Credit early warning method and system for cooperation between enterprises

Abstract

The invention relates to the field of enterprise credit risk management, in particular to a credit early warning method and system for cooperation between enterprises, and the method comprises the steps: collecting multi-dimensional data related to the cooperation between the enterprises from a plurality of data sources, and dynamically evaluating the credibility between the enterprises through a multi-dimensional trust evaluation model. And the historical and future trends of external environment variables are captured through a high-dimensional environment response model, credibility evaluation and environment response data are fused, and finally multi-level processing is performed on credit risks through a high-dimensional situation prediction model. According to the invention, the accuracy of credit risk assessment is improved, the influence of the external environment on the cooperation credit risk between enterprises can be dynamically reflected, and more accurate and reliable enterprise credit early warning is realized.

Description

Technical Field

[0001] The present invention relates to the field of enterprise credit risk management, and in particular to a credit early warning method and system for cooperation between enterprises. Background Art

[0002] In the process of inter-enterprise cooperation, credit early warning is crucial to maintaining the trust relationship between enterprises. The credit early warning system can help enterprises identify potential credit risks early, so as to take appropriate countermeasures to prevent possible losses and adverse effects. In a complex business environment, the cooperation relationship between enterprises is not only affected by internal factors such as financial data and contract performance, but also by external variables such as market environment and industry changes. Therefore, it is of great practical significance to establish an effective credit early warning system that can dynamically evaluate and predict credit risks in inter-enterprise cooperation.

[0003] Existing enterprise credit risk early warning methods (Chinese invention patent, publication number: CN115496357A, name: Enterprise credit risk early warning method and system integrating forum text temporal features) usually rely on static data analysis models, such as Logit models and some traditional machine learning models; these methods mainly focus on the financial status and historical transaction data of the enterprise, ignoring the impact of dynamic changes in the external environment on the credit of the enterprise; at the same time, existing models have limitations in processing multidimensional data and nonlinear relationships, resulting in insufficient early warning accuracy and inability to fully capture complex dynamic changes in credit risk; in addition, existing methods lack in-depth consideration of the time dimension, especially when involving long-term cooperative relationships, and cannot provide accurate credit risk assessment. Summary of the invention

[0004] In view of the many problems existing in the above-mentioned prior art, the present invention provides a credit early warning method and system for inter-enterprise cooperation. The present invention constructs a high-dimensional situational prediction model by dynamically analyzing and integrating inter-enterprise cooperation data and external environmental variables to achieve multi-level and multi-dimensional accurate prediction and early warning of corporate credit risks; improves the accuracy of credit risk prediction, and can promptly reflect the impact of changes in the external environment on corporate credit, thereby providing enterprises with more timely and targeted risk response strategies.

[0005] A credit early warning method for inter-enterprise cooperation includes the following steps:

[0006] Collect financial transaction data, contract performance data, supply chain flow data, and social interaction data between enterprises from multiple data sources, clean and standardize the collected data, and generate standardized multi-dimensional interaction data;

[0007] The standardized multi-dimensional interaction data is input into the multi-dimensional trust evaluation model, and the trust of inter-enterprise cooperation is dynamically evaluated through the multi-dimensional trust evaluation model to generate multi-dimensional trust evaluation results; the external environmental variable data is integrated, and the historical and future trends of the external environmental variables are captured through the high-dimensional environmental response model, and the multi-dimensional trust evaluation results are integrated with the high-dimensional environmental response model to generate trust-environment optimization data;

[0008] The trust-environment optimization data is input into the high-dimensional scenario prediction model, and the trust-environment optimization data is processed at multiple levels through the adaptive multi-scale analysis and hierarchical fractal structure analysis of the high-dimensional scenario prediction model to generate joint credit risk prediction data;

[0009] The joint credit risk prediction data is input into the smart contract system, which triggers credit rating and risk control measures. The execution process of the smart contract system is dynamically adjusted through a hierarchical feedback control system to generate optimized credit ratings and final credit warning results.

[0010] Preferably, the multi-dimensional trust evaluation model calculates the trust of inter-enterprise cooperation through the following formula:

[0011]

[0012] Among them, S represents the trust degree of cooperation between enterprises, and its value is used to quantify the strength of trust relationship between enterprises; w i Represents the weight factor associated with different dimensional data, weight factor w i The size of determines the influence weight of each dimension data in the overall trust evaluation; i (D i ) represents the relationship between each dimension data D i The corresponding functional relationship, function D i Used to convert dimension data D i Converted into a numerical form suitable for trust evaluation; D i It represents the data of each dimension in the standardized multi-dimensional interaction data, including the standardized results of financial transaction data, contract fulfillment data, supply chain flow data and social interaction data between enterprises; n represents the number of dimensions involved in the trust calculation, representing the total number of different data dimensions involved in the analysis in the model.

[0013] Preferably, the weight factor w in the multidimensional trust evaluation model is i It is obtained by training the machine learning algorithm on the historical cooperation data, and the weight factor w is dynamically adjusted according to the new cooperation data. i , the process includes:

[0014] For each historical cooperation dataset, the weight factor w is trained using a supervised learning algorithm i , so that the trust predicted by the model is as close as possible to the actual trust;

[0015] As new cooperation data is continuously introduced, the weight factor w is adjusted using an online learning algorithm. i Perform real-time updates;

[0016] In the weight factor w i During the dynamic adjustment process, the system also considers the potential impact of external factors such as market fluctuations and industry changes on the trust relationship.

[0017] Preferably, the high-dimensional environmental response model generates environmental response field data using a multi-factor random field modeling method, wherein the multi-factor random field modeling method is represented by the following formula:

[0018] R = ∫ Ω h(x)×φ(x)dx

[0019] Among them, R represents environmental response, which means the comprehensive impact of different environmental variables on enterprise trust in the entire market environment; Ω represents the domain of environmental variables, which is the set of all potential variables in the market environment; h(x) represents the impact factor function of each environmental variable, which quantifies the specific impact of each environmental variable on enterprise trust; φ(x) represents the probability density function of environmental variables, which describes the distribution and occurrence probability of environmental variables in the market environment.

[0020] Preferably, the high-dimensional environmental response model further processes the environmental response field data through a distribution nested model to capture the historical and future trends of environmental variables, specifically including:

[0021] Conduct time series analysis on the historical data of each environmental variable to determine its past change patterns and generate historical trend data;

[0022] Using the distribution nesting model, historical trend data is projected into the future, the changing trend of future environmental variables is predicted, and future trend data is generated;

[0023] The predicted future trends are combined with historical trends to generate comprehensive environmental response field data that include historical and future trends. The comprehensive environmental response field data are then fused with multidimensional trust assessment results to generate trust-environment optimization data as input data for the high-dimensional scenario prediction model.

[0024] Preferably, the high-dimensional context prediction model analyzes the trust-environment optimization data by adaptive multi-scale analysis, wherein the multi-scale analysis is performed according to the following formula:

[0025]

[0026] Among them, P(t) represents the scenario prediction result at time t, which represents the credit risk prediction value of the enterprise at the future time point t; g j (t) represents the multi-scale analysis function related to time t, and the function g j (t) is used to describe the changes in situational characteristics at different time scales; j (λ j ) represents the decomposition function at the jth scale, and the decomposition function Ψ j (λ j ) is used to decompose situational features into influencing factors of different scales; j represents the parameter related to the jth scale, parameter λ j The decomposition weights of the situational features at this scale are adjusted to ensure that the model can flexibly respond to analysis needs at different scales; m represents the number of analysis scales, representing the number of multi-scale analysis levels applied in the model.

