Fuzzy comprehensive evaluation-based enterprise risk evaluation method and system, and storage medium

Abstract

The invention discloses an enterprise risk evaluation method and system based on fuzzy comprehensive evaluation and a storage medium, and relates to the technical field of enterprise risk evaluation. At present, it is difficult to accurately evaluate risks in combination with actual influence factors of enterprise risks. According to the invention, material purchase, project progress, project quality and contract performance risks are dynamically monitored and evaluated by obtaining enterprise operation data, including monitoring material damage, specification difference and timeliness, and evaluating material purchase risks; evaluating project progress risks in combination with the task dependency relationship and external factors; based on a dynamic weight mechanism, analyzing the influence of design defects, construction errors and raw material price fluctuation on project quality; and dynamically adjusting the risk prediction contract performance capability by using a fuzzy membership function and a time sequence model. According to the technical scheme, the evaluation real-time performance is improved by a dynamic adjustment mechanism; various complex factors are comprehensively considered, and diversified requirements of enterprises are met; accurate risk prediction is realized, and powerful support is provided for decision making.

Description

Technical Field

[0001] The present invention relates to the technical field of enterprise risk assessment, and in particular to an enterprise risk assessment method, system and storage medium based on fuzzy comprehensive evaluation. Background Art

[0002] Whether in project management, financial management, supply chain management, market expansion, or customer relationship maintenance, enterprises must identify, assess and respond to various potential risks.

[0003] At present, when enterprises conduct risk monitoring and assessment, they usually conduct risk assessment based on static analysis and regular inspection. Although this method can provide risk identification and prevention to a certain extent, it often cannot meet the needs of enterprises in the face of rapidly changing environments due to the following reasons, including: Limitations and uncertainty: Traditional methods mostly rely on deterministic data and static models, and are difficult to handle complex, uncertain, and ambiguous information; Lack of real-time performance: In the enterprise operation environment, especially in dynamic project management and market environment fluctuations, traditional risk management methods may be difficult to respond and adjust in time; Therefore, for enterprises, the risks they face are not only diverse but also highly uncertain. Therefore, how to achieve dynamic monitoring and analysis of enterprise risks after determining the actual factors affecting enterprise risks, and further achieve accurate evaluation of enterprise risks is a problem that needs to be solved at present. Summary of the invention

[0004] In view of the above-mentioned shortcomings of the prior art, the present invention provides an enterprise risk assessment method, system and storage medium based on fuzzy comprehensive evaluation, which can effectively solve the problem in the prior art that it is difficult to accurately assess the risk in combination with the actual influencing factors of the enterprise risk due to the dynamic state of the enterprise risk.

[0005] To achieve the above objectives, the present invention is implemented through the following technical solutions: The present invention provides an enterprise risk assessment method based on fuzzy comprehensive evaluation, comprising the following steps: Obtain the status information data of the enterprise's operation, determine the enterprise's material procurement risk, project progress risk, project quality risk and contract performance risk, and evaluate the enterprise's current operating risk value. Obtain materials from suppliers and monitor the damage area, degree of non-standardization, and timeliness of the materials to determine the procurement risks of the materials; Introduce external factors that affect the project progress, determine the task dependencies in the project, combine the external factors with the task dependencies and determine the project progress risk based on the fuzzy evaluation system; Among them, the external influencing factors include weather and market changes; Introducing the time dimension and environmental changes and clarifying the impact of market changes on project quality risks based on the fuzzy evaluation system, among which, based on the dynamic weight adjustment mechanism, the weights of sub-indicators are dynamically adjusted according to time and environmental changes. The sub-indicators include: design flaws, construction error rates, and raw material price fluctuations; When determining the contract performance risk, the indicators that affect the contract performance risk are judged and the fuzzy membership function is determined. The fuzzy membership function is dynamically adjusted based on the time series model, and the enterprise's future performance risk is predicted to determine the contract performance risk.

[0006] Compared with traditional static analysis methods, this technical solution can respond to changes in the enterprise risk environment in real time, and provide more accurate and timely risk assessment through dynamic adjustment mechanisms (such as sub-indicator weight adjustment, fuzzy membership function adjustment). Considering a variety of complex and uncertain factors, such as the damage of material procurement, external influences on project progress (weather, market changes) and environmental factors of project quality, the enterprise's risks are comprehensively assessed. Through fuzzy comprehensive evaluation, it is possible to effectively deal with the complexity, ambiguity and uncertainty in enterprise risk management, especially in a dynamically changing operating environment, where traditional methods are difficult to solve these problems. Combining time series models and fuzzy membership functions, the technical solution can predict future performance risks, provide early warnings for enterprises, and help enterprises take effective countermeasures. Through accurate assessment of various risks of enterprises, this technical solution can provide more powerful data support for enterprise management and help enterprises make more scientific decisions.

