Bid invitation risk early warning method and system based on big data

Abstract

The invention discloses a bid invitation risk early warning method and system based on big data, relates to the technical field of bid invitation risk early warning, and solves the technical problems that in bid invitation risk early warning, the risk that a bidding enterprise fulfills a project is difficult to adaptively analyze, and weight factors with autonomous changes are difficult to set for different projects. According to the method, the anti-risk factors are obtained by analyzing the basic data; analyzing the target data to obtain a scheme factor; analyzing the historical bid-winning data to obtain capability factors; integrating the anti-risk factor, the scheme factor and the ability factor into a bid invitation risk factor representing the performance risk of the bidding enterprise; setting an early warning threshold 2 based on the bid invitation risk factors of the plurality of bidding enterprises, and obtaining a preset early warning threshold 1; performing risk early warning on the bidding enterprise based on the early warning threshold value I and the early warning threshold value II; according to the method, the complex enterprise qualification can be converted into a measurable risk index, and preposed management of the bid invitation risk is realized.

Description

Technical Field

[0001] The present invention belongs to the field of bidding management and relates to bidding risk early warning technology, specifically a bidding risk early warning method and system based on big data. Background Art

[0002] Bidding and procurement are crucial for modern businesses and government agencies to acquire goods, services, and projects. They not only impact project cost control but also directly impact quality, schedule, and compliance. In a complex market environment, the bidding process can be subject to various potential risks, such as unqualified bidders, bid rigging, and contract performance issues. To mitigate risks and improve the compliance and fairness of bidding and procurement activities, big data-based bidding risk warning systems and methods are gaining increasing attention.

[0003] At present, most bidding risk warning methods and systems based on big data find it difficult to adaptively analyze the risks of bidding companies in fulfilling projects during bidding risk warning, and cannot accurately provide advance warning of bidding risks, which increases the potential risks of bidding. At the same time, most bidding risk warning methods and systems based on big data find it difficult to set weight factors with autonomous changes for different projects. They only analyze based on experience or fixed weight factors, which cannot be applied to complex and changeable bidding projects, reducing the flexibility of the system.

[0004] Therefore, the present invention discloses a bidding risk early warning method and system based on big data, which are used to solve the above technical problems. Summary of the Invention

[0005] The present invention aims to solve at least one of the technical problems existing in the prior art; to this end, the present invention proposes a bidding risk warning method and system based on big data, which is used to solve the technical problems that it is difficult to adaptively analyze the risks of bidding companies in fulfilling projects, and it is difficult to set autonomously changing weight factors for different projects in bidding risk warning. The present invention solves the above problems by collecting the basic, target and historical winning bid data of bidding companies, analyzing and generating anti-risk factors, solution factors and capability factors, comprehensively evaluating the bidding risks of bidding companies and setting warning thresholds.

[0006] To achieve the above objectives, the first aspect of the present invention provides a bidding risk early warning method based on big data, comprising: Collect basic data, target data, and historical winning bid data of bidding companies; basic data includes qualification certificates, performance, and working capital; target data includes the text of bidding documents and progress control plans; Analyze the basic data to obtain the risk resistance factor that reflects the risk resistance of the bidding enterprise; Analyze the target data to obtain the solution factors that reflect the quality of the bidding enterprise's solution; Analyze historical bid-winning data to obtain capability factors that reflect the professional capabilities of bidding companies; The anti-risk factor, solution factor and capability factor are integrated into the bidding risk factor that reflects the performance risk of the bidding enterprise; Setting a second warning threshold based on the bidding risk factors of several bidding companies, and obtaining a pre-set first warning threshold; Risk warnings are issued to bidding companies based on warning threshold one and warning threshold two.

[0007] Preferably, the collection of basic data, target data and historical bid-winning data of bidding enterprises includes: After authorization by the bidding enterprise, the bidding enterprise's qualification certificate, performance and working capital, as well as the bid amount of the bidding enterprise's historical winning projects, the number of bidding enterprises and the performance of the bidding enterprise are obtained from the database; Through the bidding documents provided by the bidding companies, obtain the text and progress control plan in the bidding documents; among them, the historical winning bid data includes the winning amount of historical winning projects, the number of bidding companies and the performance of the bidding companies.

[0008] It should be noted that the performance in this invention refers to the performance of the full year before the bidding of the bidding project begins.

[0009] Preferably, the analysis of the basic data to obtain the risk resistance factor that reflects the risk resistance of the bidding enterprise includes: A1: Extract the bidding company's qualification certificate and determine whether it meets the qualification standards. If yes, jump to A2. If no, remove the current bidding company's bidding qualification. The qualification standards are the required qualification certificate standards set manually for the bidding project. A2: Extract the working capital LZ of the bidding enterprises, obtain the variance of several working capital LZs, and determine whether the variance is less than the variance threshold. If so, calculate the average value of the working capital LZs to obtain the standard working capital BLZ. If not, remove the working capital LZ with the largest difference from the mode of the working capital LZs from the several working capital LZs and re-determine the variance until the variance of the remaining working capital LZs is less than the variance threshold. Then, calculate the average value of the remaining working capital LZs to obtain the standard working capital BLZ. The variance threshold is obtained through experience. A3: Extract the performance Y of the bidding company i , based on the performance Y of the bidding company i Determine the current bidding company's risk resistance factor KZ i ; The anti-risk factor KZ i Satisfies the following formula: KZi =exp(α1×Y i / ∑(Y i )+α2×LZ / BLZ); where i is the bidding enterprise number, n is the number of bidding enterprises in this tender; ∑ is the summation function, and the summation range is [1,n]; α1 and α2 are both manually set proportional adjustment coefficients, and α1+α2=1, α1>α2.