[0027] Preferably, the high-dimensional scenario prediction model further optimizes the scenario prediction results through hierarchical fractal structure analysis to capture the complex nonlinear relationship in the data, specifically including:

[0028] The scenario prediction results generated by multi-scale analysis are decomposed hierarchically to form fractal structures of different complexity;

[0029] By calculating the fractal dimension, the strength of nonlinear relationships in each hierarchy is quantified to identify underlying complex patterns in the data;

[0030] The nonlinear relationships at various levels are integrated to generate fractal-enhanced situational response field data, which include the results of multi-scale analysis and fractal structure analysis.

[0031] Preferably, the smart contract system dynamically adjusts the execution process of the smart contract system through a hierarchical feedback control system to generate an optimized credit rating and a final credit warning result, and the hierarchical feedback control system includes:

[0032] Data collection module, used to collect credit rating data and risk control measures execution data in real time during the execution of smart contracts;

[0033] Data analysis module, which identifies deviations and potential problems in the execution process by comparing the collected execution data with the expected results;

[0034] The dynamic adjustment module adjusts the execution strategy of the smart contract in real time according to the results of the data analysis module to ensure the optimal effect of credit rating and risk control measures.

[0035] Preferably, the feedback control in the smart contract system performs path-to-path interactive correction of data through a multi-path response mechanism, and performs cluster analysis on the corrected data through an iterative clustering algorithm, specifically including:

[0036] In the process of interaction correction between pathways, the response data generated by different pathways are compared to identify and correct the inconsistencies between the data, ensuring that all pathway data are consistent with each other in multiple dimensions;

[0037] Use iterative clustering algorithms to perform cluster analysis on the corrected data to identify the most representative credit risk patterns;

[0038] The cluster analysis results are output as the final credit warning data.

[0039] A system for implementing the credit early warning method for inter-enterprise cooperation, comprising:

[0040] The data collection and processing module is used to collect financial transaction data, contract performance data, supply chain flow data and social interaction data between enterprises from multiple data sources, and clean and standardize the collected data to generate standardized multi-dimensional interaction data;

[0041] A multi-dimensional trust evaluation module is used to input the standardized multi-dimensional interaction data into the multi-dimensional trust evaluation model, dynamically evaluate the trust of cooperation between enterprises through the multi-dimensional trust evaluation model, and generate a multi-dimensional trust evaluation result;

[0042] The environmental response fusion module is used to integrate external environmental variable data and capture the historical and future trends of external environmental variables through a high-dimensional environmental response model. It integrates the multi-dimensional trust assessment results with the high-dimensional environmental response model to generate trust-environment optimization data.

[0043] The scenario prediction module is used to input the trust-environment optimization data into the high-dimensional scenario prediction model, perform multi-level processing on the trust-environment optimization data through the adaptive multi-scale analysis and hierarchical fractal structure analysis of the high-dimensional scenario prediction model, and generate joint credit risk prediction data;

[0044] The smart contract execution module is used to input the joint credit risk prediction data into the smart contract system, trigger credit rating and risk control measures through the smart contract system, and dynamically adjust the execution process of the smart contract system through a hierarchical feedback control system to generate optimized credit ratings and final credit warning results.

[0045] Compared with the prior art, the advantages and beneficial effects of the present invention are:

[0046] The present invention uses a multi-dimensional trust assessment model to comprehensively consider multi-dimensional data in inter-enterprise cooperation, including financial transaction data, contract performance data, supply chain flow data, and inter-enterprise social interaction data, and dynamically assess the trust of inter-enterprise cooperation, thereby achieving a more accurate credit risk assessment.

[0047] The present invention captures and integrates the historical and future trends of external environmental variables through a high-dimensional environmental response model to generate trust-environment optimization data, thus solving the problem of insufficient consideration of external environmental factors in the prior art.

[0048] The present invention further optimizes credit risk prediction data through a high-dimensional scenario prediction model, using adaptive multi-scale analysis and hierarchical fractal structure analysis, and improves the multi-level processing capability of corporate credit risk, thereby enhancing the accuracy and reliability of credit warnings. BRIEF DESCRIPTION OF THE DRAWINGS

[0049] Figure 1 It is a schematic diagram of the process of the present invention;

[0050] Figure 2 It is a simplified flow chart of the multi-dimensional trust evaluation model in the present invention;

[0051] Figure 3 It is a simplified flow chart of the high-dimensional environmental response model in the present invention;

[0052] Figure 4 A simplified flow chart of feedback control of the smart contract system in the present invention;

[0053] Figure 5 It is a structural block diagram of the system of the present invention. DETAILED DESCRIPTION

[0054] Hereinafter, embodiments of the present disclosure will be described with reference to the accompanying drawings. However, it should be understood that these descriptions are exemplary only and are not intended to limit the scope of the present disclosure. In the following detailed description, for ease of explanation, many specific details are set forth to provide a comprehensive understanding of the embodiments of the present disclosure. However, it is obvious that one or more embodiments may also be implemented without these specific details. In addition, in the following description, descriptions of known structures and technologies are omitted to avoid unnecessary confusion of the concepts of the present disclosure.

[0055] The terms used herein are only for describing specific embodiments and are not intended to limit the present disclosure. The terms "comprise", "include", etc. used herein indicate the existence of the features, steps, operations and / or components, but do not exclude the existence or addition of one or more other features, steps, operations or components.

[0056] All terms (including technical and scientific terms) used herein have the meanings commonly understood by those skilled in the art, unless otherwise defined. It should be noted that the terms used herein should be interpreted as having a meaning consistent with the context of this specification, and should not be interpreted in an idealized or overly rigid manner.

[0057] like Figure 1 As shown, a credit early warning method for cooperation between enterprises includes the following steps:

[0058] Collect financial transaction data, contract performance data, supply chain flow data, and social interaction data between enterprises from multiple data sources, clean and standardize the collected data, and generate standardized multi-dimensional interaction data;

[0059] Financial transaction data reflects the economic exchanges and financial stability between enterprises. These data usually include payment records, invoices, credit line usage, etc. By analyzing this information, we can judge a company's payment ability and credit habits. Contract performance data focuses on whether the company fulfills its contractual obligations on time and in quantity. The timeliness and accuracy of contract performance are important bases for judging the company's performance ability and credibility. Supply chain flow data shows the role and behavior patterns of enterprises in the supply chain, especially logistics information, inventory management and supplier relationships. Analyzing these data can identify potential risks in the supply chain and the business stability of enterprises. Inter-enterprise social interaction data covers informal communication and cooperation between enterprises, including emails, meeting minutes and other communication records. These data help supplement the cooperative relationships and potential risks that cannot be reflected in traditional financial and contract data.

[0060] After data collection is completed, the system will clean and standardize the data. Data cleaning is to remove noise, duplication and invalid data, ensure the quality of input data, and thus improve the accuracy of model evaluation. Standardization is to unify data from different sources into a consistent format and measurement unit, so that the data can be effectively compared and integrated in subsequent analysis models. For example, financial data may need to unify the currency unit, and time data may need to be standardized to the same time format. This process not only improves the operability of the data, but also ensures the consistency of the data in the subsequent analysis process.

