Regional enterprise migration early warning analysis method and system based on big data

Abstract

The invention relates to a regional enterprise migration early warning analysis method and system based on big data. The method comprises the following steps: constructing a multi-dimensional industry equilibrium domain based on a regional enterprise migration database; acquiring operation characteristics of the target enterprise in the current operation cycle; predicting the operating characteristics of the next operating cycle through the migration tendency prediction model in combination with the operating characteristics of the current operating cycle; generating a migration risk state vector based on the operation characteristics of the current operation cycle and the operation characteristics of the next operation cycle, and determining the migration development trend of the target enterprise; determining an ecological site of the target enterprise according to the migration risk state vector, and performing an equilibrium pressure test on the regional industry ecological system along the migration development trend to obtain an unbalance critical point of a multi-dimensional industry equilibrium domain; calculating the distance between the ecological site of the target enterprise and the imbalance critical point to obtain the industrial ecological matching degree of the target enterprise; and when the industrial ecological matching degree is smaller than an ecological matching threshold value, activating an enterprise retention intervention mechanism.

Description

Technical Field

[0001] The present invention relates to the field of enterprise migration early warning, and in particular to a regional enterprise migration early warning analysis method and system based on big data. Background Art

[0002] Enterprise relocation is a microcosm of regional industrial shifts, impacting both the originating and destinations differently. For the originating region, enterprise relocation can result in reduced tax revenue, shifted supply chains, and even severely impact local structural adjustments and industrial upgrades. For the destination, however, enterprise relocation can lead to increased tax revenue, enhanced vitality, and even the potential for new economic growth.

[0003] Therefore, regional enterprise migration has become a core phenomenon in the restructuring of the global industrial landscape. The large-scale migration of enterprises will lead to high volatility in regional tax revenue and a continued increase in the risk of industrial chain disruptions. However, the existing early warning mechanism has three major technical bottlenecks: 1. Dynamic perception lag: relying on quarterly economic reports (such as tax amounts and employment data), it is unable to capture real-time operational changes before migration (such as a sharp drop in capacity utilization and delayed supply chain response), and the early warning delay can be as long as 6-8 months. 2. Broken system linkages: Analyzing factors such as policies, supply chains, and talent independently ignores the transmission chain of "tax adjustment → investment reduction → talent loss". 3. Lack of precise intervention: adopting a "one-size-fits-all" retention policy, low-risk enterprises are over-subsidized, while high-risk enterprises are under-measured.

[0004] Therefore, it is urgent to fully consider the factors affecting enterprise migration, use multi-source big data to mine the data patterns of historically migrated enterprises, and make accurate predictions on enterprise migration. Summary of the Invention

[0005] To address the shortcomings of the existing technology, the present invention provides a method and system for regional enterprise migration early warning analysis based on big data. The method comprises the following steps:

[0006] Constructing a multi-dimensional industrial equilibrium domain based on the regional enterprise migration database;

[0007] Acquiring real-time operating data of a target enterprise, and extracting active operating characteristic data and passive operating characteristic data from the real-time operating data;

[0008] Extracting the temporal variation pattern of the active operation feature data as the active operation temporal feature, and extracting the temporal variation pattern of the passive operation feature data as the passive operation temporal feature;

[0009] Establishing a migration tendency prediction model based on the active operation time series characteristics and the passive operation time series characteristics, extracting active operation characteristics of the current operation cycle based on the active operation time series characteristics, and then extracting passive operation characteristics of the current operation cycle based on the passive operation time series characteristics;

[0010] Predicting the active and passive operating characteristics of the next operating cycle based on the migration tendency prediction model and the active and passive operating characteristics of the current operating cycle;

[0011] generating a migration risk state vector of the target enterprise based on the active and passive operating characteristics of the current operating cycle and the active and passive operating characteristics of the next operating cycle, and determining a migration development trend of the target enterprise based on the migration risk state vector of the target enterprise;

[0012] Obtain the ecological site of the target enterprise based on the migration risk state vector of the target enterprise, and conduct a balanced stress test on the regional industrial ecosystem along the migration development trend to obtain the imbalance critical point of the multi-dimensional industrial equilibrium domain;

[0013] Calculate the distance between the ecological site of the target enterprise and the critical point of imbalance in the multi-dimensional industrial equilibrium domain to obtain the industrial ecological matching degree of the target enterprise;

[0014] When the industrial ecological matching degree is less than the ecological matching threshold, the enterprise retention intervention mechanism is activated.