[0007] In summary, the advantages of this technical solution are: first, dynamically adjust risk assessment, improve real-time performance, and adapt to the rapidly changing operating environment; second, comprehensively consider a variety of complex factors and assess risks; third, deal with uncertainty and fuzzy information through fuzzy evaluation to improve the accuracy of risk prediction; fourth, based on time series models and fuzzy membership functions, accurate prediction of performance risks can be achieved, early warning can be given, and effective support can be provided for corporate decision-making.

[0008] Furthermore, the method for determining material procurement risks is as follows: Set a monitoring period to monitor the materials provided by the supplier, set the daily sampling volume and conduct sampling inspections during the monitoring period, and calculate the damaged area, non-standard degree, and non-time validity of the sample materials, including: Define sampling inspection methods, including: Build a dataset based on the materials provided by suppliers , which contains Different materials, each with characteristics, including breakage rate , price fluctuations , Transportation time , transport and transfer nodes , Weather severity Determine the risk level of the materials, mark the corresponding types of materials as low-sampling materials or high-sampling materials according to the risk level, and assign corresponding sampling quantities to the low-sampling materials or high-sampling materials to obtain sample materials.

[0009] Furthermore, the damaged area is determined as follows: The deep learning model is used to identify the contours of damaged areas in the material images. The damaged area of ​​each material image is calculated based on the contours and the cumulative damaged area is obtained by summing up. The model architecture of the deep learning model is: The input image undergoes multi-layer convolution operations to extract edge features.

[0010] in, Indicates Layer Output feature maps, Indicates The number of input channels of the layer, Indicates The convolution kernel acts on the input feature map, Indicates The biased top of the layer, represents the activation function;

[0011] in, represents the probability map of the output, Indicates The convolution kernel parameters of the layer, Indicates The output feature map of the layer, represents the convolution operation, Indicates The biased top of the layer, Represents the activation function.

[0012] Furthermore, the degree of heterogeneity is obtained by defining the geometric impact value and appearance impact value of the materials; The time validity is obtained by obtaining the time value of the sampled material exceeding the preset time, and the time values ​​of multiple sampled materials are recorded as the time validity.

[0013] Furthermore, the method for determining the project schedule risk is: Assume that the external influencing factors are , including weather, market tasks The impact assessment value of is a fuzzy number, expressed as a membership function ; Modeling dependencies between tasks: If the project has tasks, each task and tasks The dependencies are determined through the task dependency matrix Represents, where the matrix Indicates the task Depends on the task , if the task Depends on the task ,but ,otherwise ; Conduct fuzzy comprehensive evaluation of risk assessment. No. The risk score of each task is combined with schedule deviation, resource issues, external factors and the impact of task dependencies, calculate the risk score for each task,

[0014] in, Indicates the task The risk score of Indicates The weight of the risk factors, Indicates The membership degree of each risk factor is Indicates the task Depends on the task , Indicates the task The risk score of Influencing tasks The number of risk factors for the risk score, Representation and Tasks The number of other tasks involved; Calculate overall project schedule risk for all projects ,

[0015] in, Represents the total number of tasks in the project.

[0016] Furthermore, the method for determining project quality risks is: A hierarchy of indicators for determining project quality risks, including: Target layer, project quality risk, Indicator layer, various risk factors , Decompose based on the indicator layer to obtain sub-indicator layers, including design defects , Construction error rate , Raw material price fluctuations , Dynamically adjust the fuzzy membership function according to time and data,

[0017] in, Indicates time point The indicator value of and denote the 10th and 90th percentiles, respectively. Indicates sub-indicators In time The fuzzy membership function value at the time; Introduce a dynamic weight adjustment mechanism to dynamically adjust the weight of each sub-indicator according to time and environmental changes.

[0018] in, Indicates time The weight of represents the smoothing factor, Represents the weight updated by real-time data; Calculate the risk value of each sub-indicator,

[0019] in, Indicates Tasks or stages at time The risk score of Indicates Risk The weight of each indicator; The weighted sum of the sub-indicator risk values ​​of each task or stage is used to obtain the comprehensive risk value. , Calculate project quality risk ,

[0020] in, Indicated in tasks or stages at time The weight of .