[0061] Through this series of processing, the system generates standardized multi-dimensional interaction data. This multi-dimensional interaction data set lays the foundation for subsequent trust assessment and environmental response analysis. It contains information on all aspects of corporate cooperation and forms a comprehensive and systematic corporate credit portrait. These data can better reflect the real cooperation status between enterprises and provide a reliable basis for credit warning. Ultimately, the standardized data ensures that in the subsequent multi-dimensional trust assessment and scenario prediction, the system can capture and warn potential risks in inter-enterprise cooperation in real time with high accuracy and stability.

[0062] The standardized multi-dimensional interaction data is input into the multi-dimensional trust evaluation model, and the trust of inter-enterprise cooperation is dynamically evaluated through the multi-dimensional trust evaluation model to generate multi-dimensional trust evaluation results; the external environmental variable data is integrated, and the historical and future trends of the external environmental variables are captured through the high-dimensional environmental response model, and the multi-dimensional trust evaluation results are integrated with the high-dimensional environmental response model to generate trust-environment optimization data;

[0063] In the present invention, the standardized multi-dimensional interaction data is input into the multi-dimensional trust assessment model, the core function of which is to dynamically evaluate the trust of inter-enterprise cooperation. The multi-dimensional trust assessment model generates an overall trust assessment result by comprehensively analyzing data of different dimensions, such as financial transaction data, contract performance data, supply chain flow data, and social interaction data between enterprises. This trust assessment is based on the weight calculation of multiple factors. These weight factors are obtained through historical data training and are dynamically adjusted with the introduction of new data, thereby ensuring that the model can promptly reflect changes in inter-enterprise cooperation relationships. For example, if a company has shown a high credit rating in past financial transactions, but has frequently experienced delayed delivery in contract performance recently, the model will automatically adjust the weights of each dimension of data, thereby reflecting the reduction in the trust of the company.

[0064] After generating the multi-dimensional trust assessment results, the system further integrates the external environmental variable data and captures the historical and future trends of the external environmental variables through the high-dimensional environmental response model. External environmental variables may include market fluctuations, policy changes, industry dynamics, and macroeconomic indicators, which usually have a significant impact on the credit risk of enterprises. The high-dimensional environmental response model uses a multi-factor random field and distribution nested model to predict the changing trend of future environmental variables through time series analysis of historical data, and integrates this information with the trust assessment results of the enterprise. In this way, the system can not only reflect the internal credit status of the enterprise in real time, but also take into account the changes in the external market environment, so as to more comprehensively assess the credit risk of the enterprise. For example, in an industry with volatile market conditions, even if the company's own financial data and contract performance are good, the high-dimensional environmental response model may still downgrade the company's credit rating due to the deterioration of the external environment to warn of potential systemic risks.

[0065] By integrating the results of multi-dimensional trust assessment with the high-dimensional environmental response model, the system generates trust-environment optimization data. This data not only contains an accurate assessment of the trust in cooperation between enterprises, but also combines the potential impact of the external environment on the credit of enterprises, thus forming a more comprehensive credit risk portrait. The effect of this process is that it greatly improves the accuracy and reliability of credit warnings. By comprehensively considering multiple variables inside and outside the enterprise, the system can identify possible credit risks earlier and more accurately, thereby helping enterprises to be more cautious when making cooperation decisions and reduce losses caused by fluctuations in the external environment or internal credit problems. In practical applications, this approach enables enterprises to maintain robust operations in a complex and changing market environment, improve risk management, and enhance market competitiveness.

[0066] Preferably, the multi-dimensional trust evaluation model calculates the trust of inter-enterprise cooperation through the following formula:

[0067]

[0068] Among them, S represents the trust degree of cooperation between enterprises, and its value is used to quantify the strength of trust relationship between enterprises; w i Represents the weight factor associated with different dimensional data, weight factor w i The size of determines the influence weight of each dimension data in the overall trust evaluation; i (D i ) represents the relationship between each dimension data D i The corresponding functional relationship, function D i Used to convert dimension data D i Converted into a numerical form suitable for trust evaluation; D iIt represents the data of each dimension in the standardized multi-dimensional interaction data, including the standardized results of financial transaction data, contract fulfillment data, supply chain flow data and social interaction data between enterprises; n represents the number of dimensions involved in the trust calculation, representing the total number of different data dimensions involved in the analysis in the model.

[0069] In the above formula, S represents the trust in the cooperation between enterprises, and its value is used to quantify the strength of the trust relationship between enterprises. The higher the trust, the more stable the cooperation relationship between the two enterprises and the lower the risk of default; conversely, low trust indicates that the cooperation may have higher risks. i is the weight factor associated with different dimensional data, the weight factor w i The size of directly affects the role of each dimension of data in the overall trust assessment. For example, in some corporate cooperation, financial transaction data may be given a higher weight because the financial status directly reflects the economic strength and debt repayment ability of the enterprise; while in other cooperation, contract performance data may be more critical and therefore have a higher weight.

[0070] Function f i (D i ) represents the relationship between each dimension data D i The corresponding functional relationship. The main function of this function is to convert the dimension data D i Converted into a numerical form suitable for trust evaluation. Because the data of different dimensions have different properties, for example, financial transaction data is in digital form, while social interaction data may be text or event records, the function f i (D i ) must be designed to take into account the characteristics of the data so that all data can be normalized to comparable values. In practical applications, f i (D i ) may use different processing methods, such as normalization, standardization or other data transformation techniques. For example, when processing financial data, f i (D i ) may apply a logarithmic transformation to reduce the influence of extreme values ​​on the results, while in social interaction data, f i (D i ) might use sentiment analysis techniques to convert text data into scores reflecting cooperative attitudes.

[0071] D iRepresents the data of each dimension in the standardized multi-dimensional interaction data. Data standardization is to eliminate the dimensional differences between different data dimensions so that they can be compared and analyzed on a unified scale. In this scheme, the standardized data includes the standardized results of financial transaction data, contract fulfillment data, supply chain flow data, and social interaction data between enterprises. These data reflect various aspects of enterprise cooperation and constitute the basic data input for trust assessment.

[0072] Finally, n represents the number of dimensions involved in the trust calculation, which represents the total number of different data dimensions analyzed in the model. In this invention, the trust evaluation of inter-enterprise cooperation not only considers conventional financial and contract data, but also incorporates dimensions such as supply chain and social interaction, which fully reflects the advantages of multi-dimensional analysis.

[0073] Through this calculation formula, the present invention can comprehensively evaluate the cooperation trust between enterprises and provide more accurate credit warning than single-dimensional analysis. The advantage of this model lies in its flexibility and dynamism: the weight factor w i It can be adjusted according to different corporate relationships and cooperation backgrounds, and the function f i (D i ) ensures the adaptability of data processing. This trust assessment model effectively improves the accuracy of credit warnings, helps companies better identify potential risks in cooperation, optimizes decision-making processes, and reduces possible credit losses. In practice, companies that apply this model can better control risks, improve the quality of cooperation, and achieve sustained and stable growth in a complex and changing market environment.