[0015] According to a preferred embodiment, constructing a multi-dimensional industry equilibrium domain based on a regional enterprise migration database includes:

[0016] Based on the regional enterprise migration database, all enterprises that have migrated are identified and marked as the benchmark migration sample set;

[0017] Randomly select a sample enterprise from the benchmark migration sample set as the target migration sample;

[0018] Obtaining full operating data of the target migration sample within a preset period before migration, and extracting active operating feature sets and passive operating feature sets from the full operating data;

[0019] Based on the active and passive management feature sets, several migration incentive dimensions of the target migration sample are determined, and a gradual equilibrium disturbance is applied to the regional industrial ecosystem along each migration incentive dimension until the regional industrial ecosystem is unbalanced, so as to obtain the pressure limit point of each migration incentive dimension;

[0020] Connecting the pressure limit points of all migration inducement dimensions to generate the equilibrium warning surface of the target migration sample;

[0021] Traverse all sample enterprises in the benchmark migration sample set, repeat the above steps, and obtain the equilibrium alert surface for each sample enterprise;

[0022] The equilibrium warning surfaces of all sample enterprises are fused through the Gaussian mixture model to generate a multi-dimensional industrial equilibrium domain.

[0023] According to a preferred embodiment, based on the active business feature set and the passive business feature set, determining several migration inducement dimensions of the target migration sample includes:

[0024] generating a strategic deployment vector based on the active business feature set, and generating an environmental response vector based on the passive business feature set;

[0025] Identifying several migration pressure sources of the target enterprise based on the strategic deployment vector and the environmental response vector, and obtaining an active impact characteristic vector and a passive conduction characteristic vector of each migration pressure source;

[0026] Determine a set of associated stressors for each migration stressor based on the active impact eigenvector and the passive conduction eigenvector of each migration stressor;

[0027] Connect each migration stress source with its associated stress sources to generate the migration stress conduction path of each migration stress source;

[0028] The gradient change direction of each migration stress conduction pathway is calculated as the migration inducement dimension of the migration stress source;

[0029] Aggregate the migration inducement dimensions of all migration pressure sources to form the migration inducement dimension set of the target enterprise.

[0030] According to a preferred embodiment, determining the associated stressor set of each migration stressor based on the active impact characteristic vector and the passive conduction characteristic vector of each migration stressor includes:

[0031] Calculating the similarity between the active impact feature vector of each migration stress source and the active impact feature vector of each other migration stress source to obtain a first correlation degree between each migration stress source and each other migration stress source;

[0032] calculating the similarity between the passive conduction characteristic vector of each migration stressor and the passive conduction characteristic vector of each other migration stressor, and obtaining a second correlation degree between each migration stressor and each other migration stressor;

[0033] Calculate the similarity between the active impact feature vector of each migration stressor and the passive conduction feature vector of each other migration stressor to obtain the third correlation between each migration stressor and each other migration stressor;

[0034] The similarity between the passive conduction characteristic vector of each migration stress source and the active impact characteristic vector of each other migration stress source is calculated to obtain a fourth correlation degree between each migration stress source and each other migration stress source.

[0035] According to a preferred embodiment, determining the associated stressor set of each migration stressor based on the active impact characteristic vector and the passive conduction characteristic vector of each migration stressor includes:

[0036] Traverse all migration stressors, and use the migration stressor being traversed as the target stressor, and the remaining migration stressors as candidate stressors of the target stressor;

[0037] Traversing all candidate stressors of the target stressor, and taking the candidate stressor being traversed as the target candidate stressor;

[0038] comparing a first correlation degree between the target stressor and the target candidate stressor with a first correlation threshold, a second correlation degree with a second correlation threshold, a third correlation degree with a third correlation threshold, and a fourth correlation degree with a fourth correlation threshold;

[0039] When the following conditions are met, the target candidate stressor is marked as an associated stressor of the target stressor,

[0040] The first degree of association is greater than a first association threshold, the second degree of association is greater than a second association threshold, the third degree of association is less than a third association threshold, and the fourth degree of association is less than a fourth association threshold;

[0041] Repeat the above steps to determine the set of associated stressors for each migration stressor.

[0042] According to a preferred implementation method, the active operating characteristic data is the operating data generated by the company's autonomous decision-making behavior, reflecting the company's strategic intentions and risk prediction capabilities, which include: R&D investment data, production capacity layout data, and cooperation network data; the passive operating characteristic data is the feedback data generated by the company's response to external shocks, reflecting the company's risk resistance capabilities, which include: policy response speed, supply chain resilience, capital chain health, and talent stability.

[0043] According to a preferred embodiment, the industrial ecological matching degree is the degree of matching between the industrial environment of the enterprise's current location and its optimal survival and development needs; the industrial ecological matching degree is used to characterize the risk of enterprise migration. The lower the industrial ecological matching degree, the higher the risk of enterprise migration, and the higher the industrial ecological matching degree, the lower the risk of enterprise migration.