[0021] Furthermore, the method for determining the contract performance risk is: Identify several sub-indicators that affect contract performance risk , are construction error rate, raw material price fluctuation, and contract performance capability; Construct a fuzzy membership function for each sub-indicator and define the membership interval.

[0022] in, Indicates sub-indicators At the moment The membership function value of , , Respectively represent the minimum, optimal and maximum values ​​of the sub-indicators. Build time series models,

[0023] in, Indicates The sub-indicators are The time series value of each task, represents a constant term, represents the autoregressive order, represents the moving average order, represents the autoregressive coefficient, represents the moving average coefficient, Indicates at time The moment The observed values ​​of the indicators, represents the number of lags in the autoregressive order, Indicates at time The error term at time, represents the number of lags in the moving average order, represents the white noise term; Fitting time series models based on historical observations to predict the future Sub-index value at the moment ; Dynamically adjust the fuzzy membership function according to the predicted value; Calculate the risk value of a single sub-indicator,

[0024] in, Indicates Task No. The risk value of each sub-indicator is Indicates sub-indicators In the future The membership function value at time , Indicates in the future Time next Task No. Dynamic weights of sub-indicators; The risk value of each task is obtained by weighting the sub-indicator risk value ; Calculate the risk value of all tasks to obtain the contract performance risk at future times .

[0025] Furthermore, based on the corresponding weight coefficients of the material procurement risk, project progress risk, project quality risk and contract performance risk, the current operating risk value of the enterprise is determined by weighted summation, and the current risk status of the enterprise is determined based on the risk area to which the operating risk value belongs.

[0026] An enterprise risk assessment system is implemented according to the enterprise risk assessment method based on fuzzy comprehensive evaluation.

[0027] A computer-readable storage medium stores a computer program, which implements the steps of any one of the above methods when executed by a processor.

[0028] Compared with the known prior art, the technical solution provided by the present invention has the following beneficial effects: By monitoring the damaged area, quality specifications and arrival time of materials provided by suppliers, the material procurement risk is calculated; the project progress risk is judged in combination with factors such as weather and dependencies; and the time dimension and environmental changes are introduced to further clarify the project quality risk. Finally, the fuzzy membership function of the indicators in the contract performance risk is dynamically adjusted based on the time series model, and the contract performance risk is obtained again. Based on the material procurement risk, project progress risk, project quality risk and contract performance risk defined by the fuzzy evaluation system, the company's current operating risk is evaluated, and multi-dimensional and actual operating conditions are used to monitor the company's risks, so as to obtain a more accurate current operating risk of the company, which is beneficial for the company to implement corresponding risk management according to the corresponding operating risk. BRIEF DESCRIPTION OF THE DRAWINGS

[0029] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the prior art descriptions are briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention, and for ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0030] Figure 1 It is a schematic diagram of the overall method of the present invention. DETAILED DESCRIPTION

[0031] In order to make the purpose, technical solution and advantages of the embodiments of the present invention clearer, the technical solution in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0032] The present invention will be further described below in conjunction with the embodiments.

[0033] Example 1 (see Figure 1 ): The enterprise risk assessment method based on fuzzy comprehensive evaluation includes the following steps: Obtain the status information data of the enterprise's operations (various information data, including but not limited to project customization, implementation, process monitoring and closing processing, that is, obtaining various information data in the enterprise), judge the enterprise's material procurement risk, project progress risk, project quality risk and contract performance risk, thereby defining the enterprise's current operating risk value, forming a preliminary monitoring and identification of the enterprise's operating risks, so as to facilitate further risk management of the enterprise in the future.

[0034] Among them, the method for determining material procurement risks is as follows: Set a monitoring period (every day during the monitoring period) to monitor the materials provided by suppliers. In order to avoid resource loss and collection of useless data (considering that the quantity of materials provided by suppliers is large and cannot be inspected one by one), set the daily sampling quantity during the monitoring period and conduct sampling inspection to calculate the damaged area, degree of non-standardization, and non-time validity of the sample materials.