[0074] like Figure 2 As shown, preferably, the weight factor w in the multidimensional trust evaluation model i It is obtained by training the machine learning algorithm on the historical cooperation data, and the weight factor w is dynamically adjusted according to the new cooperation data. i , the process includes:

[0075] For each historical cooperation data set, the weight factor w is trained using a supervised learning algorithm i , so that the trust predicted by the model is as close as possible to the actual trust;

[0076] As new cooperation data is continuously introduced, the weight factor w is adjusted using an online learning algorithm. i Carry out real-time updates;

[0077] In the weight factor w i During the dynamic adjustment process, the system also considers the potential impact of external factors such as market fluctuations and industry changes on the trust relationship.

[0078] Weight factor wi The initial setting is based on historical cooperation data sets. At this stage, the system uses a supervised learning algorithm to train the weight factors, with the goal of making the trust predicted by the model as close to the actual trust as possible. The core of the supervised learning algorithm is to train the model through the annotation results in the historical data (that is, the actual cooperation results), so that it can learn from the input data the degree of influence of each dimension on the final trust. For example, the system may identify that financial transaction data has a greater influence in a certain type of cooperation by analyzing multiple successful and failed cooperation cases, while in another type of cooperation, contract performance data may be more critical. Through this training process, the system is able to assign appropriate weights to each dimension, thereby making more accurate predictions in future cooperation evaluations.

[0079] As new cooperation data is continuously introduced, the system needs to maintain the ability to dynamically adjust the weight factor. At this time, the online learning algorithm plays an important role. The online learning algorithm allows the model to process and learn new data in real time without stopping operation, thereby continuously updating the weight factor w. i This real-time update mechanism is very important, especially when the market environment changes rapidly, and the cooperative relationships and risk factors between enterprises may change at any time. Through online learning, the system can quickly adapt to these changes and reflect them in the trust evaluation. For example, in a certain period, when market volatility intensifies, the system can adjust the weight factor through online learning, so that the weight of supply chain flow data in the trust evaluation increases, thereby timely warning of potential cooperation risks.

[0080] In addition, the weight factor w i In the dynamic adjustment process, the system not only considers the impact of historical data, but also takes into account external factors such as market fluctuations and industry changes. The purpose of this is to ensure that the model can more comprehensively consider the external environment that may affect the cooperative relationship when predicting trust. For example, during an economic recession, although a company's historical cooperation data has performed well, the system may reduce the company's trust weight due to the deterioration of the external market environment, thereby issuing an early warning. This consideration of external factors makes the trust assessment model not only data-driven, but also able to flexibly respond to changes in the external environment, improving the accuracy and practicality of predictions.

[0081] Preferably, the high-dimensional environmental response model generates environmental response field data using a multi-factor random field modeling method, wherein the multi-factor random field modeling method is represented by the following formula:

[0082] R = ∫ Ω h(x)×φ(x)dx

[0083] Among them, R represents environmental response, which means the comprehensive impact of different environmental variables on enterprise trust in the entire market environment; Ω represents the domain of environmental variables, which is the set of all potential variables in the market environment; h(x) represents the impact factor function of each environmental variable, which quantifies the specific impact of each environmental variable on enterprise trust; φ(x) represents the probability density function of environmental variables, which describes the distribution and occurrence probability of environmental variables in the market environment.

[0084] In the present invention, the high-dimensional environmental response model is an important tool for analyzing and quantifying the impact of the external market environment on the trust of inter-enterprise cooperation. The model generates environmental response field data through a multi-factor random field modeling method, thereby providing a quantitative analysis of the comprehensive effects of various variables in the market environment. This analysis plays a key role in credit early warning, especially when the external environment fluctuates greatly.

[0085] In the above formula, R represents environmental response, which is the quantitative result of the comprehensive impact of different variables in the market environment on the trust of enterprises. Environmental response R is an important quantitative indicator that reflects how external environmental factors affect the cooperation stability and credit risk of enterprises. For example, in a market environment, factors such as policy changes, economic fluctuations and industry dynamics will have different degrees of impact on the credit status of enterprises, and these impacts are ultimately reflected comprehensively through the environmental response model.

[0086] Ω represents the definition domain of the environmental variables, which covers all potential variables in the market environment and is the basis of the environmental response model. For example, the definition domain Ω can include macroeconomic indicators (such as GDP growth rate, inflation rate), industry-specific market dynamics (such as raw material price fluctuations, market demand changes), and policy factors (such as tax policy adjustments, trade tariff changes). By determining this definition domain, the model can comprehensively consider all variables in the market that may affect corporate credit.

[0087] h(x) represents the impact factor function of each environmental variable, which is used to quantify the specific impact of each environmental variable on the trust of the enterprise. The role of the impact factor function h(x) is to convert the impact of different environmental variables into measurable values. For example, some variables may have a direct impact on the trust of the enterprise. For example, rising interest rates may increase the financing costs of the enterprise, thereby reducing its ability to perform its obligations. Through the impact factor function h(x), the model can accurately capture and quantify the impact of these variables.

[0088] φ(x) represents the probability density function of environmental variables, which is used to describe the distribution and probability of occurrence of environmental variables in the market environment. The introduction of the probability density function φ(x) allows the model to consider not only the impact strength of each variable, but also the possibility of these variables appearing in different market scenarios. For example, some market fluctuations may be short-term accidental events with a low probability of occurrence, while other factors may be long-term, persistent sources of stress with a high probability of occurrence. Through φ(x), the model can more accurately reflect the comprehensive impact of the market environment on corporate trust.

[0089] During the implementation process, this high-dimensional environmental response model can effectively capture and quantify the impact of complex market environments on corporate partnerships in practical applications. For example, in some industries, such as manufacturing or industries with high dependence on raw materials, fluctuations in raw material prices in the external market are often a key factor affecting corporate credit. Through this model, the system can identify the specific impact of raw material price fluctuations on corporate trust and incorporate it into the overall credit risk assessment. This enables companies to better foresee potential credit risks when facing drastic changes in the market environment, take countermeasures in advance, and reduce losses.

[0090] like Figure 3 As shown, preferably, the high-dimensional environmental response model further processes the environmental response field data through a distribution nested model to capture the historical and future trends of environmental variables, specifically including:

[0091] Conduct time series analysis on the historical data of each environmental variable to determine its past change patterns and generate historical trend data;

[0092] Using the distribution nesting model, historical trend data is projected into the future, the changing trend of future environmental variables is predicted, and future trend data is generated;

[0093] The predicted future trends are combined with historical trends to generate comprehensive environmental response field data that include historical and future trends. The comprehensive environmental response field data are then fused with multidimensional trust assessment results to generate trust-environment optimization data as input data for the high-dimensional scenario prediction model.

[0094] The system performs time series analysis on the historical data of each environmental variable. The core principle of this step is to generate historical trend data by analyzing the changing patterns of environmental variables over a period of time. Time series analysis is a common statistical method used to identify the changing patterns of data over time, such as cyclical fluctuations, long-term trends, or the impact of emergencies on data. In the present invention, time series analysis helps the system understand the historical behavior of environmental variables, such as seasonal fluctuations in raw material prices, long-term trends in interest rates, or cyclical changes in industry demand. These historical trend data not only reflect the past behavior patterns of the variables, but also provide a basis for future predictions.