[0044] According to a preferred embodiment, the migration inducement dimension is a dimension that triggers enterprise migration risks, including but not limited to talent stability dimension, supply chain dependence dimension, capital resilience dimension and policy sensitivity dimension.

[0045] According to a preferred embodiment, the migration pressure source is a specific risk entity or event that triggers enterprise migration risk, including but not limited to: improvement of environmental protection standards, regional talent loss, skyrocketing prices of core raw materials, and reduced customer orders.

[0046] The regional enterprise migration early warning and analysis system based on big data described in the present invention comprises: a balancing domain construction module, a migration risk prediction module, a matching analysis module, and an intelligent intervention execution module, wherein each module has a communication connection;

[0047] The equilibrium domain construction module constructs a multi-dimensional industry equilibrium domain based on the regional enterprise migration database;

[0048] The migration risk prediction module generates a migration risk state vector for the target enterprise based on the target enterprise's real-time operating data, determines the migration development trend of the target enterprise based on the migration risk state vector, and then obtains the ecological site of the target enterprise based on the migration risk state vector. It then performs a balanced stress test on the regional industrial ecosystem along the migration development trend to obtain the imbalance critical point of the multi-dimensional industrial equilibrium domain;

[0049] The matching analysis module calculates the distance between the ecological site of the target enterprise and the critical point of imbalance in the multi-dimensional industrial equilibrium domain to obtain the industrial ecological matching degree of the target enterprise;

[0050] The intelligent intervention execution module activates the enterprise retention intervention mechanism when the industrial ecological matching degree is less than the ecological matching threshold.

[0051] The present invention has the following beneficial effects:

[0052] 1. By directly capturing the company's real-time operating data, instead of traditional quarterly economic reports, risk signals can be identified several months before the company actually makes a relocation decision, solving the problem of serious lag in existing early warnings.

[0053] 2. The established analytical model captures the dynamic correlation and transmission paths between key factors such as policy adjustments, rising costs, and talent loss, overcoming the drawbacks of analyzing factors separately.

[0054] 3. Based on the degree of industrial ecosystem matching, the actual risk levels of different enterprises can be objectively distinguished. Intervention measures can be allocated accordingly, avoiding ineffective investment in low-risk enterprises and insufficient measures for high-risk enterprises.

[0055] 4. Transforming the traditional approach of “responding to enterprise relocation after the fact” into “predicting risks before the fact” has enhanced the stability of the regional industrial ecology. BRIEF DESCRIPTION OF THE DRAWINGS

[0056] Figure 1 A structural block diagram of a regional enterprise migration early warning and analysis system based on big data provided by an exemplary embodiment;

[0057] Figure 2 A flowchart of a regional enterprise migration early warning analysis method based on big data is provided as an exemplary embodiment. DETAILED DESCRIPTION

[0058] Exemplary embodiments will be described in detail herein, examples of which are illustrated in the accompanying drawings. In the following description, when referring to the drawings, like numbers in different figures represent like or similar elements unless otherwise indicated. The embodiments described in the following exemplary embodiments are not intended to represent all possible embodiments consistent with the present invention. Rather, they are merely examples of apparatus and methods consistent with certain aspects of the present invention, as detailed in the appended claims.

[0059] The terms used in this invention are for the purpose of describing specific embodiments only and are not intended to limit the invention. The singular forms "a," "the," and "the" used in this invention and the appended claims are also intended to include plural forms unless the context clearly indicates otherwise. It should also be understood that the term "and / or" as used herein refers to and includes any or all possible combinations of one or more of the associated listed items.

[0060] It should be understood that although the terms "first," "second," "third," etc. may be used in the present invention to describe various information, such information should not be limited to these terms. These terms are merely used to distinguish information of the same type from one another. For example, first information may also be referred to as second information, and similarly, second information may also be referred to as first information, without departing from the scope of the present invention. Depending on the context, the term "if" as used herein may be interpreted as "when," "when," or "in response to determining."

[0061] See also Figure 1 The regional enterprise migration early warning and analysis system based on big data of the present invention may include: a balancing domain construction module, a migration risk prediction module, a matching analysis module and an intelligent intervention execution module, and each module has a communication connection;

[0062] The equilibrium domain construction module constructs a multi-dimensional industry equilibrium domain based on the regional enterprise migration database;

[0063] The migration risk prediction module generates a migration risk state vector for the target enterprise based on the target enterprise's real-time operating data, determines the migration development trend of the target enterprise based on the migration risk state vector, and then obtains the ecological site of the target enterprise based on the migration risk state vector. It then performs a balanced stress test on the regional industrial ecosystem along the migration development trend to obtain the imbalance critical point of the multi-dimensional industrial equilibrium domain;

[0064] The matching analysis module calculates the distance between the ecological site of the target enterprise and the critical point of imbalance in the multi-dimensional industrial equilibrium domain to obtain the industrial ecological matching degree of the target enterprise;

[0065] The intelligent intervention execution module activates the enterprise retention intervention mechanism when the industrial ecological matching degree is less than the ecological matching threshold.

