchical structures, an intermediate model is constructed by using the output of the lower-level computational model as the input variable of the upper-level computational model.
[0030] S33: For complex scenarios involving multiple types of materials and horizontal dependencies, multiple simple and / or medium-sized models can be visually connected and logically combined through an orchestration interface, defining the data flow and dependencies between models, thereby constructing complex models.
[0031] S4: Common Parameter Configuration: During model orchestration, common parameters are defined and referenced in all calculation formulas that require them. Common parameters can be recommended based on historical task data by building the algorithm model.
[0032] S5: Model Execution and Output: Trigger the execution of the orchestrated computational model. The engine schedules the execution of each computational formula according to the dependencies and finally outputs the comprehensive measurement results.
[0033] S6: Model Reuse. Supports storing complex models as templates. When performing similar tasks, the system automatically recommends and supports adaptive configurations, quickly responding to task calculation requirements.
[0034] like Figure 1 As shown, the device of this invention comprises five core modules: a measurement standard normalization module, a calculation formula construction module, a calculation model orchestration engine, a common parameter management module, and an experience coefficient learning and recommendation module. These modules work together to achieve flexible model construction.
[0035] Example 1: Measurement Standard Normalization Taking a certain "camping water supply" measurement standard as an example, as shown in the table below, the original data is in tabular form, and the table styles vary greatly due to different standards. The data formats are not uniform, and computer parsing programs often need to be customized to meet the parsing requirements.
[0036]
[0037] After processing by the standard normalization module, it is uniformly converted into a tree structure, such as... Figure 2 As shown, computers can use the same parsing program to parse and calculate various types of heterogeneous standard data.
[0038] Example 2: The computational model construction process, such as Figure 3 As shown, this paper takes the calculation of material requirements for a complex environment emergency rescue and disaster relief mission as an example to introduce the construction process of simple, medium, and complex models. The material requirements for emergency rescue and disaster relief missions include emergency rescue expenses (composed of several costs), emergency food rations, emergency water, fuel, and transport vehicles.
[0039] (1) Simple model construction Simplified Model A: Calculation of Total Injury Costs Based on Injury Grade. Injury costs based on injury grade are a component of living expenses during disaster relief and rescue operations.
[0040]
[0041] In the formula: —Estimated total cost of injury for a certain level of injury (in yuan); P实力 —Personnel strength; e 受伤 —The percentage of injuries out of total casualties is based on experience; e 伤情i —The percentage of the total number of injured persons expected to be of a certain injury level is an empirical value; B 伤标i —Standard for injury compensation for a certain level of injury (RMB per person) Depending on the severity of the disaster relief efforts, the percentage of wounded casualties as a percentage of total troop strength is a range. This example assumes that wounded casualties account for 20% of the total troop strength. Of the wounded, it is estimated that 40% will suffer third-degree injuries, 36% second-degree injuries, and 24% first-degree injuries. The standards for issuing penalties for each level are as follows.
[0042] Reference Table for Injury Compensation Payment Standards
[0043] The corresponding tree data structure: Root node: Injury compensation standards Leaf node 1: First-class (Attribute: Value: X, Unit: Yuan / per person per time) Leaf node 2: Second class (Attribute: Value: Y, Unit: Yuan / per person per time) Leaf node 3: Third-class (Attribute: Value: Z, Unit: Yuan / person / time) Calculate the cost of first-degree injury.
[0044] Define formulas using the formula editor:
[0045] Among them, e 负伤 , Experience values can be set as variables, allowing users to customize their values. This is a public parameter that is also used when calculating other resources. For standard purposes, variables can be referenced from the aforementioned tree-structured standards. Similarly, editable formulas can be used to calculate the second-degree injury compensation. Third-degree injury compensation .
[0046] (2) Medium model construction Moderate Model B: Calculating Injury Costs for Each Injury Level
[0047] Where i = 1, 2, ..., n represents different injury levels.
[0048] Based on the above injury severity classification, use the formula editor to define the formula.
[0049] Because model B directly incorporates the calculation results of model A, it constitutes a nesting.
[0050] Medium Model C: Calculating the funding for disaster relief and rescue missions .
[0051] Emergency rescue and disaster relief mission expenses consist of several components, including living expenses. It is one of them. Use the formula editor to define model C and calculate... :
[0052] Variables used in the calculation process of model C It references the output of model B, forming a secondary nesting.
[0053] Similarly, the same method is used to calculate the task cost. Other component costs.
[0054] Finally, use the formula editor to define model D and calculate... :
[0055] The calculation process of model D uses variables It references the output of model C, forming multiple nested structures.