[0010] Preferably, the analysis of the target data to obtain solution factors that reflect the quality of the bidding enterprise's solution includes: Extracting the text of the bidding document, the progress control plan, and the plan factors from the historical reference database; wherein the historical reference database includes the text of several bidding documents and the progress control plan, and the plan factors set by the experts based on the text of the bidding document and the progress control plan; Integrate the text of the bidding document, the schedule control plan, and the plan factors into several sets of training data and test data; use the training data to train the artificial intelligence model, use the test data to test the trained artificial intelligence model, and adjust the artificial intelligence model based on the test results; ultimately, obtain a document evaluation model whose input is the text of the bidding document and the schedule control plan, and whose output is the plan factors of the bidding document; wherein the artificial intelligence model includes a BP neural network model and an RBF neural network model; The text of the bidding document and the progress control plan of the bidding enterprise are input into the document evaluation model to obtain the plan factor of the bidding enterprise.

[0011] It should be noted that the scheme factor is an evaluation value obtained by experts through analysis based on the similarity between the text of the bid document and the text of other bid documents, and the ratio of the progress of each week in the progress control plan set by the bid document to the progress set by the tender; the higher the similarity between the text of the bid document and the text of the other bid documents, the greater the probability of collusion between the two bid documents, and the smaller the corresponding scheme factor; the greater the ratio of the progress of each week in the progress control plan set by the bid document to the progress set by the tender, the faster the bidding company can complete the project, and the larger the corresponding scheme factor.

[0012] Preferably, the analysis of historical bid-winning data to obtain capability factors that reflect the professional capabilities of bidding companies includes: Extract bidding companies one by one, integrate the winning amount of each winning project in the bidding company's historical winning data into a data set V, and obtain the bidding quality evaluation function of the bidding company based on formula (1): : (1); Wherein, j is the number of the winning project, and its value range is [1, m]; is the winning bid amount of the j-th winning project in the dataset V, is the maximum amount of winning bids in the dataset V, is the minimum value of the winning bid amount in the data set V; It is a manually set time attenuation coefficient, and its value range is (0,1]; The interval between the winning time of each winning project and the current time; Extract the performance of each bidding company in each winning project from the bidding company's historical winning data Based on formula (2), the competitive environment factor of the bidding enterprise is obtained : (2); Among them, r is the number of the bidding enterprise corresponding to the winning project numbered j; is the number of bidding companies that won the bid for project j, and BY is the standard performance set manually; Based on formula (3), the capability factor of the current bidding enterprise is obtained : (3); Among them, BQ is the manually set standard quality evaluation function, and BC is the manually set competitive environment factor; and They are all manually set proportional adjustment coefficients greater than 0, and + =1, .

[0013] It is worth noting that the present invention realizes a multi-dimensional and refined evaluation of the bidding enterprise's capabilities through dynamic time decay, amount standardization, competition environment quantification and indexation comprehensive evaluation; by balancing the fairness of project scale, focusing on the timeliness of recent performance and highlighting the differentiation of high-quality enterprises, it provides the tendering party with a scientific and objective decision-making basis, while encouraging enterprises to continuously improve the quality and competitiveness of bidding.

[0014] It should be noted that in formula (1), the numerator It is to realize the logarithmic transformation of the amount and eliminate the dimension difference; the denominator It is used to compress the results into the interval [0,1] to form a relative value index; An exponential decay model is adopted to reflect the evaluation principle that “recent projects are more valuable”.

[0015] Preferably, the anti-risk factor, solution factor and capability factor are integrated into a bidding risk factor that reflects the performance risk of the bidding enterprise, including: Extract the anti-risk factor KZ of the bidding enterprise i 、Scheme factors AZ iand capability factor NL i , obtain the anti-risk factor KZ i 、Scheme factors AZ i and capability factor NL i The corresponding adjustment coefficient is based on the anti-risk factor KZ i 、Scheme factors AZ i and capability factor NL i and the corresponding adjustment coefficient to determine the bidding risk factor ZF of the bidding enterprise i ; The bidding risk factor ZF i Satisfies the following formula: ZF i =exp(-μ×(θ1×KZ i +θ2×AZ i +θ3×NL i )); where μ is an manually set amplitude adjustment coefficient greater than 0, and the value range of μ is (0,2]; θ1 is the anti-risk factor KZ i The corresponding adjustment coefficient, θ2 is the solution factor AZ i The corresponding adjustment coefficient, θ3 is the capacity factor NL i The corresponding adjustment coefficient.

[0016] It should be noted that the performance risk of the bidding enterprise can be understood as the performance risk of the bidding enterprise on the bidding project after winning the bid.

[0017] Preferably, the anti-risk factor KZ is obtained i 、Scheme factors AZ i and capability factor NL i The corresponding adjustment coefficients include: Get the risk resistance factor KZ of the bidding companies in this tender i Determine the anti-risk factor KZ i The proportion factor ZB1 satisfies the following formula: ; Among them, the max() function is the maximum value function. Anti-risk factor KZ i The percentile of is the scheme factor AZ i The percentile of is the capability factor NL i The percentile of , percentile x is set manually; Get the proposal factors AZ of the bidding companies in this tender i Determine the solution factors AZ i The proportion factor ZB2 satisfies the following formula: ; Get the capability factor NL of the bidding enterprise in this tender i Determine the capability factor NL i The proportion factor ZB3 satisfies the following formula: ; Based on the proportion factors ZB1, ZB2, and ZB3, the adjustment coefficients θ1, θ2, and θ3 are determined; the adjustment coefficient θ1 satisfies the following formula: θ1=(ZB1 / BZ1) / (ZB1 / BZ1+ZB2 / BZ2+ZB3 / BZ3); where BZ1 is the artificially set anti-risk factor KZ i The standard proportion factor, BZ2 is the manually set scheme factor AZ i The standard proportion factor, BZ3 is the manually set capacity factor NL i The standard proportion factor of The adjustment coefficient θ2 satisfies the following formula: θ2=(ZB2 / BZ2) / (ZB1 / BZ1+ZB2 / BZ2+ZB3 / BZ3); The adjustment coefficient θ3 satisfies the following formula: θ3=(ZB3 / BZ3) / (ZB1 / BZ1+ZB2 / BZ2+ZB3 / BZ3).