[0095] The system uses the distribution nesting model to project historical trend data into the future and generate future trend data. The distribution nesting model is an advanced statistical method that predicts the future trend of environmental variables by combining known historical trends with future uncertainties. Unlike simple time series forecasts, the distribution nesting model can handle more complex situations with higher uncertainty to ensure the reliability of the forecast. For example, in an extreme market environment, such as an economic crisis or policy mutation, the distribution nesting model can generate possible future trend data by simulating different future scenarios. This method not only takes into account the continuity of historical data, but also introduces the processing of future uncertainties, making the forecast results more flexible and responsive.

[0096] The system combines the predicted future trends with historical trends to generate comprehensive environmental response field data that includes historical and future trends. This comprehensive data integrates all information from the past and the future, forming a complete understanding of environmental variables. The effect of this step is that the system can more accurately grasp the evolution of environmental variables in the time dimension, thereby providing a more comprehensive environmental analysis for corporate credit warnings. For example, when predicting the market demand of a certain industry, the comprehensive environmental response field data can reflect the past market demand trends and possible future changes, helping companies adjust their strategies in advance and avoid potential cooperation risks.

[0097] The system fuses the generated comprehensive environmental response field data with the multi-dimensional trust assessment results to generate trust-environment optimization data. The principle of this data fusion process is to combine the internal trust assessment of inter-enterprise cooperation with the comprehensive analysis of the external environment to form a more accurate credit warning result. The trust-environment optimization data is used as input data for the high-dimensional scenario prediction model, which directly affects the subsequent credit risk prediction. Through this fusion, the present invention can ensure that enterprises always have a comprehensive understanding of cooperative relationships and early warning capabilities in a complex and changing market environment.

[0098] For example, in a particular market, assuming that raw material prices experience unprecedented drastic fluctuations, time series analysis may show short-term extreme fluctuations, but through the distribution nesting model, the system can predict that such fluctuations may continue and have a wider impact in the future. Combining this prediction result with the company's financial status and contract performance capabilities, the system may conclude that the risk of cooperation has increased, thereby triggering a credit warning and helping the company take timely measures to prevent possible credit risks.

[0099] The trust-environment optimization data is input into the high-dimensional scenario prediction model, and the trust-environment optimization data is processed at multiple levels through the adaptive multi-scale analysis and hierarchical fractal structure analysis of the high-dimensional scenario prediction model to generate joint credit risk prediction data;

[0100] Trust-environment optimization data is input into a high-dimensional scenario prediction model. Adaptive multi-scale analysis is a key technology in the model. Its core principle is to capture long-term trends and short-term fluctuations in the data by analyzing data at different time scales. This multi-scale analysis method can identify the characteristics of data at different time scales. For example, short-term market fluctuations may have an immediate impact on corporate cooperation, while long-term trends may reflect deep changes in the market or cyclical laws of the industry. In the context of credit warning, adaptive multi-scale analysis can help the system identify potential credit risks, regardless of whether they are caused by short-term market fluctuations or long-term trends. This multi-scale processing method is particularly suitable for complex market environments and can provide more accurate credit risk predictions.

[0101] Hierarchical fractal structure analysis further processes these multi-scale data. Fractal structure analysis is a method used to reveal complex nonlinear relationships in data. Fractal structures are self-similar, which means that data may show similar patterns at different scales. In credit warning, corporate behavior or market response is often not linear, so traditional linear analysis methods may not accurately capture these complex relationships. Through fractal structure analysis, the system is able to identify hidden patterns in the data, such as consistent behavior that companies may exhibit when facing similar market conditions, or different responses in different market environments. In this way, the system can more accurately predict the credit risk of inter-enterprise cooperation.

[0102] In this process, after multi-scale and fractal structure analysis, the trust-environment optimization data generates joint credit risk prediction data. This data not only contains the comprehensive analysis results of the enterprise's trust and external environment, but also reveals the complex relationships and potential risks in the data through multi-level processing. The joint credit risk prediction data is the final output of the present invention and is directly used for decision support of the enterprise credit early warning system. Through this data, enterprises can identify potential risks in cooperation earlier, such as possible defaults or emergencies in the market environment, so as to take preventive measures and reduce credit losses.

[0103] For example, in a certain cooperation project, if the trust-environment optimization data shows that the company's financial situation is good, but the market environment shows large fluctuations, through adaptive multi-scale analysis, the system may identify the potential impact of market fluctuations on the company's cash flow in the short term, and through fractal structure analysis, the system may find that the company has historically shown a pattern of financial constraints in similar market fluctuations. Combined with these analysis results, the joint credit risk prediction data may indicate that the company has potential default risks, thereby issuing a credit warning and prompting the management to take countermeasures such as adjusting the financing strategy in advance or increasing capital reserves.

[0104] Preferably, the high-dimensional context prediction model analyzes the trust-environment optimization data by adaptive multi-scale analysis, wherein the multi-scale analysis is performed according to the following formula:

[0105]

[0106] Among them, P(t) represents the scenario prediction result at time t, which represents the credit risk prediction value of the enterprise at the future time point t; g j (t) represents the multi-scale analysis function related to time t, and the function g j (t) is used to describe the changes in situational characteristics at different time scales; j (λ j ) represents the decomposition function at the jth scale, and the decomposition function Ψ j (λ j ) is used to decompose situational features into influencing factors of different scales; j represents the parameter related to the jth scale, parameter λ j The decomposition weights of the situational features at this scale are adjusted to ensure that the model can flexibly respond to analysis needs at different scales; m represents the number of analysis scales, representing the number of multi-scale analysis levels applied in the model.

[0107] In the present invention, the high-dimensional scenario prediction model analyzes the trust-environment optimization data through adaptive multi-scale analysis, and its core purpose is to accurately predict the credit risk of enterprises at future time points. Multi-scale analysis is a technology that can decompose and process data at different time scales. It is essential for identifying short-term fluctuations and long-term trends, and capturing complex change patterns in data.

[0108] formula It is the mathematical basis of this analysis method, where P(t) represents the scenario prediction result at time t, that is, the credit risk prediction value of the enterprise at a specific point in the future. This prediction value directly reflects the risk level of the enterprise in future cooperation and is the key output of the credit early warning system.

[0109] In the formula, g j (t) is a multi-scale analysis function related to time t. This function is used to describe the changes in situational characteristics at different time scales. In the application of credit warning, the credit risk of an enterprise may be affected by multiple time scale factors, such as short-term market fluctuations, quarterly performance, or long-term industry trends. j (t) function, the system is able to analyze these factors at different time scales, ensuring that not only recent sharp fluctuations are taken into account, but also the impact of long-term trends on corporate credit is captured.

[0110] Ψ j (λ j ) represents the decomposition function at the jth scale, which decomposes the situational features into influencing factors at different scales. In practical applications, this means that the system can identify and decompose various factors that affect the credit risk of an enterprise and process them according to their importance at different time scales. For example, in the short term, raw material price fluctuations may be the main influencing factor of the enterprise's credit risk, while in the long term, the capital structure or industry cycle of the enterprise may be more critical. j (λ j ) decomposition function, the system can flexibly respond to the influencing factors at different time scales, thereby generating more accurate situational prediction results.

[0111] Parameter λ j It is used to adjust the decomposition weight at the jth scale to ensure that the model can flexibly respond to analysis needs at different scales. Adjusting this parameter enables the system to assign different importance to factors at different time scales according to actual conditions. For example, in a rapidly changing market environment, short-term factors may require higher weights, while in a relatively stable market, long-term trends may be more important. By adjusting λ j parameters, the system can adapt to different market environments, thus maintaining the accuracy and relevance of forecasts.