[0056] (3) Complex model arrangement Planning a disaster relief exercise involves calculating mission costs, emergency rations, camping water, fuel, and transport vehicles. The calculations integrate common parameters such as troop strength and duration in the ration and fuel demand models with the mission cost model. The number of transport vehicles is determined by the combined weight of emergency rations, camping water, and the total volume of fuel. In the orchestration interface, users can connect the outputs of these three models to define a complex model E: Number of vehicles = (Emergency ration weight + Camping water weight) / Maximum carrying capacity per vehicle + Fuel volume / Maximum tanker truck volume. This orchestration constructs a complex dependency model across material categories.
[0057] Application of common parameters: In all the models mentioned above, "emergency rescue and disaster relief forces" and "mission duration" can be set as common parameters. When a mission starts, the user only needs to input "emergency forces" and "mission duration" once in the common parameter management module. All models that reference these two parameters will automatically use these values during calculation, ensuring data consistency.
[0058] Example 3: Intelligent Recommendation Based on Historical Data Mining Plan an emergency drill mission of "high-altitude mountainous terrain, winter, brigade-level scale, 10 days". System intelligent recommendation process: (1) Feature matching: The system extracts the new task feature T_new and matches it with the historical database. Three similar historical tasks are found: Task X: {Mountainous terrain, winter, regimental level, 7 days}, Excellent performance (95), Utilization coefficient {Cold-weather supplies redundancy coefficient: 1.5} Task Y: {High altitude, autumn, brigade level, 14 days}, performance: good (85), utilization coefficient: {cold-weather supplies redundancy coefficient: 1.3} Task Z: {High-altitude winter, battalion level, 5 days}, Excellent performance (96), Utilization coefficient {Cold-weather supplies redundancy coefficient: 1.8} (2) Weighted calculation: Calculate the composite weight of similarity and performance. Recommended values for common parameters: ,in, For historical mission With the current task The similarity is obtained by vectorizing the keyword combinations of the tasks and then calculating the cosine similarity. For historical mission Performance value, For historical mission The common parameter values. For the current task The recommended values for public parameters are obtained by combining the keyword matching degree and performance value of historical tasks to weighted values. The resulting recommended values are more accurate and better suited to actual task scenarios. Specifically, in this embodiment: For the "redundancy coefficient of cold-weather supplies", the recommended value is... .
[0059] (3) Results presentation: The system recommends setting the coefficient to 1.6. The recommended value of 1.6 is adopted as the initial value, and the goal of "minimizing transportation costs" is set. The system calls the real-time optimization algorithm to search in a small range around 1.6, and finally recommends that the coefficient be fine-tuned to 1.55, which can further reduce transportation costs by 8% while meeting the guarantee requirements.
[0060] In summary, this invention provides a highly flexible and scalable demand measurement model generation scheme through a combination of technologies including normalization, atomization, arrangement and nesting, and intelligent recommendation, effectively solving the problem of automated measurement in uncertain scenarios.
[0061] Compared with the prior art, the present invention has the following significant advantages: 1. Strong adaptability to various scenarios: By normalizing and atomizing the measurement standards and calculation formulas, this device can quickly assemble measurement models to cope with unknown scenarios, just like "building blocks", which fundamentally solves the problems of rigidity and the need for secondary development in traditional systems.
[0062] 2. High user autonomy: It provides an intuitive formula editor and model arrangement interface, allowing domain experts (non-programmers) to directly participate in model building, which greatly reduces the technical threshold and development cycle, and realizes "what you think is what you get" model building.
[0063] 3. Excellent ability to handle complex relationships: Based on the proposed dual technology of "nesting + orchestration", it can not only describe the vertical multi-layer composition relationship between materials (nesting) well, but also flexibly handle the horizontal cross-dependency and combination relationship (orchestration), and can represent the extremely complex security logic in reality.
[0064] 4. Low maintenance and expansion costs: Since all models are built based on a unified normalization standard and method, adding new standards or models only requires configuration on the front end, without modifying the core back-end devices, which greatly reduces the long-term maintenance and expansion costs of the system.
[0065] 5. Enhanced intelligence: The system has formed an intelligent closed loop of "data-driven and continuous evolution", which solves the problem of traditional systems becoming more rigid with use, and instead becomes more intelligent with use.
[0066] The above description is only a preferred embodiment of the present invention. It should be noted that for those skilled in the art, several improvements and modifications can be made without departing from the technical principles of the present invention, and these improvements and modifications should also be considered within the scope of protection of the present invention.