[0018] Preferably, the second warning threshold is set based on the bidding risk factors of several bidding enterprises, including: Extract the first warning threshold from the database, obtain the number of bidding enterprises in this tender whose bidding risk factor is less than the first warning threshold, and determine whether the ratio LB of the number of bidding enterprises to the number of bidding enterprises in this tender exceeds the ratio threshold; if yes, determine the second warning threshold YZ2 based on the ratio LB, and the second warning threshold YZ2 satisfies the following formula: YZ2=YZ1 / (1+ln(LB+1)); where YZ1 is the warning threshold 1, and both the warning threshold 1 and the proportional threshold are manually set; No, extract the manually set warning threshold 2 from the database.

[0019] Preferably, the risk warning for the bidding enterprise based on the first warning threshold and the second warning threshold includes: B1: Extract bidding risk factors ZF of bidding enterprises in sequence i , determine the bidding risk factor ZF iWhether it exceeds the warning threshold 1; if yes, a warning message is issued to the tenderer indicating that the current bidding enterprise has bidding risks and is not recommended for selection; if no, jump to B2; B2: Determine the bidding risk factor ZF i Is it less than the warning threshold 2? If yes, the information that the current bidding enterprise is a high-quality enterprise will be sent to the tendering party; if no, the information that the current bidding enterprise is a medium-quality enterprise will be sent to the tendering party.

[0020] The second aspect of the present invention provides a bidding risk warning system based on big data, comprising: a risk analysis module, and an information collection module, a risk warning module and a database connected thereto; The information collection module is used to collect basic data, target data and historical bid-winning data of bidding enterprises; basic data includes qualification certificates, performance and working capital; target data includes the text of bidding documents and progress control plan; The risk analysis module is used to analyze basic data to obtain a risk resistance factor that reflects the bidder's risk resistance ability, analyze target data to obtain a solution factor that reflects the quality of the bidder's solution, analyze historical winning bid data to obtain a capability factor that reflects the bidder's professional ability, and combine the risk resistance factor, solution factor, and capability factor into a bidding risk factor that reflects the bidder's performance risk. The risk warning module is used to set a warning threshold 2 based on the bidding risk factors of several bidding companies, and obtain a pre-set warning threshold 1; and to issue risk warnings to the bidding companies based on the warning threshold 1 and the warning threshold 2.

[0021] Compared with the prior art, the present invention has the following beneficial effects: 1. The present invention obtains an anti-risk factor that reflects the anti-risk ability of the bidding enterprise by analyzing basic data; obtains a solution factor that reflects the quality of the bidding enterprise's solution by analyzing target data; obtains a capability factor that reflects the professional capability of the bidding enterprise by analyzing historical winning bid data; integrates the anti-risk factor, solution factor and capability factor into a bidding risk factor that reflects the performance risk of the bidding enterprise; sets a warning threshold 2 based on the bidding risk factors of several bidding enterprises, and obtains a pre-set warning threshold 1; and issues risk warnings to the bidding enterprises based on the warning threshold 1 and the warning threshold 2, thereby solving the technical problems of being difficult to adaptively analyze the risks of the bidding enterprises in performing projects and setting weight factors with autonomous changes for different projects in bidding risk warnings; the present invention can convert complex enterprise qualifications into measurable risk indicators, and realize pre-positioned management of bidding risks.

[0022] 2. This invention achieves a multi-dimensional and refined evaluation of bidding companies' capabilities through dynamic time decay, amount standardization, competitive environment quantification, and indexed comprehensive evaluation. By balancing the fairness of project scale, focusing on the timeliness of recent performance, and highlighting the differentiation of high-quality companies, it provides tenderers with a scientific and objective basis for decision-making, while also incentivizing companies to continuously improve the quality and competitiveness of their bids.

[0023] 3. When integrating the anti-risk factor, solution factor and capability factor into the bidding risk factor, the present invention does not design a fixed adjustment coefficient, but flexibly sets it according to the comprehensive situation of the anti-risk factor, solution factor and capability factor of the bidding enterprises in this bidding; compared with the fixed coefficient method, the present invention upgrades risk management from a static defense dominated by human experience to a dynamic immune system driven by machine intelligence by constructing a closed-loop model of "data perception → feedback optimization", forming a continuously optimized precipitation mechanism, so that the bidding risk assessment system has the ability to change autonomously, and realizes the accuracy and adaptability of risk assessment. BRIEF DESCRIPTION OF THE DRAWINGS

[0024] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0025] Figure 1 Schematic diagram of the operating steps of the present invention; Figure 2 Schematic diagram of the system module of the present invention; Figure 3 The figure is a schematic diagram of the operation steps for obtaining the anti-risk factor of a bidding enterprise according to the present invention. DETAILED DESCRIPTION

[0026] The technical solutions of the present invention will be clearly and completely described below in conjunction with the embodiments. Obviously, the embodiments described are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0027] See also Figure 1-Figure 2 The first embodiment of the present invention provides a bidding risk early warning method based on big data, comprising: Collect basic data, target data, and historical winning bid data of bidding companies; basic data includes qualification certificates, performance, and working capital; target data includes the text of bidding documents and progress control plans; Analyze the basic data to obtain the risk resistance factor that reflects the risk resistance of the bidding enterprise; Analyze the target data to obtain the solution factors that reflect the quality of the bidding enterprise's solution; Analyze historical bid-winning data to obtain capability factors that reflect the professional capabilities of bidding companies; The anti-risk factor, solution factor and capability factor are integrated into the bidding risk factor that reflects the performance risk of the bidding enterprise; Setting a second warning threshold based on the bidding risk factors of several bidding companies, and obtaining a pre-set first warning threshold; Risk warnings are issued to bidding companies based on warning threshold one and warning threshold two.

