Scientific achievement transformation risk assessment method and system based on big data analysis

Abstract

The invention discloses a scientific achievement transformation risk assessment method and system based on big data analysis, and relates to the technical field of scientific achievement transformation risk assessment. The scientific achievement transformation risk assessment method based on big data analysis comprises the following steps: acquiring transformation feature data of scientific achievements, including structural performance data, application adaptation data and prototype test data; based on the conversion characteristic data, a conversion characteristic index set of the scientific research achievements is analyzed, and the conversion characteristic index set comprises a structure stability index, an adaptive capacity index and a conversion maturity index; based on the transformation characteristic index set, transformation risk assessment indexes of the scientific research achievements are analyzed, judgment and analysis are conducted on the transformation risk assessment indexes and a plurality of preset transformation risk assessment intervals, and each transformation risk assessment interval corresponds to a risk level. And quantitative analysis of different feature data is combined, so that the evaluation result is more accurate.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of scientific research achievement transformation risk assessment, in particular to a scientific research achievement transformation risk assessment method and system based on big data analysis. BACKGROUND

[0002] Under the background of accelerating technological innovation, the transformation of scientific research achievements has become a key factor in promoting technological progress and industrial upgrading. However, the transformation process of scientific research achievements is complex and uncertain, involving technology, market, policy and other factors. Traditional scientific research achievement transformation evaluation methods mainly rely on qualitative analysis and expert experience, which often cannot comprehensively and objectively evaluate the actual transformation potential and risks of scientific research achievements.

[0003] With the rapid development of big data technology, the evaluation method of scientific research achievement transformation has gradually transformed into a quantitative and data-driven direction. Big data analysis technology can extract effective information from massive data, help to comprehensively analyze various characteristics of scientific research achievements, and thus provide more accurate and systematic evaluation methods. However, existing big data-based evaluation methods still have certain limitations, at least including the following problems:

[0004] Existing technologies usually rely on qualitative analysis and lack systematic evaluation tools based on big data analysis and quantitative models, which makes it difficult to comprehensively and accurately evaluate the actual impact of various key factors on the transformation effect in the transformation process of scientific research achievements. Existing technologies usually only focus on a single indicator, such as the performance of laboratory environment, however, the multi-dimensional factors such as adaptability and maturity involved in the transformation process have not been effectively quantitatively analyzed. Especially when facing different actual application scenarios, existing methods fail to fully consider the environmental adaptability, system compatibility and stability of technology and other key factors, resulting in that scientific research achievements are often difficult to well integrate with existing technologies or market demands in actual application, causing the failure or poor effect of technology transformation. SUMMARY

[0005] In view of the deficiencies of the prior art, the present application provides a scientific research achievement transformation risk assessment method and system based on big data analysis, which solves the problem of lack of multi-dimensional and dynamic quantitative evaluation in the prior art, which makes it difficult to accurately predict the transformation potential and risks of scientific research achievements.

[0006] To achieve the above object, the present application is implemented by the following technical solutions: a scientific research achievement transformation risk assessment method based on big data analysis, comprising the following steps: obtaining transformation characteristic data of scientific research achievements, including structural performance data, application adaptation data and prototype test data; based on the transformation characteristic data, analyzing a transformation characteristic index set of the scientific research achievements, including a structural stability index, an adaptation capability index and a transformation maturity index; based on the transformation characteristic index set, analyzing a transformation risk assessment index of the scientific research achievements, and performing judgment analysis with a plurality of preset transformation risk assessment intervals, and each transformation risk assessment interval corresponds to a risk level;

[0007] Wherein, the specific formula for calculating the transformation risk assessment index of the scientific research achievements is as follows: ; wherein, , , , The transformation risk assessment index, the structural stability index, the adaptation capability index and the transformation maturity index of the scientific research achievements are in turn, , , The structural risk assessment coefficient, the adaptation risk assessment coefficient and the maturity risk assessment coefficient stored in the database are in turn.

[0008] Further, the structural performance data includes a prototype maximum power consumption value, a prototype minimum working temperature value, a prototype maximum working temperature value, a prototype response delay measured value in each functional trigger test, and a prototype output error measured value in each functional operation process.

[0009] Further, the specific steps for analyzing the structural stability index of the scientific research achievements are as follows: based on a pre-trained response characteristic analysis network, the prototype response delay measured value in each functional trigger test and the prototype output error measured value in each functional operation process of the scientific research achievements are analyzed to obtain response characteristic data of the scientific research achievements, including prototype average response time, prototype maximum response time, prototype response time fluctuation amplitude, prototype maximum error value, prototype error standard deviation and prototype error offset; the prototype maximum power consumption value, the prototype minimum working temperature value and the prototype maximum working temperature value of the scientific research achievements are respectively combined with the response characteristic data for comprehensive analysis to obtain the structural stability index of the scientific research achievements.