[0112] Finally, m represents the number of analysis scales, which represents the number of multi-scale analysis levels applied in the model. This number determines how many time scales of situational features the system can handle, thus affecting the sophistication and depth of the forecast. In actual applications, the number of analysis scales needs to be set according to the specific application scenario. Usually, when dealing with complex market environments, more analysis scales can provide more detailed risk analysis.

[0113] For example, suppose a company is working with a partner with an unstable supply chain. Trust-environment optimization data has shown that the partner has performed well in past contract performance, but the current market environment shows drastic fluctuations in raw material prices. Through adaptive multi-scale analysis, the system may identify that short-term fluctuations in raw material prices have a greater impact on the partner's ability to fulfill its contract, so in the short term, g j (t) and Ψ j (λ j ) will increase. On a long-term scale, the system may find that the capital structure of the enterprise is relatively stable, so it assigns a higher weight to the long-term factor. Finally, the system combines the multi-scale analysis results to generate a scenario prediction result P(t), which indicates that the partner may face a higher performance risk in the short term, thereby triggering a corresponding credit warning.

[0114] Preferably, the high-dimensional scenario prediction model further optimizes the scenario prediction results through hierarchical fractal structure analysis to capture the complex nonlinear relationship in the data, specifically including:

[0115] The scenario prediction results generated by multi-scale analysis are decomposed hierarchically to form fractal structures of different complexity;

[0116] By calculating the fractal dimension, the strength of nonlinear relationships in each hierarchy is quantified to identify underlying complex patterns in the data;

[0117] The nonlinear relationships at various levels are integrated to generate fractal-enhanced situational response field data, which include the results of multi-scale analysis and fractal structure analysis.

[0118] The system will hierarchically decompose the scenario prediction results generated by multi-scale analysis to form fractal structures of different complexity. Fractal structure is a concept used in mathematics to describe certain geometric shapes with self-similarities and complex forms. In credit risk prediction, the behavior of enterprises and market reactions often have multiple levels of complexity, which manifests itself under different time scales and market conditions. Through hierarchical decomposition, the system can decompose these complex credit risk characteristics into multiple different levels, each representing a specific level of complexity. For example, when a company faces a relatively stable market environment, it may show a consistent credit behavior pattern, while when the market fluctuates violently, its behavior pattern may become more complex and unpredictable. Through the hierarchical decomposition of the fractal structure, the system can capture these different levels of behavior characteristics separately, thereby making a more detailed analysis of credit risk.

[0119] The system quantifies the strength of nonlinear relationships in each hierarchy by calculating the fractal dimension. Fractal dimension is a mathematical indicator used to quantify the complexity of fractal structures. In the present invention, it is used to evaluate the complexity and nonlinearity of credit risk characteristics at each level. Nonlinear relationships are usually not captured by traditional linear models, and these relationships often manifest as asymmetric responses of enterprises under different market conditions. For example, a company may gradually improve its credit during a period of economic growth, while its credit deteriorates rapidly during a recession. This nonlinear behavior can be quantified and identified by the fractal dimension. By calculating the fractal dimension, the system can accurately evaluate the strength of these nonlinear relationships and thus identify potential credit risk patterns.

[0120] The system integrates nonlinear relationships at all levels to generate fractal-enhanced situational response field data. This data set not only contains the results of multi-scale analysis, but also integrates the deep information of fractal structure analysis to form a comprehensive and comprehensive credit risk assessment result. The core of this integration process is that the system can make a comprehensive analysis of the credit risk of the enterprise at different complexities and time scales and generate a unified risk assessment result. Fractal-enhanced situational response field data can provide more accurate and comprehensive prediction results than traditional risk assessment methods, allowing enterprises to better foresee and manage potential risks when making cooperation decisions.

[0121] For example, in a long-term supply chain, a company's credit performance may perform well under stable market conditions, but show significant fluctuations when the market is turbulent. Through fractal structure analysis, the system can identify the complex nonlinear relationships behind these fluctuations, such as how the company's credit behavior pattern changes when raw material prices fluctuate sharply. By integrating this information, fractal-enhanced situational response field data can warn of possible credit risks that may arise in the future for the company, prompting partners to take preventive measures, such as re-evaluating credit limits or requiring stricter contract terms.

[0122] like Figure 4 As shown, the joint credit risk prediction data is input into the smart contract system, the credit rating and risk control measures are triggered by the smart contract system, and the execution process of the smart contract system is dynamically adjusted through a hierarchical feedback control system to generate optimized credit ratings and final credit warning results.

[0123] Preferably, the smart contract system dynamically adjusts the execution process of the smart contract system through a hierarchical feedback control system to generate an optimized credit rating and a final credit warning result, and the hierarchical feedback control system includes:

[0124] Data collection module, used to collect credit rating data and risk control measures execution data in real time during the execution of smart contracts;

[0125] Data analysis module, which identifies deviations and potential problems in the execution process by comparing the collected execution data with the expected results;

[0126] The dynamic adjustment module adjusts the execution strategy of the smart contract in real time according to the results of the data analysis module to ensure the optimal effect of credit rating and risk control measures.

[0127] The data acquisition module included in the system is responsible for collecting credit rating data and risk control measures execution data in real time during the execution of smart contracts. The data acquisition module is the basis of the entire feedback control system, which ensures that the system can obtain key data in the execution process in real time. These data include the company's current credit rating, the progress of contract execution, and the actual effect of control measures taken against potential risks. By collecting this data in real time, the system can have a clear understanding of the execution of smart contracts, thereby providing reliable basic data for subsequent analysis and adjustments.

[0128] The data analysis module analyzes the collected execution data and identifies deviations and potential problems in the execution process by comparing the actual execution results with the expected results. The core principle of the data analysis module is difference analysis, that is, identifying possible deviations in the execution of smart contracts, such as changes in credit ratings that do not meet expectations, or the effects of risk control measures that do not meet expected standards. Through this analysis, the system can discover potential execution problems and identify factors that may affect credit ratings. For example, if a company has an unexpected default during the execution of a smart contract, the data analysis module can immediately identify this deviation and mark it as a potential credit risk.

[0129] Based on the results of the data analysis module, the dynamic adjustment module adjusts the execution strategy of the smart contract in real time to ensure the best effect of credit rating and risk control measures. The role of the dynamic adjustment module is to enable the system to flexibly respond to various problems that arise during the execution process. By adjusting the execution strategy in real time, the system can optimize the path of contract execution based on the latest execution data and analysis results. For example, if the system finds that a certain risk control measure is not effective, the dynamic adjustment module can immediately adjust the measure or introduce new control measures to ensure that the final credit rating and warning results can reflect the current real risk level.

[0130] This hierarchical feedback control system is effective in credit early warning. Through real-time data collection, accurate data analysis and flexible dynamic adjustment, the system can ensure the accuracy and effectiveness of credit risk management at every step of the smart contract execution. For example, in a certain cooperation project, the system can monitor the changes in the company's credit rating and dynamically adjust the contract terms according to the actual execution situation, or introduce new risk control measures in a timely manner to prevent potential credit risks. This real-time feedback and adjustment mechanism enables enterprises to maintain the effectiveness of credit risk management in the face of an uncertain market environment and reduce losses caused by failure to identify and handle credit risks in a timely manner.