[0028] This application collects basic data, target data, and historical winning bid data of bidding companies, including: After authorization by the bidding enterprise, the bidding enterprise's qualification certificate, performance and working capital, as well as the bid amount of the bidding enterprise's historical winning projects, the number of bidding enterprises and the performance of the bidding enterprise are obtained from the database; Through the bidding documents provided by the bidding companies, obtain the text and progress control plan in the bidding documents; among them, the historical winning bid data includes the winning amount of historical winning projects, the number of bidding companies and the performance of the bidding companies.

[0029] It should be noted that the database can be understood as a storage system that includes data generated by the storage system, manually set data, and website public data.

[0030] It should be noted that the performance in this invention refers to the performance of the full year before the bidding of the bidding project begins.

[0031] See also Figure 3 In this application, the basic data is analyzed to obtain risk resistance factors that reflect the risk resistance of the bidding enterprise, including: A1: Extract the bidding company's qualification certificate and determine whether it meets the qualification standards. If yes, jump to A2; if no, remove the current bidding company's bidding qualification. The qualification standards are the required qualification certificate standards set manually for the bidding project. A2: Extract the working capital LZ of the bidding companies, obtain the variance of several working capital LZs, and determine whether the variance is less than the variance threshold. If so, calculate the average value of the working capital LZs to obtain the standard working capital BLZ. If not, remove the working capital LZ with the largest difference from the mode of the working capital LZs and re-determine the variance until the variance of the remaining working capital LZs is less than the variance threshold. Then, calculate the average value of the remaining working capital LZs to obtain the standard working capital BLZ. The variance threshold is obtained through experience. A3: Extract the performance Y of the bidding company i , based on the performance Y of the bidding company i Determine the current bidding company's risk resistance factor KZ i ; The anti-risk factor KZ i Satisfies the following formula: KZ i =exp(α1×Y i / ∑(Y i )+α2×LZ / BLZ); where i is the bidding enterprise number, n is the number of bidding enterprises in this tender; ∑ is the summation function, and the summation range is [1,n]; α1 and α2 are both manually set proportional adjustment coefficients, and α1+α2=1, α1>α2.

[0032] It should be noted that the present invention indirectly analyzes the risk resistance of bidding companies through their performance and working capital. The more performance and working capital a bidding company has, the stronger its risk resistance can be indirectly indicated, which is more beneficial to the implementation of subsequent bidding projects.

[0033] It should be noted that the working capital LZ of the bidding enterprises withdrawn for this tender is the working capital LZ of all bidding enterprises for this tender.

[0034] It should be noted that α1 and α2 are both manually set proportional adjustment coefficients, and α1>α2 because: α1 is multiplied by data related to the performance of the bidding enterprise, and α2 is multiplied by data related to the working capital of the bidding enterprise; for the risk resistance of the bidding enterprise, performance is more able to reflect the current economic size of the bidding enterprise and the ability to mobilize additional funds in an emergency than working capital. Therefore, the proportional adjustment coefficient α1 set in the present invention is greater than the proportional adjustment coefficient α2.

[0035] In this example, if the standard working capital BLZ = 10 million, the working capital LZ of bidding enterprise 1 = 15 million, and the performance Y1 of bidding enterprise 1 = 50 million, the total performance of the bidding enterprises in this tender is 500 million; Based on the formula KZ1=exp(α1×Y1 / ∑(Y i )+α2×LZ / BLZ)=exp(0.6×5000 / 50000+0.4×1500 / 1000)=1.93, and the risk resistance factor KZ1 of bidding enterprise 1 is obtained as 1.93; wherein, in this embodiment, the proportional adjustment coefficient α1 is 0.6, and the proportional adjustment coefficient α2 is 0.4.

[0036] In this application, the target data is analyzed to obtain proposal factors that reflect the quality of the bidding enterprise's proposal, including: Extracting the text of the bidding document, the progress control plan, and the plan factors from the historical reference database; wherein the historical reference database includes the text of several bidding documents and the progress control plan, and the plan factors set by the experts based on the text of the bidding document and the progress control plan; Integrate the text of the bidding document, the schedule control plan, and the plan factors into several sets of training data and test data; use the training data to train the artificial intelligence model, use the test data to test the trained artificial intelligence model, and adjust the artificial intelligence model based on the test results; ultimately, obtain a document evaluation model whose input is the text of the bidding document and the schedule control plan, and whose output is the plan factors of the bidding document; wherein the artificial intelligence model includes a BP neural network model and an RBF neural network model; The text of the bidding document and the progress control plan of the bidding enterprise are input into the document evaluation model to obtain the plan factor of the bidding enterprise.

[0037] It should be noted that the scheme factor is an evaluation value obtained by experts through analysis based on the similarity between the text of the bid document and the text of other bid documents, and the ratio of the progress of each week in the progress control plan set by the bid document to the progress set by the tender; the higher the similarity between the text of the bid document and the text of the other bid documents, the greater the probability of collusion between the two bid documents, and the smaller the corresponding scheme factor; the greater the ratio of the progress of each week in the progress control plan set by the bid document to the progress set by the tender, the faster the bidding company can complete the project, and the larger the corresponding scheme factor.

[0038] Specifically, the trained artificial intelligence model is tested using test data, and the specific steps for adjusting the artificial intelligence model based on the test results are as follows: The text of the bidding document and the progress control plan in the test data are input into the trained artificial intelligence model to obtain the corresponding plan factors, and the corresponding plan factors are compared with the corresponding plan factors in the test data. When the difference between the two is within the threshold, which is obtained based on experience, there is no need to adjust the parameters, and the next set of test data is tested; if it is not within the threshold, the corresponding parameters are adjusted until the difference between the two is within the threshold, and then the next set of test data is tested. When the number of test data whose output maximum safe deceleration is within the threshold for all test data accounts for 90% or more of the total test data, a document evaluation model is obtained with the text of the bidding document and the progress control plan as input and the plan factor of the bidding document as output.