[0035] The method for determining the damaged area is as follows: Obtain a dataset of damaged material images with edge annotations (e.g., damaged material image data in various scenarios, including different lighting conditions (bright, dim), material properties (metal, cloth, plastic, etc.), and damage types (cracks, wear, notches, etc.)); The edge information in the material image is learned through the deep learning model, and the edge contour detection is performed. The algorithm architecture of the deep learning model is as follows: Define the input raw material image

[0036] in, , , Respectively represent the height, width and number of channels of the input image, represents the set of real numbers; The input image undergoes multi-layer convolution operations to extract edge features.

[0037] in, Indicates Layer Output feature maps, Indicates The number of input channels of the layer, Indicates The convolution kernel acts on the input feature map, Indicates The biased top of the layer, represents the activation function, where it should be noted that since the edge contour of the damaged area usually has strong texture changes or gradient changes, the feature extraction of the convolution layer can capture these features from the data, so as to distinguish the damaged area from the background area through feature extraction, even if the characteristics of the damaged area vary in different scenes; Contour detection probability generation, probability map output by convolutional network , represents the probability that each pixel belongs to the edge,

[0038] in, Indicates The convolution kernel parameters of the layer, Indicates The output feature map of the layer, represents the convolution operation, Indicates The biased top of the layer, Represents the activation function, which is used to map the output value to , so the probability map can be transformed into Binarization generates a binary image of the contour, which is used to detect the damaged contour in the material image by inputting the material image collected during the monitoring period, and thereby calculate the cumulative damaged area of ​​multiple sampling samples.

[0039] Among them, the method of sampling inspection is defined, including: Build a dataset based on the materials provided by suppliers , which contains Different materials, each with characteristics, including breakage rate (history, can collect 3 days, 5 days), price fluctuations (Take the maximum fluctuation value and obtain the data of the current material market), transportation time (the time it takes for materials to arrive at their destination from the place of shipment), transportation and transfer nodes (Number of material transfer centers passed through, material transfer centers passed during the transportation process. Generally speaking, if the materials cannot be directly delivered, they are usually transferred to multiple transfer centers. Therefore, the materials will frequently be loaded and unloaded at the transfer centers. During the loading and unloading process, it is usually inevitable that the materials will be bumped or hit, etc.) Weather severity (It is defined according to the rainfall, snowfall and temperature in the weather at the location where the materials are transported during the transportation process. According to the safe rainfall, snowfall and temperature required for the materials, the deviation difference is determined between the rainfall, snowfall and temperature in the weather at the location where the materials are transported and the corresponding safe rainfall, snowfall and temperature (such as the difference between the rainfall at the location and the safe rainfall). The weather severity is defined based on the product of multiple deviation differences. The weather is considered because some materials are affected by bad weather during the transportation process, such as materials); The corresponding characteristics (breakage rate , price fluctuations , Transportation time , transport and transfer nodes and weather severity ) are assigned corresponding weight coefficients respectively, and the risk level values ​​of different materials provided by the current supplier are obtained by weighted summation. The preset calibrated risk value is obtained, so that the material types corresponding to the risk level values ​​less than the calibrated risk value are marked as low-sampling materials (small number of samples), and the material types corresponding to the risk level values ​​higher than the calibrated risk value are marked as high-sampling materials (large number of samples). Set a data set The sampling quantity is used to assign corresponding sampling quantities to low-sampled materials and high-sampled materials respectively. For the actual sampling quantities of low-sampled materials and high-sampled materials, it is only necessary to increase the sampling quantity of high-sampled materials and reduce the sampling quantity of low-sampled materials when determining the sampling quantity. The actual sampling quantity is not limited in this embodiment. Therefore, when determining the procurement risk in the subsequent period, it can ensure that the sampled materials can reflect the actual situation, and avoid the uneven distribution of sampled materials, less sampled materials or the generation of useless sampled materials, which affects the accuracy of subsequent judgment on the procurement risk of materials.