[0010] Further, the response feature analysis network comprises an input processing layer and a feature extraction layer, and the specific steps of obtaining the response feature data of the scientific research achievement are as follows: in the input processing layer of the response feature analysis network, the response delay measured value of the prototype of the scientific research achievement in each function trigger test and the output error measured value in each function running process are subjected to standardization processing to obtain standard processing response data; in the feature extraction layer of the response feature analysis network, the standard processing response data is subjected to statistical analysis, the response feature data is extracted, and output processing is performed.

[0011] Further, the application adaptation data comprises a prototype rated working voltage, a prototype installation volume, a prototype communication rate measured value in each communication session, and a prototype system timestamp value when each fault event occurs.

[0012] Further, the specific steps of analyzing the adaptation capability index of the scientific research achievement are as follows: based on the pre-trained performance feature analysis network, the prototype communication rate measured value of the prototype of the scientific research achievement in each communication session and the prototype system timestamp value when each fault event occurs are subjected to feature analysis to obtain the performance feature data of the scientific research achievement, including the prototype average communication rate, the prototype maximum communication rate, the prototype communication rate fluctuation amplitude, the prototype fault event occurrence frequency, and the prototype average fault interval length; the prototype rated working voltage and the prototype installation volume of the scientific research achievement are respectively combined with the adaptation capability feature data for comprehensive analysis to obtain the adaptation capability index of the scientific research achievement.

[0013] Further, the performance feature analysis network comprises a data preprocessing layer and a performance analysis layer, and the specific steps of obtaining the performance feature data of the scientific research achievement are as follows: in the data preprocessing layer of the performance feature analysis network, the prototype communication rate measured value of the prototype of the scientific research achievement in each communication session and the prototype system timestamp value when each fault event occurs are subjected to standardization processing to obtain standard processing performance feature data; in the performance analysis layer of the performance feature analysis network, the standard processing performance feature data is subjected to statistical analysis, the performance feature data is extracted, and output processing is performed.

[0014] Further, the prototype test data comprises a prototype single function test time consumption, a prototype cumulative integration test number, a prototype manufacturing qualified rate, and a prototype function test pass rate.

[0015] Further, the specific steps of analyzing the transformation maturity index of the scientific research achievement are as follows: the prototype single function test time consumption of the scientific research achievement is read and subjected to standardization processing; the standardization processed prototype single function test time consumption of the scientific research achievement is respectively combined with the corresponding prototype cumulative integration test number, prototype manufacturing qualified rate, and prototype function test pass rate for comprehensive analysis to obtain the transformation maturity index of the scientific research achievement.

[0016] The application discloses a scientific research achievement transformation risk assessment system based on big data analysis.

[0017] The application has the following beneficial effects:

[0018] (1) The scientific research achievement transformation risk assessment method based on big data analysis can comprehensively consider structural performance data, application adaptation data and prototype test data of scientific research achievements, calculate structural stability indexes, adaptation capability indexes and transformation maturity indexes, and combine quantitative analysis of different characteristic data, so that the assessment result is more accurate, scientific research personnel and decision makers can better identify risks that may be faced by scientific research achievements in the transformation process according to the comprehensive indexes, avoid misjudgment caused by single-dimensional assessment, optimize transformation strategies of scientific research achievements, improve the transformation success rate, and ensure that scientific research achievements can smoothly enter the market.

[0019] (2) The scientific research achievement transformation risk assessment method based on big data analysis can predict various risks that may be encountered in the transformation process of scientific research achievements in real time and accurately, for example, by collecting and analyzing prototype test data in real time, the response delay, power consumption, temperature change and other characteristics of scientific research achievements are evaluated, so that potential risks in the transformation process, such as insufficient adaptability or poor technical stability, can be dynamically identified, and the structural stability, adaptation capability and transformation maturity are analyzed in combination with power consumption, working temperature, response delay and other data of the prototype, and then comprehensive assessment indexes such as structural stability, adaptation capability and transformation maturity are obtained, and scientific research personnel can adjust the technology transformation strategy in advance based on the assessment result, identify and avoid potential risks, not only reduce uncontrollable factors in the transformation process, but also significantly improve the transformation efficiency of scientific research achievements from the laboratory to the market, and provide more scientific decision support for scientific research personnel and enterprises.

