Intelligent decision-making system and method based on system modeling and potential calculation

Abstract

The invention discloses an intelligent decision-making system and method based on system modeling and calculation potential calculation, and aims to solve the problem of game calculation and optimization in a complex system and convert a classical game principle into a computable engineering framework. The system is used for resource allocation, risk prediction and optimal strategy generation in a multi-subject system, and the architecture of the system comprises an input layer, a calculation layer, an output layer and a feedback layer. The system firstly collects multi-dimensional variables such as economy, resources and information, the multi-dimensional variables are mapped into a multi-field-domain input matrix, and then a calculation layer calculates the potential energy difference of the system through a potential calculation function based on the matrix and generates a map. And finally, the feedback layer performs dynamic self-calibration through real-time data iteration model weight, so that the potential energy difference of the system tends to be stably converged. According to the method, a technical structure of multi-dimensional input, potential calculation and feedback closed loop is realized, the method can be applied to complex systems in multiple fields of business, economy, supply chains and the like, and the optimal solution of resource configuration and system stability is realized through parameter matrix and model self-calibration.

Description

TECHNICAL FIELD

[0001] The application belongs to the technical field of artificial intelligence auxiliary decision-making system, and particularly relates to an intelligent decision-making system and method based on system modeling and potential calculation. BACKGROUND

[0002] In modern society, organizations, enterprises and industrial ecology are facing highly coupled and uncertain environments. Multivariate interaction, supply chain vulnerability, information redundancy and decision-making delay make it difficult for traditional experience-based analysis to form effective judgments. Existing AI models focus on single-field prediction and lack cross-field feedback mechanisms and system advantage evaluation capabilities. The "potential difference judgment" and "win first, then fight" principles in the classic "Sun Tzu Art of War" embody the idea of system potential and game balance. However, traditional literature does not form an engineering reproducible structure, making it difficult to apply in modern complex systems.

[0003] Therefore, the present application proposes an intelligent decision-making system and method based on system modeling and potential calculation to realize the "potential recognition-strategy generation-feedback iteration" process of complex systems through engineering means, to solve the following problems existing in the prior art: 1. Lack of system potential calculation methods that can be used across fields; 2. Lack of dynamic feedback multi-agent game optimization mechanism; 3. Lack of reproducible models to implement classical game principles through technical means. SUMMARY

[0004] The purpose of the present application is to provide an intelligent decision-making system and method based on system modeling and potential calculation to solve the above technical problems.

[0005] To achieve the above application purposes, the technical solutions adopted by the present application are as follows: The embodiment of the present application provides an intelligent decision-making system based on system modeling and potential calculation, which is used for resource allocation, risk prediction and optimal strategy generation in a multi-agent system, comprising: An input layer is used to collect multi-dimensional variables such as economy, resources, information, ecology and organizational behavior, and map the multi-dimensional variables into a multi-field input matrix FieldMatrix; A calculation layer is configured with a potential algorithm, which is used to calculate the system potential difference ΔP through a potential function f based on the multi-field input matrix FieldMatrix, dynamic weight coefficient Weight and time step Time, to generate a system potential difference graph; An output layer is used to generate an optimized path, a cooperation strategy and a resource scheduling scheme according to the system potential difference graph; a feedback layer for dynamically self-calibrating based on real-time data iteration model weights to stabilize the system potential difference.

[0006] In some embodiments, the computing layer is configured with the following functional modules: a FieldRemap module for abstracting the five elements of classical game theory into seven field variables to form the FieldMatrix; a ComputeAdvantage module for calculating the system potential difference ΔP and key node weights based on system dynamics; a VirtualRealSwitch module for performing virtual-real scenario switching operations to test robustness and output optimal paths; a ChainCollapse module for identifying system fragile chains and constructing minimum cost optimization paths; a FeedbackLoop module for real-time updating of parameter weights to achieve the dynamic self-calibration.

[0007] In some embodiments, the calculation of the system potential difference ΔP satisfies: ΔP=f(FieldMatrix, Weight, Time).