[0040] Among them, the method for determining the degree of non-standardization of materials is as follows: Define the geometric influence value, obtain the geometric dimensions of each sample, including length, width, height, thickness, and volume), and subtract the geometric dimensions from the calibrated geometric dimensions to obtain the difference (the difference between the length, width, height, thickness, and volume of the sample and the corresponding calibrated length, width, height, thickness, and volume), and calculate the sum of the differences (for example, the difference between the length of the sample and the calibrated length, the sum between the width of the sample and the calibrated width, etc.) to obtain the geometric influence value; Define the appearance impact value, detect the appearance status of each sampled material, detect the length of the scratch, the depth of the crack, the number of bubbles, and the degree of deformation (the degree of deformation can be identified by the above-mentioned deep learning model to identify the coordinates of the actual edge points of the current material image , based on the coordinates of the standard edge points , according to the formula , Indicates the deviation between the actual edge or geometric point and the standard reference point. If the value is higher than the threshold, it is determined to be deformed, and the number of deformations of the sampled samples is counted), the length of the scratch, the depth of the crack, the number of bubbles, and the degree of deformation of each sampled sample are calculated, and the sum of multiple sampled samples is obtained; The corresponding weight coefficients are assigned according to the appearance influence value and the geometric influence value, and the degree of material non-standardization is obtained by taking the weighted sum of the two.

[0041] The method for determining heterochronic validity is: Obtain whether the time for the sampled materials to arrive at the destination does not exceed the preset time. If it exceeds the preset time, obtain the time value of each sampled material exceeding the preset time, obtain the time values ​​of multiple sampled materials and record them as different time validity; Therefore, corresponding weight coefficients are assigned based on the damaged area, degree of non-standardization and non-temporal validity (the weight coefficients are preset values ​​based on actual conditions and will not be repeated here), and the weighted sum of these three factors is used to obtain the company's fuzzy material procurement risk.

[0042] Different from the above, in this embodiment, the project progress risk, project quality risk and contract performance risk are all calculated in a fuzzy evaluation manner to achieve an accurate assessment of enterprise risks.

[0043] The method for determining project schedule risk is: Considering that weather and market will affect the progress of projects in the enterprise, it is necessary to determine the impact of external factors on the progress of projects. Assume that the external influencing factors are , such as weather, market conditions, etc. (see below for explanation of tasks) The impact assessment value of is a fuzzy number, expressed as a membership function , whose evaluation value is ambiguous, for example: Weather, which can be assessed as Therefore, the impact of external factors on progress can be expressed by the fuzzy membership function as follows: ; Considering that the tasks in a project (infrastructure construction, equipment installation and commissioning, etc.) are usually interdependent, that is, the delay of some tasks may affect the progress of other tasks, the dependency modeling between tasks is carried out: Assume that the project has tasks, each task and tasks The dependencies are determined through the task dependency matrix Represents, where the matrix Indicates the task Depends on the task , if the task Depends on the task ,but ,otherwise ; Conduct fuzzy comprehensive evaluation of risk assessment. Risk score for each task It is determined by both internal and external factors, therefore: Assume The risk score of each task is affected by internal factors such as schedule deviation (referring to the gap between the actual task progress and the planned progress, such as the task is not completed on time, there are delays, and the task is completed ahead of time, resulting in resource allocation problems or insufficient preparation for subsequent tasks. Excessive deviation may cause delays in the overall project progress and increase costs, and continuous schedule deviations will affect the dependencies of the project and increase risk propagation) and resource issues (referring to obstacles or deficiencies related to resources during project execution, such as manpower, key personnel cannot be on duty or personnel skills are insufficient; materials, key materials are not delivered on time, or quality problems make them unusable; funds, budget overruns, resulting in certain task resources cannot be paid on time, etc.). Insufficient resources may cause task delays and increase execution difficulty. It should be noted that schedule deviation reflects the accuracy of task completion time and helps identify risks that may cause delays in construction period. Resource issues reflect the degree of resource guarantee required for tasks and help identify risks caused by insufficient or improper resource allocation. In addition, it is also affected by external factors (such as weather, market changes, etc.) and the impact of task dependencies, that is, the risk score of each task is calculated through the fuzzy weighted sum formula.

[0044] in, Indicates the task The risk score of Indicates The weight of each risk factor (such as schedule deviation, resource issues, etc.), Indicates The membership degree of each risk factor is expressed by fuzzy numbers. Indicates the task Depends on the task , Indicates the task The risk score of Influencing tasks The number of risk factors for the risk score, including schedule deviation, resource issues, weather, market changes, Representation and Tasks The number of other tasks related to it, especially the dependencies between tasks, such as Depends on Task 2 and Task 3, then is 2; Therefore, the tasks in the above Risk score It is obtained by comprehensively evaluating various risk factors within the task (such as schedule deviation, resource issues, etc.) and external influencing factors (such as weather, market changes, etc.), and considering the dependencies between tasks, to achieve the actual overall score of a single task or project stage. Therefore, the overall project schedule risk of all projects can be obtained. ,

[0045] in, It represents the total number of tasks in the project. Then, by incorporating external influencing factors and task dependencies into the fuzzy evaluation system, the project schedule risk of the enterprise can be more comprehensively evaluated.