Claims
1. A flexible orchestration-based adaptive demand estimation model generation device, characterized in that, The device includes: a measurement standard normalization module, a calculation formula construction module, a calculation model orchestration engine, a public parameter management module, and an experience coefficient learning and recommendation module; Measurement Standard Normalization Module: This module is used to normalize the input material measurement standards, which have various heterogeneous data formats, into a tree-structured data that can be uniformly recognized within the device. Calculation formula construction module: Provides a graphical or script-based formula editor, supporting user-defined calculation formulas; the calculation formulas are normalized into combinations of variables, constants, and operators; Computational model orchestration engine: Employs a combination of nesting and orchestration techniques to build multi-level computational models, including simple, medium, and complex models. Common Parameter Management Module: Used to define and manage common parameters shared in multiple different calculation formulas; ensures that in a single calculation task, all calculation formulas that reference the common parameter use the same assignment; Experience coefficient learning recommendation module: It is used to analyze the data generated by the execution of historical tasks and provide intelligent initial value recommendations for common parameters in new calculation tasks, thereby realizing the system's self-learning and continuous optimization. 2.The flexible orchestration based adaptive demand estimation model generation apparatus of claim 1, wherein, The measurement standard normalization module defines a set of core components and their organization rules, mapping measurement standards from different sources and in different formats into structured tree nodes and attributes. 3.The flexible orchestration based adaptive demand estimation model generation device of claim 1, wherein, In the calculation formula construction module, the variable types support single values, single-level sequences, standard data in tree structures, and calculation results from other formulas. Constants use common parameters. 4.The flexible orchestration based adaptive demand estimation model generation apparatus of claim 1, wherein, The experience coefficient learning recommendation module includes: Historical task archive: Used to store a complete record of each task execution, including task feature vectors, the combination of coefficients used, and performance evaluation after task completion; Task Feature Analysis and Similarity Matching Unit: Used to parse the features of new tasks and search for historical tasks with similar features in the historical archive; Performance-weighted recommendation engine: Based on similar historical tasks and their performance, it calculates and recommends the initial values of each experience coefficient in new tasks through a weighted fusion algorithm. Recommendation confidence assessment unit: Used to assess the reliability of recommendation results and provide users with a reference. 5.The flexible orchestration based adaptive demand estimation model generation device of claim 1, wherein, The simple model calls a single calculation formula that does not reference the results of other formulas. 6.The flexible orchestration based adaptive demand estimation model generation apparatus of claim 5, wherein, The intermediate model uses a multi-layered nested calculation formula technique to handle material measurement needs with multiple dependencies; variables in the upper-level formula reference the calculation results of the lower-level formula, and the nesting level is unlimited.
7. The flexible orchestration based adaptive demand estimation model generation apparatus as claimed in claim 6, wherein, Complex models use computational formula arrangement techniques to arrange, combine, and associate multiple simple and / or medium-sized models.
8. A flexible orchestration-based adaptive demand estimation model generation method, characterized in that, The method includes the following steps: S1: Measurement Standard Normalization: Receives heterogeneous material measurement standards from external input, parses and converts them into unified tree-structured data according to predefined normalization rules, and uses them as the source of variables for subsequent calculations; S2: Calculation Formula Definition: Through the formula editor, user input is received to create or edit calculation formulas. Formula elements include variables, constants, and operators, and the source of variable references is defined. Variable types support single values, single-level sequences, standard data in tree structures, and references to calculation results from other formulas. Constants use common parameters. S3: Computational Model Orchestration: Build simple, medium, or complex models to meet different needs; S4: Common Parameter Configuration: During model orchestration, common parameters are defined and referenced in all calculation formulas that require these parameters; common parameters are recommended based on historical task data by constructing a weighted algorithm model. S5: Model Execution and Output: Trigger the execution of the orchestrated computational model. The engine schedules the execution of each computational formula according to the dependency relationship and finally outputs the comprehensive calculation result. S6: Model reuse; It supports storing complex models as templates, and when performing similar tasks, the system automatically recommends and supports adaptive configurations to quickly respond to task calculation needs.
9. The adaptive demand measurement model generation method based on flexible orchestration as described in claim 8, characterized in that, In S3, S31: For simple needs, build a simple model and directly associate it with the formulas and variables defined in step S2; S32: To address the need for hierarchical structures, an intermediate model is constructed by using the output of the lower-level computational model as the input variable of the upper-level computational model. S33: For complex scenarios involving multiple types of materials and horizontal dependencies, multiple simple and / or medium-sized models can be visually connected and logically combined through an orchestration interface, defining the data flow and dependencies between models, thereby constructing complex models.
10. The method for generating an adaptive demand measurement model based on flexible orchestration as described in claim 8, characterized in that, In step S4, the public parameters are recommended based on historical task data by constructing a weighted algorithm model, specifically including: Recommended values for common parameters: ,in, For historical mission With the current task The similarity is obtained by vectorizing the keyword combinations of the tasks and then calculating the cosine similarity. For historical mission Performance value, For historical mission The common parameter values; For the current task Recommended values for common parameters.