[0039] In this application, historical bid-winning data is analyzed to obtain capability factors that reflect the professional capabilities of bidding companies, including: Extract bidding companies one by one, integrate the winning amount of each winning project in the bidding companies’ historical winning data into a data set V, and obtain the bidding quality evaluation function of the bidding companies based on formula (1): : (1); Wherein, j is the number of the winning project, and its value range is [1, m]; is the winning bid amount of the j-th winning project in the dataset V, is the maximum amount of winning bids in the dataset V, is the minimum value of the winning bid amount in the data set V; It is a manually set time attenuation coefficient, and its value range is (0,1]; The interval between the winning time of each winning project and the current time; Extract the performance of each bidding company in each winning project from the bidding company's historical winning data Based on formula (2), the competitive environment factor of the bidding enterprise is obtained : (2); Among them, r is the number of the bidding enterprise corresponding to the winning project numbered j; is the number of bidding companies that won the bid for project j, and BY is the standard performance set manually; Based on formula (3), the capability factor of the current bidding enterprise is obtained : (3); Among them, BQ is the manually set standard quality evaluation function, and BC is the manually set competitive environment factor; and They are all manually set proportional adjustment coefficients greater than 0, and + =1, .

[0040] It is worth noting that the present invention realizes a multi-dimensional and refined evaluation of the bidding enterprise's capabilities through dynamic time decay, amount standardization, competition environment quantification and indexation comprehensive evaluation; by balancing the fairness of project scale, focusing on the timeliness of recent performance and highlighting the differentiation of high-quality enterprises, it provides the tendering party with a scientific and objective decision-making basis, while encouraging enterprises to continuously improve the quality and competitiveness of bidding.

[0041] It is worth noting that formula (2) combines the competition density with the opponent's strength to achieve a two-dimensional evaluation of the competition intensity. If a bidding enterprise can win when there are many bidders and the opponent is strong, it can indicate that the bidding enterprise is strong and can increase the bidding enterprise's success rate in this bidding.

[0042] It is worth noting that in the present invention, in formula (1), The function assigns higher weight to recently awarded projects. Older projects contribute less to the capability assessment. This aligns with the industry's rapidly evolving technology landscape and volatile market environment. It ensures that assessment results are more closely aligned with a company's current capabilities, preventing them from relying on past successes to increase their chances of winning, and increasing the timeliness of analysis of bidding companies.

[0043] It is worth noting that in formula (1), the present invention converts the winning bid amount into The function is mapped to the interval [0,1] to eliminate the excessive impact of extremely large or small projects on the evaluation results; it enables small and medium-sized enterprises to improve their scores through multiple high-quality bids even if the bid amounts are low, avoiding the advantage of large enterprises relying on a single project monopoly.

[0044] It is worth noting that in the present invention, in formula (2), The function adjusts the intensity of competition. The more bidding companies there are, the more difficult it is to win, which reflects the value of the project. It indirectly indicates the quality of competition by comparing the performance of competitors with the standard performance. If the opponent is strong, the quality of the company's victory in the project will be higher. By distinguishing the difference between "easy to win the bid" and "winning in fierce competition", it reflects the true competitiveness of the company in different environments.

[0045] It is worth noting that in formula (3), the present invention gives priority to the quality of the bid while taking into account the competitive environment, which is in line with the evaluation logic of "quality first, environment second"; through linear combination, the quality of the bid and the competitive environment are transformed into exponential growth, so that the ability factors of high-scoring enterprises are significantly higher than the average level, which makes it easier for tenderers to quickly identify enterprises with outstanding comprehensive capabilities and improve decision-making efficiency.

[0046] It should be noted that in formula (1), the numerator It is to realize the logarithmic transformation of the amount and eliminate the dimension difference; the denominator It is used to compress the results into the interval [0,1] to form a relative value index; An exponential decay model is adopted to reflect the evaluation principle that “recent projects are more valuable”.

[0047] It should be noted that The time decay coefficient is set manually, and its value range is (0,1]. It is set according to the technical unfamiliarity time of each industry. The faster the technical unfamiliarity time of the industry, the higher the time decay coefficient. The larger the setting, the slower the industry's technology becomes unfamiliar, and the time decay coefficient The smaller the setting.

[0048] It should be noted that It is the interval between the winning time of each winning project and the current time, in months.

[0049] It should be noted that and All of them are manually set proportional adjustment coefficients greater than 0. Because: The product is the bidding quality evaluation function of the bidding enterprise relevant data, The factor multiplied by the competitive environment of the bidding enterprise When analyzing the capabilities of bidding companies, the quality of the projects won by the bidding companies more truly reflects the comprehensive strength of the companies in terms of technology, management, reputation, etc., compared with the competitive environment of the bidding at that time; therefore, the proportional adjustment coefficient designed by the present invention Greater than the proportional adjustment coefficient .

[0050] In this application, the risk resistance factor, solution factor and capability factor are integrated into the bidding risk factors that reflect the performance risk of the bidding enterprise, including: Extract the anti-risk factor KZ of the bidding enterprise i 、Scheme factors AZ i and capability factor NL i , obtain the anti-risk factor KZ i 、Scheme factors AZ i and capability factor NL i The corresponding adjustment coefficient is based on the anti-risk factor KZ i 、Scheme factors AZ i and capability factor NL i and the corresponding adjustment coefficient to determine the bidding risk factor ZF of the bidding enterprise i ; The bidding risk factor ZF i Satisfies the following formula: ZF i =exp(-μ×(θ1×KZ i +θ2×AZ i +θ3×NL i )); where μ is an manually set amplitude adjustment coefficient greater than 0, and the value range of μ is (0,2]; θ1 is the anti-risk factor KZ i The corresponding adjustment coefficient, θ2 is the solution factor AZ i The corresponding adjustment coefficient, θ3 is the capacity factor NL i The corresponding adjustment coefficient.