[0020] (3), the scientific research achievement transformation risk assessment system based on big data analysis, through the introduction of the transformation data acquisition module, so as to automatically obtain the structural performance, application adaptability and prototype test data of scientific research achievements from multiple data sources, eliminate the low efficiency problem of relying on manual collection and processing data in traditional method, the transformation feature acquisition module based on automatic data analysis can quickly calculate the structural stability index, adaptability index and transformation maturity index, ensure the efficiency and reliability of the evaluation process, through the transformation risk assessment module, the system can automatically judge and evaluate the risk level of scientific research achievements on the basis of real-time acquisition of transformation characteristic data, help scientific researchers and enterprises get accurate transformation risk assessment results in the shortest time, the automation function of the system greatly saves time and labor cost, improves the efficiency of scientific research achievement transformation, especially suitable for scientific and technological enterprises and scientific research institutions which need to make quick decisions and respond to complex changes.

[0021] Of course, implementing any product of the present application does not necessarily require all the advantages described above to be achieved simultaneously. BRIEF DESCRIPTION OF DRAWINGS

[0022] Figure 1 A flow chart of a scientific research achievement transformation risk assessment method based on big data analysis.

[0023] Figure 2 A flow chart of the specific steps of obtaining response characteristic data of scientific research achievements in a scientific research achievement transformation risk assessment method based on big data analysis.

[0024] Figure 3 A block diagram of a scientific research achievement transformation risk assessment system based on big data analysis. DETAILED DESCRIPTION

[0025] Please refer to Figure 1 The embodiment of the present application provides a technical scheme: a scientific research achievement transformation risk assessment method based on big data analysis, comprising the following steps: obtaining transformation characteristic data of scientific research achievements, including structural performance data, application adaptation data and prototype test data; based on the transformation characteristic data, analyzing the transformation characteristic index set of scientific research achievements, including the structural stability index, the adaptability index and the transformation maturity index; based on the transformation characteristic index set, analyzing the transformation risk assessment index of scientific research achievements, and judging and analyzing with a plurality of preset transformation risk assessment intervals, and each transformation risk assessment interval corresponds to a risk level, including but not limited to the following examples:

[0026] The transformation risk assessment index ∈ [0-1.0]: the risk level is high risk, indicating that the scientific research achievement faces higher transformation risk.

[0027] The transformation risk evaluation index ∈ (1.0-2.0]: the risk level is high, indicating that the scientific research results may face certain risks.

[0028] The transformation risk evaluation index ∈ (2.0-3.0]: the risk level is medium, indicating that the scientific research results have a certain degree of technical maturity, but there are still some potential problems to be solved.

[0029] The transformation risk evaluation index ∈ (3.0-4.0]: the risk level is low, indicating that the scientific research results are basically mature, the technology is stable, the adaptability and market potential are strong, and the transformation process risk is low.

[0030] The transformation risk evaluation index > 4.0: the risk level is extremely low, indicating that the scientific research results are fully mature, and the technical indicators and market adaptability are very excellent, with almost no transformation risk.

[0031] The specific formula for calculating the transformation risk evaluation index of scientific research results is as follows: ; wherein, , , , The transformation risk evaluation index, the structural stability index, the adaptability index, and the transformation maturity index of scientific research results are, , , The structural risk evaluation coefficient, the adaptability risk evaluation coefficient, and the maturity risk evaluation coefficient stored in the database are.

[0032] It needs to be explained that the specific acquisition steps of the structural risk evaluation coefficient , the adaptability risk evaluation coefficient , and the maturity risk evaluation coefficient stored in the database are: sum the structural stability index, the adaptability index, and the transformation maturity index of scientific research results for proportion analysis, and the proportion analysis result is the corresponding coefficient.

[0033] The specific implementation example of calculating the transformation risk evaluation index of scientific research results is as follows, and the existing parameters are as follows:

[0034] The structural stability index of scientific research results is about 0.745.

[0035] The adaptability index of scientific research results is about 0.832.

[0036] The transformation maturity index of scientific research results is about 0.653.

[0037] The structural risk evaluation coefficient stored in the database is 0.745 / (0.745+0.832+0.653)≈0.334.

[0038] The stored adaptation risk assessment coefficient in the database = 0.832 / (0.745+0.832+0.653)≈0.373.

[0039] The stored maturity risk assessment coefficient in the database = 0.653 / (0.745+0.832+0.653)≈0.293.

[0040] The above data are respectively substituted into the specific formula for calculating the transformation risk assessment index of scientific research results to obtain:

[0041] The transformation risk assessment index of scientific research results = ((0.745 / (1+0.745))^0.334) + ((0.832 / (1+0.832))^0.373) + ((0.653 / (1+0.653))^0.293)≈2.259.

[0042] Specifically, the structural performance data include the maximum power consumption value of the prototype, the minimum working temperature value of the prototype, the maximum working temperature value of the prototype, the measured response delay value of the prototype in each functional trigger test, and the measured output error value of the prototype in each functional operation process.