[0008] In some embodiments, the potential function Differential equations, dynamic programming models, or neural network structures can be used.

[0009] In some embodiments, the FeedbackLoop module performs dynamic adjustment of parameter weights to raise the potential of low potential nodes and stabilize the system as a whole.

[0010] Embodiments of the present application also provide an intelligent decision-making method based on system modeling and potential calculation, which is applied to complex system game calculation and optimization, including the following steps: Step S1: Collecting multi-dimensional variables such as economy, resources, information, ecology, and organizational behavior, and mapping the multi-dimensional variables into a multi-field input matrix FieldMatrix through the FieldRemap module; Step S2: Through the ComputeAdvantage module, calculating the system potential difference based on the FieldMatrix, dynamic weight coefficient Weight, and time step Time through the potential function, and generating a system potential difference graph; Step S3: Through the VirtualRealSwitch module, performing virtual-real field switching operations in the potential difference graph to deduce strategy diversity and output optimal paths; Step S4: Identify the key fragile chain in the system through the ChainCollapse module, and construct the minimum cost optimization path; Step S5: Output the optimization path, cooperation strategy and resource scheduling scheme; Step S6: Real-time update the parameter weight through the FeedbackLoop module, realize dynamic feedback loop, so that the system potential energy difference converges stably.

[0011] The embodiment of the application further provides an electronic device, including a memory and a processor, the memory stores a computer program, and the processor realizes the intelligent decision method described above when executing the computer program.

[0012] The embodiment of the application further provides a computer readable storage medium, which stores a computer program, and the computer program is executed by a processor to realize the intelligent decision method described above.

[0013] Compared with the prior art, the embodiment of the application has the beneficial effects that: The intelligent decision system and method based on system modeling and potential calculation provided by the embodiment of the application systematize and calculate the classical game logic, realize the reproducible transformation of the algorithm layer, form a complex system simulation tool applicable to business, economy, ecology, supply chain and other fields, realize the technical structure of multi-dimensional input, potential calculation and feedback loop, realize the optimal solution of resource allocation and system stability through parameter matrix and model self-calibration. BRIEF DESCRIPTION OF DRAWINGS

[0014] In order to more clearly illustrate the technical solutions in the embodiments of the application, the following will briefly introduce the drawings needed to be used in the embodiments or prior art description. Obviously, the drawings in the following description are only some embodiments of the application, and other drawings can be obtained by those skilled in the art without creative labor.

[0015] Figure 1 A flow chart of the intelligent decision method based on system modeling and potential calculation provided by the embodiment of the application is shown.

[0016] Figure 2 The system potential energy difference map (ΔP distribution initial state) provided by the embodiment of the application is shown.

[0017] Figure 3 The system optimization process diagram (feedback and parameter adjustment) provided by the embodiment of the application is shown.

[0018] Figure 4 The optimized system potential energy difference map (ΔP_new distribution) provided by the embodiment of the application is shown. DETAILED DESCRIPTION

[0019] In order to make the technical problems to be solved by the present application, technical solutions and beneficial effects clearer, the present application will be further described in detail below with reference to the drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and not intended to limit the present application.

[0020] It should be noted that when an element is referred to as being "fixed" or "set" on another element, it can be directly on the other element or indirectly on the other element. When an element is referred to as being "connected" to another element, it can be directly connected to the other element or indirectly connected to the other element.

[0021] In addition, the terms "first", "second" are only for descriptive purposes and cannot be understood as indicating or implying relative importance or implicitly indicating the number of the indicated technical features. Therefore, the features defined with "first", "second" can explicitly or implicitly include one or more of the features. In the description of the present application, the meaning of "a plurality of" is two or more, unless otherwise specifically limited.

[0022] In order to illustrate the technical solutions described in the present application, the following will be described by specific embodiments.

[0023] Please refer to Figure 1 As shown in the drawings, the embodiments of the present application provide an intelligent decision-making system and method based on system modeling and potential calculation, which converts the classical game principle (systematic game thought derived from "The Art of War") into a technical implementation scheme of a computable framework, which is used for resource allocation, risk prediction and optimal strategy generation in a multi-agent system.