[0131] Preferably, the feedback control in the smart contract system performs path-to-path interactive correction of data through a multi-path response mechanism, and performs cluster analysis on the corrected data through an iterative clustering algorithm, specifically including:

[0132] In the process of interaction correction between pathways, the response data generated by different pathways are compared to identify and correct the inconsistencies between the data, ensuring that all pathway data are consistent with each other in multiple dimensions;

[0133] Use iterative clustering algorithms to perform cluster analysis on the corrected data to identify the most representative credit risk patterns;

[0134] The cluster analysis results are output as the final credit warning data.

[0135] The multi-path response mechanism plays a key role in the data processing process. The mechanism compares the response data generated on different paths to identify and correct inconsistencies between data. The core principle of inter-path interaction correction is to ensure that data obtained from different sources or different calculation paths are consistent with each other in multiple dimensions. This is crucial in credit early warning systems because data inconsistency may lead to incorrect credit risk assessments. For example, a company's financial data and contract performance data may be processed through different paths respectively. If these two types of data deviate during the processing process, the system may produce inaccurate credit rating results. Through inter-path interaction correction, the system can identify these inconsistencies and make necessary corrections to ensure that the final data is consistent in all dimensions, thereby improving the accuracy of the assessment.

[0136] The corrected data will be clustered and analyzed using an iterative clustering algorithm. The iterative clustering algorithm is a commonly used machine learning method for grouping large amounts of data to identify the most representative patterns. In the present invention, the goal of cluster analysis is to identify those patterns that best represent the credit risk of the enterprise from a large amount of corrected data. Through this analysis, the system can extract core information from complex data sets and form an accurate judgment of the credit risk of the enterprise. For example, when processing the cooperation data of multiple enterprises, the iterative clustering algorithm can identify which enterprises show similar credit risk patterns under similar market conditions, thereby helping the system predict the credit risks of these enterprises in future cooperation.

[0137] Through this cluster analysis, the system is able to simplify complex data sets into a set of key credit risk patterns and output these patterns as the final credit warning data. This approach not only improves the accuracy of credit warnings, but also makes credit risk assessment results more intuitive and easy to understand. Based on these results, business managers can quickly identify the risk points that need the most attention and take appropriate risk control measures. For example, if the cluster analysis results show that a certain type of enterprise tends to show a higher risk of default when the supply chain is unstable, the enterprise can adjust its supply chain strategy in advance to avoid potential losses.

[0138] For example, in a global supply chain cooperation project, data from multiple suppliers may enter the credit early warning system through different paths. These data may include financial status, contract performance records, market environment, etc. In the process of data processing, due to the different calculation methods and data sources of different paths, certain inconsistencies may occur. If not corrected, these inconsistencies may lead to errors in credit risk assessment. Through the multi-path response mechanism, the system can identify and correct the inconsistencies of these data to ensure that the data of all paths are consistent in dimension. Subsequently, the system uses an iterative clustering algorithm to cluster and analyze these corrected data to identify the most representative credit risk patterns. Finally, these patterns are output as credit early warning data, providing accurate risk assessment information for corporate managers, enabling them to make more informed decisions.

[0139] like Figure 5 As shown, a system for implementing the credit early warning method for inter-enterprise cooperation includes:

[0140] The data collection and processing module is used to collect financial transaction data, contract performance data, supply chain flow data, and social interaction data between enterprises from multiple data sources, and clean and standardize the collected data to generate standardized multi-dimensional interaction data; the data collection and processing module is responsible for collecting key data on inter-enterprise cooperation from multiple data sources, including financial transactions, contract performance, supply chain flow, and social interaction between enterprises. These data are cleaned and standardized and converted into multi-dimensional interaction data in a unified format. This process ensures the accuracy and comparability of the data, providing a solid foundation for subsequent evaluation.

[0141] The multidimensional trust assessment module is used to input the standardized multidimensional interaction data into the multidimensional trust assessment model, dynamically evaluate the trust of inter-enterprise cooperation through the multidimensional trust assessment model, and generate a multidimensional trust assessment result; the multidimensional trust assessment module receives the processed multidimensional interaction data and dynamically calculates the trust of inter-enterprise cooperation through the multidimensional trust assessment model. The model comprehensively considers data from different dimensions, adjusts the weight factors in real time, and generates trust assessment results that reflect the cooperative relationship between enterprises. Through this module, the system can quickly identify potential credit risks in cooperation and provide accurate trust assessment.

[0142] The environmental response fusion module is used to integrate external environmental variable data, and capture the historical and future trends of external environmental variables through a high-dimensional environmental response model, and fuse the multi-dimensional trust assessment results with the high-dimensional environmental response model to generate trust-environment optimization data; the environmental response fusion module combines the multi-dimensional trust assessment results with external environmental variable data, and captures the historical trends and future changes of the external environment through a high-dimensional environmental response model, and fuses these data to generate trust-environment optimization data. This process enables the system to consider the impact of the external environment on corporate credit, thereby providing a more comprehensive perspective for credit risk assessment.

[0143] The scenario prediction module is used to input the trust-environment optimization data into the high-dimensional scenario prediction model, and perform multi-level processing on the trust-environment optimization data through the adaptive multi-scale analysis and hierarchical fractal structure analysis of the high-dimensional scenario prediction model to generate joint credit risk prediction data; The scenario prediction module is responsible for inputting the trust-environment optimization data into the high-dimensional scenario prediction model, and perform multi-level processing on these data through adaptive multi-scale analysis and hierarchical fractal structure analysis to generate joint credit risk prediction data. The role of this module is to accurately predict the credit risk in inter-enterprise cooperation through multi-scale and multi-level analysis, so that the system can warn of possible risk events in advance.

[0144] The smart contract execution module is used to input the joint credit risk prediction data into the smart contract system, trigger the credit rating and risk control measures through the smart contract system, and dynamically adjust the execution process of the smart contract system through a hierarchical feedback control system to generate optimized credit ratings and final credit warning results. The smart contract execution module uses the joint credit risk prediction data to trigger the corresponding credit rating and risk control measures through the smart contract system. In addition, the system dynamically adjusts the execution of smart contracts in real time through a hierarchical feedback control system to ensure the best effect of credit ratings and risk control measures. The module ultimately generates optimized credit ratings and credit warning results, providing real-time risk management decision support for enterprises.

[0145] Those skilled in the art will appreciate that the embodiments of the present application may be provided as methods, systems or computer program products. Therefore, the present application may take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware.

[0146] The above are only embodiments of the present application and are not intended to limit the present application. For those skilled in the art, the present application may have various changes and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application should be included in the scope of the claims of the present application.

Claims

1. A credit warning method for inter-enterprise cooperation, characterized in that: The following steps are involved: Collect financial transaction data, contract performance data, supply chain flow data, and social interaction data between enterprises from multiple data sources, clean and standardize the collected data, and generate standardized multi-dimensional interaction data; The standardized multi-dimensional interaction data is input into the multi-dimensional trust evaluation model, and the trust of inter-enterprise cooperation is dynamically evaluated through the multi-dimensional trust evaluation model to generate multi-dimensional trust evaluation results; the external environmental variable data is integrated, and the historical and future trends of the external environmental variables are captured through the high-dimensional environmental response model, and the multi-dimensional trust evaluation results are integrated with the high-dimensional environmental response model to generate trust-environment optimization data; The trust-environment optimization data is input into the high-dimensional scenario prediction model, and the trust-environment optimization data is processed at multiple levels through the adaptive multi-scale analysis and hierarchical fractal structure analysis of the high-dimensional scenario prediction model to generate joint credit risk prediction data; The joint credit risk prediction data is input into the smart contract system, which triggers credit rating and risk control measures. The execution process of the smart contract system is dynamically adjusted through a hierarchical feedback control system to generate optimized credit ratings and final credit warning results.