[0066] See also Figure 2 The present invention provides a regional enterprise migration early warning analysis method based on big data, which includes:

[0067] S1. Construct a multi-dimensional industrial equilibrium domain based on the regional enterprise migration database.

[0068] Optionally, the multidimensional industrial equilibrium domain is a regional industrial system stability boundary model constructed through historical migration enterprise data, which represents the tolerance limit of the industrial ecology under different pressure combinations.

[0069] Optionally, the regional enterprise migration database includes all enterprises in the region that have migrated and relevant business data of the migrated enterprises.

[0070] In a preferred embodiment, constructing a multi-dimensional industry balance domain based on a regional enterprise migration database includes:

[0071] Based on the regional enterprise migration database, all enterprises that have migrated are identified and marked as the benchmark migration sample set;

[0072] Randomly select a sample enterprise from the benchmark migration sample set as the target migration sample;

[0073] Obtaining full operating data of the target migration sample within a preset period before migration, and extracting active operating feature sets and passive operating feature sets from the full operating data;

[0074] Based on the active and passive management feature sets, several migration incentive dimensions of the target migration sample are determined, and a gradual equilibrium disturbance is applied to the regional industrial ecosystem along each migration incentive dimension until the regional industrial ecosystem is unbalanced, so as to obtain the pressure limit point of each migration incentive dimension;

[0075] Connecting the pressure limit points of all migration inducement dimensions to generate the equilibrium warning surface of the target migration sample;

[0076] Traverse all sample enterprises in the benchmark migration sample set, repeat the above steps, and obtain the equilibrium alert surface for each sample enterprise;

[0077] The equilibrium warning surfaces of all sample enterprises are fused through the Gaussian mixture model to generate a multi-dimensional industrial equilibrium domain.

[0078] Optionally, the full amount of operating data includes all operating data of several operating cycles (the specific value is set according to actual conditions) before the migration.

[0079] Preferably, the migration inducement dimension is the dimensional level that triggers enterprise migration risks, including but not limited to the talent stability dimension, supply chain dependence dimension, capital resilience dimension and policy sensitivity dimension. The same migration inducement dimension includes multiple migration pressure sources.

[0080] Optionally, progressive equilibrium disturbance refers to the precise detection of critical points of enterprise migration risks through micro-scale, multi-frequency external shock tests under the dynamic equilibrium state of the industrial system.

[0081] Optionally, the pressure limit point of the migration incentive dimension refers to the maximum limit of the impact of the corresponding migration incentive dimension that the enterprise can withstand without making a migration decision.

[0082] In a preferred embodiment, based on the active business feature set and the passive business feature set, determining several migration inducement dimensions of the target migration sample includes:

[0083] generating a strategic deployment vector based on the active business feature set, and generating an environmental response vector based on the passive business feature set;

[0084] Identifying several migration pressure sources of the target enterprise based on the strategic deployment vector and the environmental response vector, and obtaining an active impact characteristic vector and a passive conduction characteristic vector of each migration pressure source;

[0085] Determine a set of associated stressors for each migration stressor based on the active impact eigenvector and the passive conduction eigenvector of each migration stressor;

[0086] Connect each migration stress source with its associated stress sources to generate the migration stress conduction path of each migration stress source;

[0087] The gradient change direction of each migration stress conduction pathway is calculated as the migration inducement dimension of the migration stress source;

[0088] Aggregate the migration inducement dimensions of all migration pressure sources to form the migration inducement dimension set of the target enterprise.

[0089] Optionally, the strategic deployment vector is used to quantify a set of multidimensional indicators of the enterprise's autonomous strategic behavior, reflecting the enterprise's proactive layout adjustment capabilities; the environmental response vector is used to quantify a set of feedback indicators for the enterprise's response to external pressure, reflecting the enterprise's environmental adaptability.

[0090] Preferably, the migration pressure source is a specific risk entity or event that triggers the enterprise migration risk, including but not limited to: improvement of environmental protection standards, regional talent loss, skyrocketing prices of core raw materials, and reduced customer orders.