[0024] Specifically, the system comprises: Input layer (Input Layer): Collecting economic, resource, information, ecological and organizational behavior and other multi-dimensional variables, and mapping the multi-dimensional variables into a multi-field input matrix (FieldMatrix).

[0025] Computation layer (Computation Layer): calculating the system potential difference through the potential algorithm and the system matrix model, for calculating the system potential difference based on the multi-field input matrix FieldMatrix, the dynamic weight coefficient Weight and the time step or the number of calculation iterations Time, through the potential function to calculate the system potential difference, to generate a system potential difference map.

[0026] Output layer (Output Layer): for generating an optimized path, a cooperation strategy and a resource scheduling scheme according to the system potential difference map.

[0027] Feedback Layer: Based on real-time data, iteratively adjust model weights to dynamically self-correct, aiming to stabilize the system potential difference.

[0028] Specifically, the system configures the following modules: M1. FieldRemap module, used to abstract the classical "five events and seven calculations" thought into modern system variable mapping.

[0029] M2. ComputeAdvantage module, used to calculate the potential difference ΔP and key node weights based on system dynamics.

[0030] M3. VirtualRealSwitch module, used for virtual-real scene switching to realize strategy diversity deduction.

[0031] M4. ChainCollapse module, used to identify system fragile chains and build minimum cost optimization paths.

[0032] M5. FeedbackLoop module, used to update parameter weights in real time to achieve continuous optimization.

[0033] As an example, the following gives the architecture pseudo code (anti-copying method): # Generic Systemic Game OS - Research Preview / Proprietary Implementation class GameOS: def __init__(self): self.field_map = FieldRemap() self.engine = ComputeAdvantage() self.switch = VirtualRealSwitch() self.chain = ChainCollapse() self.feedback = FeedbackLoop() def run(self, system_input): F = self.field_map.load(system_input) S = self.engine.evaluate(F) R = self.switch.transform(S) P = self.chain.detect(R) return self.feedback.update(P) Wherein, the module algorithm logic is a proprietary implementation, which meets the reproducibility requirements.

[0034] The model principle described above is explained as follows.

[0035] (1) FieldRemap module Used to abstract the five elements of "Dao, Heaven, Earth, General, Law" in Sun Tzu's Art of War as: • Dao → System consistency and goal alignment; • Heaven → External environment and time rhythm; • Earth → Resource distribution and path constraints; • General → Decision mechanism and execution force; • Law → Rule system and collaboration protocol.

[0036] And extended to modern seven field variables (economy, information, resources, organization, technology, ecology, public opinion) to form an input matrix.

[0037] (2) ComputeAdvantage module Calculate ΔP = f(FieldMatrix, Weight, Time) through the system dynamics model, output the potential energy difference and node weight atlas. Specifically: FieldMatrix is the multi-field input matrix, which is used to represent the state variables of different elements in the system, which can include structured data such as economy, resources, information, organization, etc. Weight is a dynamic weight coefficient, which reflects the influence degree of each element on the system potential energy at different stages, with a value range of 0 ≤ Wi ≤ 1, which can be set by rules or adjusted by historical data. Time is the time step or calculation iteration number of system evolution.

[0038] The potential function is used to calculate the potential energy difference in the multi-variable space, which can adopt linear or nonlinear mapping form (such as differential equation set, dynamic programming model, or neural network structure) to realize the calculation of system potential energy difference (ΔP).

[0039] The potential calculation result generates a "Strategic Potential Map" through node analysis, which is used to identify the advantage nodes and vulnerable chains in the system.

[0040] The present application does not limit the specific function form or the value of the weight parameter to protect the commercial implementation details; but the skilled person in the art can reproduce an algorithmic potential calculation process with equivalent functional effects according to the above structure, parameter definition and input-output logic.

[0041] (3) VirtualRealSwitch module Performing "virtual-real switching" operation in the potential map: Virtual field simulation assumes scenarios to test robustness; Real field calculates real data to output the optimal path.