2. The credit warning method for inter-enterprise cooperation according to claim 1 is characterized in that: The multi-dimensional trust evaluation model calculates the trust level of inter-enterprise cooperation through the following formula: Among them, S represents the trust degree of cooperation between enterprises, and its value is used to quantify the strength of trust relationship between enterprises; w i Represents the weight factor associated with different dimensional data, weight factor w i The size of determines the influence weight of each dimension data in the overall trust evaluation; i (D i ) represents the relationship between each dimension data D i The corresponding functional relationship, function D i Used to convert dimension data D i Converted into a numerical form suitable for trust evaluation; D i It represents the data of each dimension in the standardized multi-dimensional interaction data, including the standardized results of financial transaction data, contract fulfillment data, supply chain flow data and social interaction data between enterprises; n represents the number of dimensions involved in the trust calculation, representing the total number of different data dimensions involved in the analysis in the model.

3. The credit warning method according to claim 2, characterized in that: The weight factor w in the multidimensional trust evaluation model i It is obtained by training the machine learning algorithm on the historical cooperation data, and the weight factor w is dynamically adjusted according to the new cooperation data. i , the process includes: For each historical cooperation dataset, the weight factor w is trained using a supervised learning algorithm i , so that the trust predicted by the model is as close as possible to the actual trust; As new cooperation data is continuously introduced, the weight factor w is adjusted using an online learning algorithm. i Carry out real-time updates; In the weight factor w i During the dynamic adjustment process, the system also considers the potential impact of external factors such as market fluctuations and industry changes on the trust relationship.

4. The credit warning method according to claim 1, characterized in that: The high-dimensional environmental response model generates environmental response field data using a multi-factor random field modeling method, wherein the multi-factor random field modeling method is represented by the following formula: R=∫ Ω h(x)×φ(x)dx Among them, R represents environmental response, which means the comprehensive impact of different environmental variables on enterprise trust in the entire market environment; Ω represents the domain of environmental variables, which is the set of all potential variables in the market environment; h(x) represents the impact factor function of each environmental variable, which quantifies the specific impact of each environmental variable on enterprise trust; φ(x) represents the probability density function of environmental variables, which describes the distribution and occurrence probability of environmental variables in the market environment.

5. The credit warning method according to claim 4, characterized in that: The high-dimensional environmental response model further processes the environmental response field data through a distribution nesting model to capture the historical and future trends of environmental variables, specifically including: Conduct time series analysis on the historical data of each environmental variable to determine its past change patterns and generate historical trend data; Using the distribution nesting model, historical trend data is projected into the future, the changing trend of future environmental variables is predicted, and future trend data is generated; The predicted future trends are combined with historical trends to generate comprehensive environmental response field data that include historical and future trends. The comprehensive environmental response field data are then fused with multidimensional trust assessment results to generate trust-environment optimization data as input data for the high-dimensional scenario prediction model.

6. The credit warning method according to claim 1, characterized in that: The high-dimensional context prediction model analyzes the trust-environment optimization data through adaptive multi-scale analysis, wherein the multi-scale analysis is performed according to the following formula: Among them, P(t) represents the scenario prediction result at time t, which represents the credit risk prediction value of the enterprise at the future time point t; g j (t) represents the multi-scale analysis function related to time t, and the function g j (t) is used to describe the changes in situational characteristics at different time scales; j (λ j ) represents the decomposition function at the jth scale, and the decomposition function Ψ j (λ j ) is used to decompose situational features into influencing factors of different scales; j represents the parameter related to the jth scale, parameter λ j The decomposition weights of the situational features at this scale are adjusted to ensure that the model can flexibly respond to analysis needs at different scales; m represents the number of analysis scales, representing the number of multi-scale analysis levels applied in the model.

7. The credit warning method according to claim 6, characterized in that: The high-dimensional scenario prediction model further optimizes the scenario prediction results through hierarchical fractal structure analysis to capture the complex nonlinear relationships in the data, including: The scenario prediction results generated by multi-scale analysis are decomposed hierarchically to form fractal structures of different complexity; By calculating the fractal dimension, the strength of nonlinear relationships in each hierarchy is quantified to identify underlying complex patterns in the data; The nonlinear relationships at various levels are integrated to generate fractal-enhanced situational response field data, which include the results of multi-scale analysis and fractal structure analysis.

8. The credit warning method according to claim 1, characterized in that: The smart contract system dynamically adjusts the execution process of the smart contract system through a hierarchical feedback control system to generate an optimized credit rating and a final credit warning result. The hierarchical feedback control system includes: Data collection module, used to collect credit rating data and risk control measures execution data in real time during the execution of smart contracts; The data analysis module identifies deviations and potential problems in the execution process by comparing the collected execution data with the expected results; The dynamic adjustment module adjusts the execution strategy of the smart contract in real time according to the results of the data analysis module to ensure the optimal effect of credit rating and risk control measures.

9. The credit warning method according to claim 8, characterized in that: The feedback control in the smart contract system performs path-to-path interactive correction of data through a multi-path response mechanism, and performs cluster analysis on the corrected data through an iterative clustering algorithm, specifically including: In the process of interaction correction between pathways, the response data generated by different pathways are compared to identify and correct the inconsistencies between the data, ensuring that all pathway data are consistent with each other in multiple dimensions; Use iterative clustering algorithms to perform cluster analysis on the corrected data to identify the most representative credit risk patterns; The cluster analysis results are output as the final credit warning data.

10. A system for implementing the credit warning method for inter-enterprise cooperation as claimed in any one of claims 1 to 9, characterized in that: include: The data collection and processing module is used to collect financial transaction data, contract performance data, supply chain flow data and social interaction data between enterprises from multiple data sources, and clean and standardize the collected data to generate standardized multi-dimensional interaction data; A multi-dimensional trust evaluation module is used to input the standardized multi-dimensional interaction data into the multi-dimensional trust evaluation model, dynamically evaluate the trust of cooperation between enterprises through the multi-dimensional trust evaluation model, and generate a multi-dimensional trust evaluation result; The environmental response fusion module is used to integrate external environmental variable data and capture the historical and future trends of external environmental variables through a high-dimensional environmental response model. It integrates the multi-dimensional trust assessment results with the high-dimensional environmental response model to generate trust-environment optimization data. The scenario prediction module is used to input the trust-environment optimization data into the high-dimensional scenario prediction model, perform multi-level processing on the trust-environment optimization data through the adaptive multi-scale analysis and hierarchical fractal structure analysis of the high-dimensional scenario prediction model, and generate joint credit risk prediction data; The smart contract execution module is used to input the joint credit risk prediction data into the smart contract system, trigger credit rating and risk control measures through the smart contract system, and dynamically adjust the execution process of the smart contract system through a hierarchical feedback control system to generate optimized credit ratings and final credit warning results.

Images