[0046] The method for determining project quality risks is: Decompose project quality risks to determine the indicator hierarchy, including: Target layer, project quality risk, Indicator layer, various risk factors , Decompose based on the defined indicator layer to obtain sub-indicator layers, including evaluation indicators, design defects , Construction error rate , Raw material price fluctuations , Dynamically adjust the fuzzy membership function according to time and data,

[0047] in, Indicates time point The indicator value of and Represents the 10th and 90th percentiles, respectively, and is dynamically adjusted based on historical data. Indicates sub-indicators In time The fuzzy membership function value at the time; Fluctuations in raw material prices By adjusting the boundary value or nonlinear shape of the membership function,

[0048] in, Indicates that the price of raw materials fluctuates at time The fuzzy membership value of Indicates that the price of raw materials fluctuates at time The minimum value of and They represent the optimal and maximum values ​​of raw material prices at the time respectively; Introduce a dynamic weight adjustment mechanism to dynamically adjust the weight of each sub-indicator according to time and environmental changes (such as market fluctuations, supply chain disruptions, natural disasters, etc., which will significantly affect the importance of certain risk indicators. For example, when the market fluctuates violently, the risk of raw material price fluctuations may require a higher weight).

[0049] in, Indicates time The weight reflects the importance of a certain indicator to the overall risk assessment at the current moment. It represents the contribution of the weight at the previous moment. By introducing the historical weight (the weight value at the previous moment), the smoothness of the weight adjustment is maintained to avoid excessive adjustment caused by short-term fluctuations in environmental changes. represents the smoothing factor, It indicates the weight updated by real-time data or expert judgment (according to current market fluctuations). It is worth noting that the dynamic weight reflects the change in the importance of the indicator in the time dimension. For example, the weight of construction error is higher in a certain stage, while the weight of customer satisfaction is higher in the delivery stage. Calculate the risk value of each sub-indicator,

[0050] in, Indicates A task or stage (such as the infrastructure construction tasks or stages mentioned above) at a certain time The risk score of Indicates Among the risks (such as material price fluctuations, construction error rate, etc.), The weight of each indicator; The sub-indicator risk values ​​of each task or stage are weighted and summed up.

[0051] in, Indicates tasks or stages at time The comprehensive risk value of Calculate project quality risk ,

[0052] in, Indicated in tasks or stages at time The weight of Therefore, by introducing the time dimension, environmental changes, and a dynamic fuzzy comprehensive evaluation framework, the impact of market changes on project quality can be reflected in real time. At the same time, the weights and affiliations are dynamically adjusted with the project stage, which is more in line with the actual scenario and improves the accuracy and practicality of project quality risk assessment.

[0053] The method for determining contract performance risk is: First, decompose the contract performance risk into several sub-indicators , respectively, Construction error rate, raw material price fluctuations, and contract performance capabilities; Construct a fuzzy membership function for each sub-indicator and define the membership interval.

[0054] in, Indicates sub-indicators At the moment The membership function value of , , Respectively represent the minimum, optimal and maximum values ​​of the sub-indicators. In this embodiment, considering that the contract performance risk may change dynamically over time, a time series model is introduced to dynamically adjust the membership function to predict future performance risks, including: For each sub-indicator, use its historical observations , Among them, assuming that the project updates progress every day, then They correspond to the first day and the second day of the project respectively, until the Tth day, then Indicates at time At this moment, Task No. The actual observed value of each sub-indicator; Therefore, to construct a time series model,

[0055] in, Indicates The sub-indicators are The time series value of each task, represents a constant term, represents the autoregressive order, represents the moving average order, represents the autoregressive coefficient, represents the moving average coefficient, Indicates at time The moment The observed values ​​of the indicators, represents the number of lags in the autoregressive order, Indicates at time The error term at time, represents the number of lags in the moving average order, represents the white noise term; Then, we fit the time series model based on historical data to predict the future The value of time is,

[0056] in, Indicates the predicted future The sub-indicator value at the moment, represents the autoregressive value predicted based on past moments, Represents the moving average residual value based on past predictions; Then, the fuzzy membership function can be adjusted dynamically according to the predicted value.