[0042] (4) ChainCollapse module Identify the key fragile chain in the system, find the intervention point that can maximize the overall stability through simulation calculation, and realize the minimum cost reconstruction.

[0043] (5) FeedbackLoop module Introduce time variable and dynamic feedback to realize real-time model updating and ensure the system to maintain stable convergence under external disturbance.

[0044] The following will be illustrated by Figures 2 to 4 , the technical principle is illustrated by the potential difference map of the system. From Figures 2 to 4 It can be seen that: 1、Output result: The total potential difference of the system ΣΔP is raised from 0.73 to 0.86; The number of risk exposure nodes is reduced from 2 to 0; The potential balance degree is improved by 18%, and the system enters the stable state running interval.

[0045] 2、The system first calculates the potential P_i and potential difference ΔP_i of each node, and automatically identifies the dominant node and fragile chain.

[0046] Subsequently, the FeedbackLoop module performs dynamic adjustment of the parameter weight Wi according to the characteristics of the fragile chain, optimizes resource allocation and node cooperation.

[0047] After optimization, the map is regenerated, and it can be seen that the potential of the low potential node is significantly improved, and the system tends to be stable as a whole.

[0048] The above shows the complete closed-loop process of the system from identification → adjustment → optimization. This process embodies the "identification and early warning + self-adaptive optimization" function of the present application: the system first identifies the potential difference node, then dynamically adjusts the parameters through the feedback module, and finally realizes structural balance and performance optimization.

[0049] In some embodiments of the application, supply chain optimization is used as a specific example.

[0050] I. Background Modern supply chains involve multiple parties and complex constraints, and bottlenecks at any node can cause system-level fluctuations. This embodiment verifies the potential and feedback effects of the above system in this scenario.

[0051] II. Input layer design Input variables include: • Resource distribution parameters (R): raw material supply, energy proportion, transportation capacity; • Economic parameters (E): cost coefficients, price fluctuations, financing availability; • Technical parameters (T): manufacturing autonomy, maturity of alternative technologies; • Information parameters (I): data sharing and transmission delay; • Organizational parameters (O): supplier collaboration, contract stability; • External parameters (X): market demand changes, policy constraints, etc.

[0052] III. Calculation layer deduction Calculate the system potential difference through the ComputeAdvantage module: ΔP = f(R, E, T, I, O, X); When the weight of a certain link W is greater than the threshold T, the system is marked as a critical node.

[0053] The VirtualRealSwitch module simulates different alternative strategies (such as path reconstruction, inventory adjustment, contract redistribution) to compare potential.

[0054] IV. Output and feedback The system outputs multiple optimization paths and adjusts ΔP and weights in real time according to the feedback module.

[0055] In the simulation results, the resource waste rate is reduced by 38%, and the overall risk exposure index is reduced by 42%.

[0056] V. Optional technical implementation conditions (Technical Implementation) The system runs on a computer server cluster or cloud architecture (including cloud general NLP models); Modules interact through API interfaces and shared databases; All data inputs are standardized and encrypted for transmission and processing; The potential calculation module can be accelerated by GPU / TPU, supporting dynamic matrix updates; The model output is displayed through a visual dashboard for decision support.

[0057] The system has a complete technical chain and a reproduction path.

[0058] Six, data and safety compliance (Data Compliance) This application only uses structured public data or simulation data sets, and does not involve any personal information or national security data. The algorithm calculation process complies with the relevant data or information security requirements, and the system output is only an auxiliary analysis result, and does not have the function of automatic execution or black box decision.

[0059] In summary, the intelligent decision system and method based on system modeling and algorithm potential calculation provided by the embodiments of the application have the following beneficial effects: 1. Systematize and computable the classical game logic, realize the reproducible transformation of the algorithm layer; 2. Form a complex system simulation tool that can be applied to business, economy, ecology, supply chain and other fields; 3. Realize the technical structure of multi-dimensional input, algorithm potential calculation and feedback loop; 4. Through parameter matrix and model self-calibration, realize the optimal solution of resource allocation and system stability.