[0091] Alternatively, the migration pressure transmission path refers to the dynamic process in which internal and external risk factors faced by enterprises are transmitted step by step through multi-dimensional channels such as industrial chains, capital chains, and talent chains, ultimately triggering migration decisions. Its core lies in revealing the evolution mechanism of risks from micro-disturbance to macro-migration behavior.

[0092] Preferably, the active impact characteristic vector is used to represent the quantitative vector of the intensity and timing characteristics of the direct active impact exerted by the migration pressure source on the enterprise; the passive conduction characteristic vector is used to represent the quantitative vector of the conduction ability and diffusion path of the cascade reaction triggered by the migration pressure source in the industrial chain.

[0093] Alternatively, associated stressors refer to external migration risk factors that are transmitted through links such as industrial chains and geographical chains and have cascading impacts on enterprises, that is, associated stressors are secondary stressors transmitted through industrial chains.

[0094] In a preferred embodiment, determining the associated stressor set of each migration stressor based on the active impact characteristic vector and the passive conduction characteristic vector of each migration stressor includes:

[0095] Calculating the similarity between the active impact feature vector of each migration stress source and the active impact feature vector of each other migration stress source to obtain a first correlation degree between each migration stress source and each other migration stress source;

[0096] calculating the similarity between the passive conduction characteristic vector of each migration stressor and the passive conduction characteristic vector of each other migration stressor, and obtaining a second correlation degree between each migration stressor and each other migration stressor;

[0097] Calculate the similarity between the active impact feature vector of each migration stressor and the passive conduction feature vector of each other migration stressor to obtain the third correlation between each migration stressor and each other migration stressor;

[0098] The similarity between the passive conduction characteristic vector of each migration stress source and the active impact characteristic vector of each other migration stress source is calculated to obtain a fourth correlation degree between each migration stress source and each other migration stress source.

[0099] In a preferred embodiment, determining the associated stressor set of each migration stressor based on the active impact characteristic vector and the passive conduction characteristic vector of each migration stressor includes:

[0100] Traverse all migration stressors, and use the migration stressor being traversed as the target stressor, and the remaining migration stressors as candidate stressors of the target stressor;

[0101] Traversing all candidate stressors of the target stressor, and taking the candidate stressor being traversed as the target candidate stressor;

[0102] comparing a first correlation degree between the target stressor and the target candidate stressor with a first correlation threshold, a second correlation degree with a second correlation threshold, a third correlation degree with a third correlation threshold, and a fourth correlation degree with a fourth correlation threshold;

[0103] When the following conditions are met, the target candidate stressor is marked as an associated stressor of the target stressor,

[0104] The first degree of association is greater than a first association threshold, the second degree of association is greater than a second association threshold, the third degree of association is less than a third association threshold, and the fourth degree of association is less than a fourth association threshold;

[0105] Repeat the above steps to determine the set of associated stressors for each migration stressor.

[0106] Optionally, the first correlation threshold, the second correlation threshold, the third correlation threshold and the fourth correlation threshold are set according to actual conditions.

[0107] Optionally, the vector similarity calculation method includes cosine similarity, Euclidean distance and Manhattan distance.

[0108] S2. Acquire the real-time operating data of the target enterprise, and extract active operating feature data and passive operating feature data from the real-time operating data, extract the time series change pattern of the active operating feature data as the active operating time series feature, and extract the time series change pattern of the passive operating feature data as the passive operating time series feature.

[0109] Optionally, the active operating characteristic data is the operating data generated by the company's autonomous decision-making behavior, reflecting the company's strategic intentions and risk prediction capabilities, which include: R&D investment data, production capacity layout data, and cooperation network data; the passive operating characteristic data is the feedback data generated by the company's response to external shocks, reflecting the company's risk resistance, which include: policy response speed, supply chain resilience, capital chain health, and talent stability.

[0110] Optionally, active operating time series characteristics are used to characterize the dynamic trajectory of the evolution of the company's autonomous decision-making behavior over time, revealing the accumulation process of active migration intentions; passive operating time series characteristics are used to characterize the dynamic trajectory of the evolution of the company's feedback to external shocks over time, revealing the accumulation process of migration intentions caused by the influence of the external environment.

[0111] S3. Establish a migration tendency prediction model based on the active operating time series characteristics and the passive operating time series characteristics, and extract the active operating characteristics of the current operating cycle based on the active operating time series characteristics, and then extract the passive operating characteristics of the current operating cycle based on the passive operating time series characteristics; predict the active operating characteristics and passive operating characteristics of the next operating cycle based on the migration tendency prediction model and the active operating characteristics and passive operating characteristics of the current operating cycle.