[0051] It should be noted that the performance risk of the bidding enterprise can be understood as the performance risk of the bidding enterprise on the bidding project after winning the bid.

[0052] It should be noted that μ is an artificially set amplitude adjustment coefficient greater than 0, and μ is used to adjust the anti-risk factor KZ of the bidding enterprise. i 、Scheme factors AZ i and capability factor NL i Tender risk factors for bidding enterprises i When other conditions remain unchanged, the larger μ is, the greater the bidding risk factor ZF of the bidding enterprise is. i The greater the impact, the smaller μ is. The bidding risk factor ZF of the bidding enterprise i The less impact.

[0053] It should be noted that the dependent variable of the exp(-μ) function decreases as the independent variable increases, and it is a downward trend curve.

[0054] In this application, the anti-risk factor KZ is obtained i 、Scheme factors AZ i and capability factor NL i The corresponding adjustment coefficients include: Get the risk resistance factor KZ of the bidding companies in this tender i Determine the anti-risk factor KZ i The proportion factor ZB1 satisfies the following formula: ; Among them, the max() function is the maximum value function. Anti-risk factor KZ i The percentile of is the scheme factor AZ i The percentile of is the capability factor NL i The percentile of , percentile x is set manually; Get the proposal factors AZ of the bidding companies in this tender i Determine the solution factors AZ i The proportion factor ZB2 satisfies the following formula: ; Get the capability factor NL of the bidding enterprise in this tender i Determine the capability factor NL i The proportion factor ZB3 satisfies the following formula: ; Based on the proportion factors ZB1, ZB2, and ZB3, the adjustment coefficients θ1, θ2, and θ3 are determined; the adjustment coefficient θ1 satisfies the following formula: θ1=(ZB1 / BZ1) / (ZB1 / BZ1+ZB2 / BZ2+ZB3 / BZ3); where BZ1 is the artificially set anti-risk factor KZ i The standard proportion factor, BZ2 is the manually set scheme factor AZ i The standard proportion factor, BZ3 is the manually set capacity factor NL i The standard proportion factor of The adjustment coefficient θ2 satisfies the following formula: θ2=(ZB2 / BZ2) / (ZB1 / BZ1+ZB2 / BZ2+ZB3 / BZ3); The adjustment coefficient θ3 satisfies the following formula: θ3=(ZB3 / BZ3) / (ZB1 / BZ1+ZB2 / BZ2+ZB3 / BZ3).

[0055] It is worth noting that there are significant differences in the risk structures of different projects in actual bidding. For example, technology-intensive projects need to increase the weight of solution factors, while capital-intensive projects need to focus on anti-risk factors. Therefore, when combining anti-risk factors, solution factors and capability factors into bidding risk factors, the present invention does not design a fixed adjustment coefficient, but flexibly sets it according to the comprehensive situation of the anti-risk factors, solution factors and capability factors of the bidding companies in this bidding. Compared with the fixed coefficient method, the present invention upgrades risk management from static defense dominated by human experience to a dynamic immune system driven by machine intelligence by constructing a closed-loop model of "data perception → feedback optimization", forming a continuously optimized precipitation mechanism, so that the bidding risk assessment system has the ability to change autonomously, and realizes the accuracy and adaptability of risk assessment.

[0056] It should be noted that Anti-risk factor KZ i The percentile of is the scheme factor AZ i The percentile of is the capability factor NL i The percentile x is set manually, and the percentile is the value at a specific percentile in the data; for example: the anti-risk factor KZ i Sort by large to small. If the manually set percentile is 60%, then the data at the 60% position is a number of anti-risk factors KZ i If there is no data at the 60% position, the data closest to the 60% position will be used as several anti-risk factors KZ i percentile of .

[0057] This application sets a second warning threshold based on the bidding risk factors of several bidding companies, including: Extract the first warning threshold from the database, obtain the number of bidding enterprises in this tender whose bidding risk factor is less than the first warning threshold, and determine whether the ratio LB of the number of bidding enterprises to the number of bidding enterprises in this tender exceeds the ratio threshold; if yes, determine the second warning threshold YZ2 based on the ratio LB, and the second warning threshold YZ2 satisfies the following formula: YZ2=YZ1 / (1+ln(LB+1)); where YZ1 is the warning threshold 1, and both the warning threshold 1 and the proportional threshold are manually set; No, extract the manually set warning threshold 2 from the database.

[0058] It should be noted that the ratio LB of the number of bidding companies to the total number of bidding companies can be explained by the formula: LB=SL / n; where SL is the number of bidding companies and n is the number of bidding companies in this tender.

[0059] It should be noted that the manually set warning threshold 2 is obtained based on the warning threshold 1 YZ1. For example, 0.75 of the warning threshold 1 YZ1 is taken as the value of the warning threshold 2.

[0060] In this application, risk warnings are issued to bidding companies based on warning thresholds 1 and 2, including: B1: Extract bidding risk factors ZF of bidding enterprises in sequence i , judge the bidding risk factor ZF i Whether it exceeds the warning threshold 1; if yes, a warning message is issued to the tenderer indicating that the current bidding enterprise has bidding risks and is not recommended for selection; if no, jump to B2; B2: Determine the bidding risk factors ZF i Is it less than the warning threshold 2? If yes, the information that the current bidding enterprise is a high-quality enterprise will be sent to the tendering party; if no, the information that the current bidding enterprise is a medium-quality enterprise will be sent to the tendering party.