[0057] in, Indicates sub-indicators In the future The membership function value at time , , , Respectively represent the sub-indicators in the future The minimum, optimal and maximum values ​​at the moment; Then, the sub-indicator weights are dynamically adjusted over time.

[0058] Indicates in the future Time next Task No. Dynamic weights of sub-indicators; Calculate the risk value of a single sub-indicator,

[0059] in, Indicates Task No. Risk value of each sub-indicator; Therefore, the risk value of each task is obtained by weighted summing up the sub-indicator risk values. ; Calculate the risk value of all tasks to obtain the contract performance risk at future times ,

[0060] in, Indicates The combined weight of the tasks.

[0061] Furthermore, based on the material procurement risk, project progress risk, project quality risk and contract performance risk obtained above, the corresponding weight coefficients (the sum of multiple weight coefficients is usually 1) are respectively assigned, and the risk status of the current operation of the enterprise is judged by weighted summation, that is, an enterprise's operating risk value is obtained, and the preset risk range in the enterprise is obtained. According to the operating risk value in the risk area to which it belongs, the current risk status of the enterprise is judged. Then, if the risk status is high, corresponding risk control is carried out, thereby achieving accurate supervision of the enterprise's operating risks.

[0062] Finally, the present invention also provides: An enterprise risk assessment system is implemented according to the enterprise risk assessment method based on fuzzy comprehensive evaluation.

[0063] A computer-readable storage medium stores a computer program, which implements the steps of any one of the above methods when executed by a processor.

[0064] Just refer to the above method and I will not go into details here.

[0065] The above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements will not cause the essence of the corresponding technical solutions to deviate from the protection scope of the technical solutions of the embodiments of the present invention.

Claims

1. The enterprise risk assessment method based on fuzzy comprehensive evaluation includes the following steps: Obtain the status information data of the enterprise's operation, judge the enterprise's material procurement risk, project progress risk, project quality risk and contract performance risk, and evaluate the enterprise's current operating risk value. It is characterized by: Obtain materials from suppliers and monitor the damage area, degree of non-standardization, and timeliness of the materials to determine the procurement risks of the materials; Introduce external factors that affect the project progress, determine the task dependencies in the project, combine the external factors with the task dependencies and determine the project progress risk based on the fuzzy evaluation system; Among them, the external influencing factors include weather and market changes; The time dimension and environmental changes are introduced to clarify the impact of market changes on project quality risks based on the fuzzy evaluation system. The sub-indicator weights are adjusted based on the dynamic weight adjustment mechanism to determine the project quality risks. The sub-indicators include: design flaws, construction error rates, and raw material price fluctuations; When determining the contract performance risk, the indicators that affect the contract performance risk are judged and the fuzzy membership function is determined. The fuzzy membership function is dynamically adjusted based on the time series model, and the enterprise's future performance risk is predicted to determine the contract performance risk.

2. The enterprise risk assessment method based on fuzzy comprehensive evaluation according to claim 1 is characterized in that: The method for determining the material procurement risk is as follows: Set a monitoring period to monitor the materials provided by the supplier, set the daily sampling volume and conduct sampling inspections during the monitoring period, and calculate the damaged area, non-standard degree, and non-time validity of the sample materials, including: Define sampling inspection methods, including: Build a dataset based on the materials provided by suppliers , which contains Different materials, each with characteristics, including breakage rate , price fluctuations , Transportation time , transport and transfer nodes , Weather severity Determine the risk level of the materials, mark the corresponding types of materials as low-sampling materials or high-sampling materials according to the risk level, and assign corresponding sampling quantities to the low-sampling materials or high-sampling materials to obtain sample materials.

3. The enterprise risk assessment method based on fuzzy comprehensive evaluation according to claim 2 is characterized in that: The method for determining the damaged area is: The deep learning model is used to identify the contours of damaged areas in the material images. The damaged area of ​​each material image is calculated based on the contours and the cumulative damaged area is obtained by summing up. The model architecture of the deep learning model is: The input image undergoes multi-layer convolution operations to extract edge features. in, Indicates Layer Output feature map, Indicates The number of input channels of the layer, Indicates The convolution kernel acts on the input feature map, Indicates The biased top of the layer, represents the activation function; in, represents the probability map of the output, Indicates The convolution kernel parameters of the layer, Indicates The output feature map of the layer, represents the convolution operation, Indicates The biased top of the layer, Represents the activation function.