[0112] Optionally, active operating characteristics are used to characterize the operating behaviors caused by the company's autonomous decision-making, reflecting its ability and intention to actively adjust its strategic layout; passive operating characteristics are used to characterize the company's feedback when responding to external pressure, revealing its ability to resist risks.

[0113] S4. Generate a migration risk state vector of the target enterprise based on the active and passive operating characteristics of the current operating cycle and the active and passive operating characteristics of the next operating cycle, and determine the migration development trend of the target enterprise based on the migration risk state vector of the target enterprise.

[0114] Alternatively, the migration development trend is the dynamic evolution trajectory of an enterprise's migration direction, migration acceleration and chain conduction effect under the pressure of a multi-dimensional industrial ecology. By quantitatively analyzing the evolution path of an enterprise from "steady-state anchoring" to "migration triggering", the spatiotemporal laws of the flow of industrial factors can be predicted.

[0115] S5. Obtain the ecological site of the target enterprise based on the migration risk state vector of the target enterprise, and conduct a balanced stress test on the regional industrial ecosystem along the migration development trend to obtain the imbalance critical point of the multi-dimensional industrial equilibrium domain.

[0116] Optionally, the ecological site is the resource coordinates of the enterprise in the industrial ecology, which is composed of the talent supply rate, industrial chain matching degree, policy adaptability, capital flow efficiency, etc.

[0117] Optionally, equilibrium stress testing is to simulate the extreme pressure of the regional industrial ecosystem by continuously applying increasing pressure along the trend of enterprise migration development until a systemic collapse is triggered, thereby accurately locating the critical point of imbalance.

[0118] Alternatively, the critical point of imbalance is the threshold boundary where the regional industrial ecosystem undergoes irreversible collapse. When the enterprise ecological site crosses this boundary, the system will lose its self-repair ability. The critical point of imbalance is the breaking point of the industrial ecological stability.

[0119] S6. Calculate the distance between the ecological site of the target enterprise and the critical point of imbalance in the multi-dimensional industrial equilibrium domain to obtain the industrial ecological matching degree of the target enterprise; when the industrial ecological matching degree is less than the ecological matching threshold, activate the enterprise retention intervention mechanism.

[0120] Alternatively, the distance between the ecological site and the critical point of imbalance can be the Mahalanobis distance.

[0121] Optionally, the industrial ecology matching degree is the degree of matching between the industrial environment of the enterprise's current location and its optimal survival and development needs; the industrial ecology matching degree is used to characterize the risk of enterprise migration. The lower the industrial ecology matching degree, the higher the risk of enterprise migration, and the higher the industrial ecology matching degree, the lower the risk of enterprise migration.

[0122] Optionally, the ecological matching threshold is set according to actual conditions.

[0123] This application directly captures the real-time business data of enterprises instead of traditional quarterly economic reports, and can identify risk signals several months before the actual relocation decision of the enterprise is made, thus solving the problem of serious lag in existing early warning. By establishing an analytical model to capture the dynamic correlation and transmission path between key factors such as policy adjustments, rising costs, and talent loss, the disadvantages of analyzing factors separately are overcome. Based on the matching degree of the industrial ecology, the actual risk levels of different enterprises can be objectively distinguished. Intervention measures are allocated accordingly to avoid ineffective investment in low-risk enterprises and insufficient measures for high-risk enterprises. The traditional "post-response to enterprise relocation" is transformed into "pre-predicting risks", which improves the stability of the regional industrial ecology.

[0124] The computer program instructions for performing the operation of the present invention can be assembly instructions, instruction set architecture (ISA) instructions, machine instructions, machine-dependent instructions, microcode, firmware instructions, state setting data, or source code or object code written in any combination of one or more programming languages, including object-oriented programming languages ​​such as Smalltalk, C++, etc., and procedural programming languages ​​such as "C" language or similar programming languages. The computer readable program instructions can be executed entirely on the user's computer, partially on the user's computer, as an independent software package, partially on the user's computer, partially on a remote computer, or completely on a remote computer or server. In the case of a remote computer, the remote computer can be connected to the user's computer through any type of network including a local area network (LAN) or a wide area network (WAN), or can be connected to an external computer (e.g., using an Internet service provider to connect via the Internet). In some embodiments, an electronic circuit, such as a programmable logic circuit, a field programmable gate array (FPGA), or a programmable logic array (PLA), is personalized by utilizing the state information of the computer readable program instructions, and the electronic circuit can execute the computer readable program instructions, thereby realizing various aspects of the present invention.

[0125] The present invention provides a non-transitory computer-readable storage medium, wherein computer instructions are stored in the non-transitory computer-readable storage medium. When the computer instructions are executed by a processor, the processor is caused to execute the above method.