[0061] The second embodiment of the present invention provides a bidding risk warning system based on big data, including: a risk analysis module, and an information collection module, a risk warning module and a database connected thereto; Information collection module: used to collect basic data, target data and historical bid-winning data of bidding companies; basic data includes qualification certificates, performance and working capital; target data includes the text of bidding documents and progress control plan; Risk Analysis Module: This module analyzes basic data to obtain risk resistance factors that reflect the bidder's risk resistance capabilities; analyzes target data to obtain solution factors that reflect the quality of the bidder's solution; and analyzes historical winning bid data to obtain capability factors that reflect the bidder's professional capabilities. The risk resistance factors, solution factors, and capability factors are combined into a bidding risk factor that reflects the bidder's performance risk. Risk warning module: used to set warning threshold 2 based on the bidding risk factors of several bidding companies, and obtain the pre-set warning threshold 1; and issue risk warnings to bidding companies based on warning threshold 1 and warning threshold 2.

[0062] Some of the data in the above formula are calculated by removing the dimensions and taking their numerical values. The formula is a formula that is closest to the actual situation obtained by software simulation of a large amount of collected data; the preset parameters and preset thresholds in the formula are set by technical personnel in this field according to actual conditions or obtained through simulation of a large amount of data.

[0063] Working principle of the present invention: Collect basic data, target data and historical bid-winning data of bidding companies; analyze basic data to obtain anti-risk factors that reflect the bidding companies' risk resistance, and analyze target data to obtain solution factors that reflect the quality of the bidding companies' solutions; analyze historical bid-winning data to obtain capability factors that reflect the bidding companies' professional capabilities; combine the anti-risk factors, solution factors and capability factors into a bidding risk factor that reflects the performance risk of the bidding companies; set a second warning threshold based on the bidding risk factors of several bidding companies, and obtain a pre-set first warning threshold; and issue risk warnings to bidding companies based on the first and second warning thresholds.

[0064] The above embodiments are only used to illustrate the technical method of the present invention and are not intended to limit the present invention. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical method of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical method of the present invention.

Claims

1. A bidding risk early warning method based on big data, characterized in that: include: Collect basic data, target data, and historical winning bid data of bidding companies; basic data includes qualification certificates, performance, and working capital; target data includes the text of bidding documents and progress control plans; Analyze the basic data to obtain the risk resistance factor that reflects the risk resistance of the bidding enterprise; Analyze the target data to obtain the solution factors that reflect the quality of the bidding enterprise's solution; Analyze historical bid-winning data to obtain capability factors that reflect the professional capabilities of bidding companies; The anti-risk factor, solution factor and capability factor are integrated into the bidding risk factor that reflects the performance risk of the bidding enterprise; Setting a second warning threshold based on the bidding risk factors of several bidding companies, and obtaining a pre-set first warning threshold; Risk warnings are issued to bidding companies based on warning threshold one and warning threshold two.

2. A bidding risk early warning method based on big data according to claim 1, characterized in that: The collection of basic data, target data and historical bid-winning data of bidding enterprises includes: After authorization by the bidding enterprise, the bidding enterprise's qualification certificate, performance and working capital, as well as the bid amount of the bidding enterprise's historical winning projects, the number of bidding enterprises and the performance of the bidding enterprise are obtained from the database; Through the bidding documents provided by the bidding companies, obtain the text and progress control plan in the bidding documents; among them, the historical winning bid data includes the winning amount of historical winning projects, the number of bidding companies and the performance of the bidding companies.

3. A bidding risk early warning method based on big data according to claim 2, characterized in that: The analysis of the basic data to obtain the risk resistance factors that reflect the risk resistance of the bidding enterprise includes: A1: Extract the bidding company's qualification certificate and determine whether it meets the qualification standards. If yes, jump to A2; if no, remove the current bidding company's bidding qualification. The qualification standards are the required qualification certificate standards set for the bidding project. A2: Extract the working capital LZ of the bidding enterprises, obtain the variance of several working capital LZs, and determine whether the variance is less than the variance threshold. If so, calculate the average value of the working capital LZs to obtain the standard working capital BLZ. If not, remove the working capital LZ with the largest difference from the mode of the working capital LZs from the several working capital LZs and re-determine the variance until the variance of the remaining working capital LZs is less than the variance threshold. Then, calculate the average value of the remaining working capital LZs to obtain the standard working capital BLZ. A3: Extract the performance Y of the bidding company i , based on the performance Y of the bidding company i Determine the current bidding company's risk resistance factor KZ i ; The anti-risk factor KZ i Satisfies the following formula: KZ i =exp(α1×Y i / ∑(Y i )+α2×LZ / BLZ); where i is the bidding enterprise number, n is the number of bidding enterprises in this tender; ∑ is the summation function, and the summation range is [1,n]; α1 and α2 are both proportional adjustment coefficients, and α1+α2=1, α1>α2.

4. The bidding risk early warning method based on big data according to claim 2 is characterized in that: The target data is analyzed to obtain solution factors that reflect the quality of the bidding enterprise's solution, including: Extracting the text of the bidding document, the progress control plan, and the plan factors from the historical reference database; wherein the historical reference database includes the text of several bidding documents and the progress control plan, and the plan factors set by the experts based on the text of the bidding document and the progress control plan; Integrate the text of the bidding document, the schedule control plan, and the plan factors into several sets of training data and test data; use the training data to train the artificial intelligence model, use the test data to test the trained artificial intelligence model, and adjust the artificial intelligence model based on the test results; ultimately, obtain a document evaluation model whose input is the text of the bidding document and the schedule control plan, and whose output is the plan factors of the bidding document; wherein the artificial intelligence model includes a BP neural network model and an RBF neural network model; The text of the bidding document and the progress control plan of the bidding enterprise are input into the document evaluation model to obtain the plan factor of the bidding enterprise.