4. The enterprise risk assessment method based on fuzzy comprehensive evaluation according to claim 2 is characterized in that: The degree of non-standardization is obtained by defining the geometric impact value and the appearance impact value of the material; The time validity is obtained by obtaining the time value of the sampled material exceeding the preset time, and the time values ​​of multiple sampled materials are recorded as the time validity.

5. The enterprise risk assessment method based on fuzzy comprehensive evaluation according to claim 1 is characterized in that: The method for determining the project schedule risk is as follows: Assume that the external influencing factors are , including weather, market tasks The impact assessment value of is a fuzzy number, expressed as a membership function ; Modeling dependencies between tasks: If the project has tasks, each task and tasks The dependencies are determined through the task dependency matrix Represents, where the matrix Indicates the task Depends on the task , if the task Depends on the task ,but ,otherwise ; Conduct fuzzy comprehensive evaluation of risk assessment. No. The risk score of each task is combined with schedule deviation, resource issues, external factors and the impact of task dependencies, calculate the risk score for each task, in, Indicates the task The risk score of Indicates The weight of the risk factors, Indicates The membership degree of each risk factor is Indicates the task Depends on the task , Indicates the task The risk score of Influencing tasks The number of risk factors for the risk score, Representation and Tasks The number of other tasks involved; Calculate overall project schedule risk for all projects , in, Represents the total number of tasks in the project.

6. The enterprise risk assessment method based on fuzzy comprehensive evaluation according to claim 1 is characterized in that: The method for determining the project quality risk is as follows: A hierarchy of indicators for determining project quality risks, including: Target layer, project quality risk, Indicator layer, various risk factors , Decompose based on the indicator layer to obtain the sub-indicator layer that affects the project quality risk, including design defects , Construction error rate , Raw material price fluctuations , Dynamically adjust the fuzzy membership function according to time and data, in, Indicates time point The indicator value of and denote the 10th and 90th percentiles, respectively. Indicates sub-indicators In time The fuzzy membership function value at the moment; Introduce a dynamic weight adjustment mechanism to dynamically adjust the weight of each sub-indicator according to time and environmental changes. in, Indicates time The weight of represents the smoothing factor, Represents the contribution of the weight at the previous moment, Represents the weight updated by real-time data; Calculate the risk value of each sub-indicator, in, Indicates Tasks or stages at time The risk score of Indicates The risk category The weight of each indicator; The weighted sum of the sub-indicator risk values ​​of each task or stage is used to obtain the comprehensive risk value. , Calculate project quality risk , in, Indicated in Tasks or stages at time The weight of .

7. The enterprise risk assessment method based on fuzzy comprehensive evaluation according to claim 1 is characterized in that: The method for determining the contract performance risk is as follows: Identify several sub-indicators that affect contract performance risk , are construction error rate, raw material price fluctuation, and contract performance capability; Construct a fuzzy membership function for each sub-indicator and define the membership interval. in, Indicates sub-indicators At the moment The membership function value of , , Respectively represent the minimum, optimal and maximum values ​​of the sub-indicators. Build time series models, in, Indicates The sub-indicators are The time series value of each task, represents a constant term, represents the autoregressive order, represents the moving average order, represents the autoregressive coefficient, represents the moving average coefficient, Indicates at time The moment The observed values ​​of the indicators, represents the number of lags in the autoregressive order, Indicates at time The error term at time, represents the number of lags in the moving average order, represents the white noise term; Fitting time series models based on historical observations to predict the future Sub-index value at the moment ; Dynamically adjust the fuzzy membership function according to the predicted value; Calculate the risk value of a single sub-indicator, in, Indicates Task No. The risk value of each sub-indicator is Indicates sub-indicators In the future The membership function value at time , Indicates in the future Time next Task No. Dynamic weights of sub-indicators; The risk value of each task is obtained by weighting the sub-indicator risk value. ; Calculate the risk value of all tasks to obtain the contract performance risk at future times .

8. The enterprise risk assessment method based on fuzzy comprehensive evaluation according to claim 1 is characterized in that: According to the material procurement risk, project progress risk, project quality risk and contract performance risk, which belong to the corresponding weight coefficients respectively, the current operating risk value of the enterprise is determined by weighted summation. According to the risk area to which the operating risk value belongs, the current risk status of the enterprise is determined.

9. An enterprise risk assessment system, characterized in that: This is achieved according to the enterprise risk assessment method based on fuzzy comprehensive evaluation described in claim 1.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 8 are implemented.

Images