[0060] The protection scope of the application covers the systematic engineering transformation of the principles of Sun Tzu Art of War and the general game calculation framework. The system can be widely used in complex system modeling and optimization in the fields of business, economy, industry, scientific research and education, and does not involve military, political or social management activities. The application is based on the method of system science, realizes the adaptive optimization of complex systems through algorithmic algorithm potential and multi-domain game modeling.

[0061] The above-described embodiments are only used to illustrate the technical solutions of the application, rather than limit them; although the application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that they can modify the technical solutions recorded in the foregoing embodiments, or make equivalent replacement for part of the technical features; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of the application, and should be included in the protection scope of the application.

Claims

1. An intelligent decision-making system based on system modeling and computational potential calculation, characterized in that, Used for resource allocation, risk prediction, and optimal strategy generation in multi-agent systems, including: The input layer is used to collect multi-dimensional variables such as economy, resources, information, ecology and organizational behavior, and map the multi-dimensional variables into a multi-field input matrix FieldMatrix; The computational layer is configured with a computational potential algorithm, which is used to calculate the system potential energy difference ΔP based on the multi-field input matrix FieldMatrix, dynamic weight coefficient Weight and time step Time, through the computational potential function f, so as to generate a system potential energy difference spectrum. The output layer is used to generate optimized paths, cooperation strategies, and resource scheduling schemes based on the system potential energy difference map. The feedback layer is used to iterate the model weights based on real-time data to achieve dynamic self-calibration, so as to stabilize the potential energy difference of the system.

2. The intelligent decision-making system according to claim 1, characterized in that, The computing layer is configured with the following functional modules: The FieldRemap module is used to abstract the five elements of classical game theory into seven modern field variables, forming the FieldMatrix. The ComputeAdvantage module is used to calculate the system potential energy difference ΔP and the weights of key nodes based on system dynamics. The VirtualRealSwitch module is used to perform virtual-real scene switching operations to test robustness and output the optimal path; The ChainCollapse module is used to identify fragile chains in the system and build the minimum-cost optimized path. The FeedbackLoop module is used to update parameter weights in real time to achieve the dynamic self-calibration.

3. The intelligent decision-making system according to claim 1, characterized in that, The calculation of the system potential energy difference ΔP satisfies: ΔP=f(FieldMatrix, Weight, Time).

4. The intelligent decision-making system according to claim 3, characterized in that, The computational potential function f can be a system of differential equations, a dynamic programming model, or a neural network structure.

5. The intelligent decision-making system according to claim 1, characterized in that, The FeedbackLoop module dynamically adjusts parameter weights to increase the potential energy of low-potential nodes, thereby stabilizing the overall system.

6. An intelligent decision-making method based on system modeling and computational potential calculation, characterized in that, Applied to game theory calculation and optimization of complex systems, it includes the following steps: Step S1: Collect multi-dimensional variables such as economy, resources, information, ecology and organizational behavior, and map the multi-dimensional variables into a multi-field input matrix FieldMatrix through the FieldRemap module; Step S2: Using the ComputeAdvantage module, based on the FieldMatrix, dynamic weight coefficient Weight, and time step Time, calculate the system potential energy difference using the computational potential function, and generate a system potential energy difference map; Step S3: Using the VirtualRealSwitch module, perform virtual and real field switching operations in the potential energy difference spectrum to deduce strategy diversity and output the optimal path; Step S4: Using the ChainCollapse module, identify the critical vulnerable chains in the system and construct the minimum cost optimization path; Step S5: Output the optimized path, cooperation strategy, and resource scheduling scheme; Step S6: Update the parameter weights in real time through the FeedbackLoop module to achieve dynamic feedback closed loop, so that the potential energy difference of the system can be stably converged.

7. An electronic device, characterized in that, It includes a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the intelligent decision-making method of claim 6.

8. A computer-readable storage medium, characterized in that, It stores a computer program, which, when executed by a processor, implements the intelligent decision-making method as described in claim 6.

Images