5. The bidding risk early warning method based on big data according to claim 2 is characterized in that: The analysis of historical bid-winning data to obtain capability factors that reflect the professional capabilities of bidding companies includes: Extract bidding companies one by one, integrate the winning amount of each winning project in the bidding company's historical winning data into a data set V, and obtain the bidding quality evaluation function of the bidding company based on formula (1): : (1)； Wherein, j is the number of the winning project, and its value range is [1, m]; is the winning bid amount of the j-th winning project in the dataset V, is the maximum amount of winning bids in the dataset V, is the minimum value of the winning bid amount in the data set V; is the time attenuation coefficient, and its value range is (0,1]; The interval between the winning time of each winning project and the current time; Extract the performance of each bidding company in each winning project from the bidding company's historical winning data Based on formula (2), the competitive environment factor of the bidding enterprise is obtained : (2)； Among them, r is the number of the bidding enterprise corresponding to the winning project numbered j; is the number of bidding companies that won the bid for project j, BY is the standard performance; Based on formula (3), the capability factor of the current bidding enterprise is obtained : (3)； Among them, BQ is the standard quality evaluation function, BC is the competitive environment factor; and are all proportional adjustment coefficients greater than 0, and + =1, .

6. The bidding risk early warning method based on big data according to claim 1 is characterized in that: The aforementioned bidding risk factors that integrate risk resistance factors, solution factors, and capability factors to represent the performance risk of bidding enterprises include: Extract the anti-risk factor KZ of the bidding enterprise i 、Scheme factors AZ i and capability factor NL i , obtain the anti-risk factor KZ i 、Scheme factors AZ i and capability factor NL i The corresponding adjustment coefficient is based on the anti-risk factor KZ i 、Scheme factors AZ i and capability factor NL i and the corresponding adjustment coefficient to determine the bidding risk factor ZF of the bidding enterprise i ; The bidding risk factor ZF i Satisfies the following formula: ZF i =exp(-μ×(θ1×KZ i +θ2×AZ i +θ3×NL i )); where μ is the amplitude adjustment coefficient greater than 0, and the value range of μ is (0,2]; θ1 is the anti-risk factor KZ i The corresponding adjustment coefficient, θ2 is the solution factor AZ i The corresponding adjustment coefficient, θ3 is the capacity factor NL i The corresponding adjustment coefficient.

7. The bidding risk early warning method based on big data according to claim 6 is characterized in that: Acquiring the anti-risk factor KZ i 、Scheme factors AZ i and capability factor NL i The corresponding adjustment coefficients include: Get the risk resistance factor KZ of the bidding companies in this tender i Determine the anti-risk factor KZ i The proportion factor ZB1 satisfies the following formula: ； Among them, the max() function is the maximum value function. Anti-risk factor KZ i The percentile of is the scheme factor AZ i The percentile of is the capability factor NL i percentile of Get the proposal factors AZ of the bidding companies in this tender i Determine the solution factors AZ i The proportion factor ZB2 satisfies the following formula: ； Get the capability factor NL of the bidding enterprise in this tender i Determine the capability factor NL i The proportion factor ZB3 satisfies the following formula: ； Based on the proportion factors ZB1, ZB2, and ZB3, the adjustment coefficients θ1, θ2, and θ3 are determined; the adjustment coefficient θ1 satisfies the following formula: θ1=(ZB1 / BZ1) / (ZB1 / BZ1+ZB2 / BZ2+ZB3 / BZ3); where BZ1 is the anti-risk factor KZ i The standard proportion factor, BZ2 is the scheme factor AZ i The standard proportion factor, BZ3 is the capacity factor NL i The standard proportion factor of The adjustment coefficient θ2 satisfies the following formula: θ2=(ZB2 / BZ2) / (ZB1 / BZ1+ZB2 / BZ2+ZB3 / BZ3); The adjustment coefficient θ3 satisfies the following formula: θ3=(ZB3 / BZ3) / (ZB1 / BZ1+ZB2 / BZ2+ZB3 / BZ3).

8. The bidding risk early warning method based on big data according to claim 1 is characterized in that: The second warning threshold is set based on the bidding risk factors of several bidding enterprises, including: Extract the first warning threshold from the database, obtain the number of bidding enterprises in this tender whose bidding risk factor is less than the first warning threshold, and determine whether the ratio LB of the number of bidding enterprises to the number of bidding enterprises in this tender exceeds the ratio threshold; if yes, determine the second warning threshold YZ2 based on the ratio LB, and the second warning threshold YZ2 satisfies the following formula: YZ2=YZ1 / (1+ln(LB+1)); where YZ1 is the warning threshold 1; No, extract the manually set warning threshold 2 from the database.

9. The bidding risk early warning method based on big data according to claim 1 is characterized in that: The risk warning for bidding enterprises based on the first and second warning thresholds includes: B1: Extract bidding risk factors ZF of bidding enterprises in sequence i , determine the bidding risk factor ZF i Whether it exceeds the warning threshold 1; if yes, a warning message is issued to the tenderer indicating that the current bidding enterprise has bidding risks and is not recommended for selection; if no, jump to B2; B2: Determine the bidding risk factor ZF i Is it less than the warning threshold 2? If yes, the information that the current bidding enterprise is a high-quality enterprise will be sent to the tendering party; if no, the information that the current bidding enterprise is a medium-quality enterprise will be sent to the tendering party.

10. A bidding risk early warning system based on big data, operating based on a bidding risk early warning method based on big data according to any one of claims 1 to 9, characterized in that: include: Risk analysis module, and its connected information collection module, risk warning module and database; The information collection module is used to collect basic data, target data and historical bid-winning data of bidding enterprises; basic data includes qualification certificates, performance and working capital; target data includes the text of bidding documents and progress control plan; The risk analysis module is used to analyze basic data to obtain a risk resistance factor that reflects the bidder's risk resistance ability, analyze target data to obtain a solution factor that reflects the quality of the bidder's solution, analyze historical winning bid data to obtain a capability factor that reflects the bidder's professional ability, and combine the risk resistance factor, solution factor, and capability factor into a bidding risk factor that reflects the bidder's performance risk. The risk warning module is used to set a warning threshold 2 based on the bidding risk factors of several bidding companies, and obtain a pre-set warning threshold 1; and to issue risk warnings to the bidding companies based on the warning threshold 1 and the warning threshold 2.

Images