[0126] Those skilled in the art will appreciate that all or part of the steps in the above method can be completed by a program to instruct related hardware (e.g., a processor, FPGA, ASIC, etc.), and the program can be stored in a readable storage medium, such as a read-only memory, a disk, or an optical disk. All or part of the steps in the above embodiment can also be implemented using one or more integrated circuits. Accordingly, each module in the above embodiment can be implemented in the form of hardware, for example, by implementing its corresponding functions through an integrated circuit, or in the form of a software functional module, for example, by executing a program / instruction stored in a memory by a processor to implement its corresponding function. The embodiments of the present invention are not limited to any particular form of combination of hardware and software.

[0127] In addition, the functional units in the various embodiments herein may be integrated into a single processing unit, each unit may exist physically separately, or two or more units may be integrated into a single unit. The aforementioned integrated units may be implemented in the form of hardware or software functional units.

[0128] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this article is essentially or the part that contributes to the existing technology, or all or part of the technical solution can be embodied in the form of a software product. The computer software product is stored in a storage medium and includes several instructions for causing a computer device (which can be a personal computer, server, or network device, etc.) to execute all or part of the steps of the method described in each embodiment of this article. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), disk or optical disk, etc. Various media that can store program code.

[0129] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. A regional enterprise migration early warning analysis method based on big data, characterized by: The following steps are involved: Constructing a multi-dimensional industrial equilibrium domain based on the regional enterprise migration database; Acquiring real-time operating data of a target enterprise, and extracting active operating characteristic data and passive operating characteristic data from the real-time operating data; Extracting the temporal variation pattern of the active operation feature data as the active operation temporal feature, and extracting the temporal variation pattern of the passive operation feature data as the passive operation temporal feature; Establishing a migration tendency prediction model based on the active operation time series characteristics and the passive operation time series characteristics, extracting active operation characteristics of the current operation cycle based on the active operation time series characteristics, and then extracting passive operation characteristics of the current operation cycle based on the passive operation time series characteristics; Predicting the active and passive operating characteristics of the next operating cycle based on the migration tendency prediction model and the active and passive operating characteristics of the current operating cycle; generating a migration risk state vector of the target enterprise based on the active and passive operating characteristics of the current operating cycle and the active and passive operating characteristics of the next operating cycle, and determining a migration development trend of the target enterprise based on the migration risk state vector of the target enterprise; Obtain the ecological site of the target enterprise based on the migration risk state vector of the target enterprise, and conduct a balanced stress test on the regional industrial ecosystem along the migration development trend to obtain the imbalance critical point of the multi-dimensional industrial equilibrium domain; Calculate the distance between the ecological site of the target enterprise and the critical point of imbalance in the multi-dimensional industrial equilibrium domain to obtain the industrial ecological matching degree of the target enterprise; When the industrial ecological matching degree is less than the ecological matching threshold, the enterprise retention intervention mechanism is activated.

2. The regional enterprise migration early warning analysis method according to claim 1 is characterized in that: Constructing a multi-dimensional industrial equilibrium domain based on the regional enterprise migration database includes: Based on the regional enterprise migration database, all enterprises that have migrated are identified and marked as the benchmark migration sample set; Randomly select a sample enterprise from the benchmark migration sample set as the target migration sample; Obtaining full operating data of the target migration sample within a preset period before migration, and extracting active operating feature sets and passive operating feature sets from the full operating data; Based on the active and passive management feature sets, several migration incentive dimensions of the target migration sample are determined, and a gradual equilibrium disturbance is applied to the regional industrial ecosystem along each migration incentive dimension until the regional industrial ecosystem is unbalanced, so as to obtain the pressure limit point of each migration incentive dimension; Connecting the pressure limit points of all migration inducement dimensions to generate the equilibrium warning surface of the target migration sample; Traverse all sample enterprises in the benchmark migration sample set, repeat the above steps, and obtain the equilibrium alert surface for each sample enterprise; The equilibrium warning surfaces of all sample enterprises are fused through the Gaussian mixture model to generate a multi-dimensional industrial equilibrium domain.

3. The regional enterprise migration early warning analysis method according to claim 2 is characterized in that: Based on the active and passive business feature sets, several migration inducement dimensions of the target migration sample are determined, including: generating a strategic deployment vector based on the active business feature set, and generating an environmental response vector based on the passive business feature set; Identifying several migration pressure sources of the target enterprise based on the strategic deployment vector and the environmental response vector, and obtaining an active impact characteristic vector and a passive conduction characteristic vector of each migration pressure source; Determine a set of associated stressors for each migration stressor based on the active impact eigenvector and the passive conduction eigenvector of each migration stressor; Connect each migration stress source with its associated stress sources to generate the migration stress conduction path of each migration stress source; The gradient change direction of each migration stress conduction pathway is calculated as the migration inducement dimension of the migration stress source; Aggregate the migration inducement dimensions of all migration pressure sources to form the migration inducement dimension set of the target enterprise.

4. The regional enterprise migration early warning analysis method according to claim 3 is characterized in that: Based on the active impact eigenvector and passive conduction eigenvector of each migration stressor, the associated stressor set of each migration stressor is determined to include: Calculating the similarity between the active impact feature vector of each migration stress source and the active impact feature vector of each other migration stress source to obtain a first correlation degree between each migration stress source and each other migration stress source; calculating the similarity between the passive conduction characteristic vector of each migration stressor and the passive conduction characteristic vector of each other migration stressor, and obtaining a second correlation degree between each migration stressor and each other migration stressor; Calculate the similarity between the active impact feature vector of each migration stressor and the passive conduction feature vector of each other migration stressor to obtain the third correlation between each migration stressor and each other migration stressor; The similarity between the passive conduction characteristic vector of each migration stress source and the active impact characteristic vector of each other migration stress source is calculated to obtain a fourth correlation degree between each migration stress source and each other migration stress source.

5. The regional enterprise migration early warning analysis method according to claim 4 is characterized in that: Based on the active impact eigenvector and passive conduction eigenvector of each migration stressor, the associated stressor set of each migration stressor is determined to include: Traverse all migration stressors, and use the migration stressor being traversed as the target stressor, and the remaining migration stressors as candidate stressors of the target stressor; Traversing all candidate stressors of the target stressor, and taking the candidate stressor being traversed as the target candidate stressor; comparing a first correlation degree between the target stressor and the target candidate stressor with a first correlation threshold, a second correlation degree with a second correlation threshold, a third correlation degree with a third correlation threshold, and a fourth correlation degree with a fourth correlation threshold; When the following conditions are met, the target candidate stressor is marked as an associated stressor of the target stressor, The first degree of association is greater than a first association threshold, the second degree of association is greater than a second association threshold, the third degree of association is less than a third association threshold, and the fourth degree of association is less than a fourth association threshold; Repeat the above steps to determine the set of associated stressors for each migration stressor.

6. The regional enterprise migration early warning analysis method according to claim 5, characterized in that: The active operating characteristic data is the operating data generated by the company's independent decision-making behavior, reflecting the company's strategic intentions and risk prediction capabilities, including: R&D investment data, production capacity layout data, and cooperation network data; the passive operating characteristic data is the feedback data generated by the company's response to external shocks, reflecting the company's risk resistance, including: policy response speed, supply chain resilience, capital chain health, and talent stability.

7. The regional enterprise migration early warning analysis method according to claim 7, characterized in that: The industrial ecological matching degree is the matching degree between the industrial environment of the enterprise's current location and its optimal survival and development needs; The industrial ecological matching degree is used to characterize the risk of enterprise migration. The lower the industrial ecological matching degree, the higher the risk of enterprise migration. The higher the industrial ecological matching degree, the lower the risk of enterprise migration.

8. The regional enterprise migration early warning analysis method according to claim 7, characterized in that: The migration incentive dimensions are dimensions that trigger enterprise migration risks, including but not limited to talent stability dimensions, supply chain dependence dimensions, financial resilience dimensions, and policy sensitivity dimensions.

9. The regional enterprise migration early warning analysis method according to claim 8, characterized in that: The migration pressure sources are specific risk entities or events that trigger enterprise migration risks, including but not limited to: improved environmental protection standards, regional talent loss, skyrocketing prices of core raw materials, and reduced customer orders.

10. A regional enterprise migration early warning analysis system based on big data, characterized by: It includes: a balance domain construction module, a migration risk prediction module, a matching analysis module and an intelligent intervention execution module, and each module has a communication connection; The equilibrium domain construction module constructs a multi-dimensional industry equilibrium domain based on the regional enterprise migration database; The migration risk prediction module generates a migration risk state vector for the target enterprise based on the target enterprise's real-time operating data, determines the migration development trend of the target enterprise based on the migration risk state vector, and then obtains the ecological site of the target enterprise based on the migration risk state vector. It then performs a balanced stress test on the regional industrial ecosystem along the migration development trend to obtain the imbalance critical point of the multi-dimensional industrial equilibrium domain; The matching analysis module calculates the distance between the ecological site of the target enterprise and the critical point of imbalance in the multi-dimensional industrial equilibrium domain to obtain the industrial ecological matching degree of the target enterprise; The intelligent intervention execution module activates the enterprise retention intervention mechanism when the industrial ecological matching degree is less than the ecological matching threshold.

Images