[0043] The specific steps for analyzing the structural stability index of scientific research results are as follows: based on the pre-trained response feature analysis network, the measured response delay value of the prototype of scientific research results in each functional trigger test and the measured output error value of the prototype in each functional operation process are analyzed to obtain the response feature data of scientific research results, including the average response time of the prototype, the maximum response time of the prototype, the response time fluctuation amplitude of the prototype, the maximum error value of the prototype, the error standard deviation of the prototype, and the error offset of the prototype; the maximum power consumption value of the prototype of scientific research results, the minimum working temperature value of the prototype, and the maximum working temperature value of the prototype are respectively combined with the response feature data for comprehensive analysis to obtain the structural stability index of scientific research results.

[0044] The pre-training steps of the response feature analysis network are as follows:

[0045] In the pre-training process of the response feature analysis network, first, the response delay measured value in each function trigger test of the prototype and the output error measured value in the function running process are preprocessed. First, all the response delay measured values and the output error measured values are standardized. For the response delay measured values, the maximum value and the minimum value are calculated, and each response delay value is subtracted by the minimum value and divided by the difference between the maximum value and the minimum value to obtain the normalized response data. For the output error measured values, the mean and the standard deviation of the error data are calculated, and the z-score standardization method is used to subtract the mean from each error value and divide by the standard deviation to ensure the dimensionality of the data. Next, through data enhancement technology, the original data is expanded to generate additional sample data under different test conditions to ensure that the network can learn more input features.

[0046] After data preprocessing and enhancement, the training phase of the model is entered. The network is trained through supervised learning, using the standardized response delay measured value and the output error measured value as input data. The goal is to minimize the error between the predicted value and the actual value by adjusting the network parameters. During the training process, mean square error (MSE) is used as the loss function to calculate the difference between each predicted output and the actual target value, and the network weights are optimized through the backpropagation algorithm. To improve the generalization ability of the model, cross-validation is used during training to divide the data into training and validation sets, ensuring that the model not only performs well on training data but also has good prediction results on unseen data. After multiple iterations of optimization, the network gradually adjusts the weights until the loss function converges, and finally a trained response feature analysis network model is obtained.

[0047] In calculating the structural stability index of scientific research achievements, the prototype maximum power consumption value, the prototype minimum operating temperature value, the prototype maximum operating temperature value, the prototype average response time, the prototype maximum response time, and the prototype response time fluctuation amplitude need to be unit-free.

[0048] The specific formula for calculating the structural stability index of scientific research achievements is as follows: ; wherein, , , , , , , , , , The structural stability index of scientific research achievements, the prototype maximum power consumption value, the prototype maximum operating temperature value, the prototype minimum operating temperature value, the prototype average response time, the prototype maximum response time, the prototype response time fluctuation amplitude, the prototype maximum error value, the prototype error standard deviation, and the prototype error offset are in turn 、 、 、 The power consumption control coefficient, the temperature difference control coefficient, the response control coefficient and the error control coefficient stored in the database are sequentially obtained.

[0049] It needs to be explained that the specific acquisition steps of the power consumption control coefficient , the temperature difference control coefficient , the response control coefficient , and the error control coefficient stored in the database are as follows:

[0050] The highest working temperature value and the lowest working temperature value of the prototype after unit removal processing of the scientific research achievements are calculated by difference to obtain the corresponding working temperature difference value;

[0051] The average response time, the maximum response time and the response time fluctuation amplitude of the prototype after unit removal processing of the scientific research achievements are summed to obtain the corresponding response sum value;

[0052] The maximum error value, the error standard deviation and the error offset of the prototype of the scientific research achievements are summed to obtain the corresponding error sum value;

[0053] The working temperature difference value, the response sum value, the error sum value and the maximum power consumption value after unit removal processing of the scientific research achievements are summed and analyzed by proportion, and the proportion analysis result is the corresponding coefficient.

[0054] In this embodiment, by introducing big data analysis and multi-dimensional feature data processing, the accuracy and reliability of the scientific research achievement transformation risk assessment are significantly improved. First, the response delay, output error and other data of the prototype are standardized and enhanced by the response feature analysis network, which can effectively eliminate the noise in the data and ensure the uniformity and comparability of the input data. This process provides an accurate basis for subsequent structure stability index calculation. Through comprehensive analysis of power consumption, temperature difference, response time and error, combined with the corresponding control coefficient, the influence of each index on the transformation of scientific research achievements can be quantified, so that the stability and adaptability can be more scientifically evaluated. This data-driven method can timely discover potential technical problems and help researchers and decision-makers optimize transformation strategies to avoid uncertainty risks. In addition, the unit removal processing and proportion analysis method enables different dimensions of features to be integrated, resulting in more comprehensive and accurate evaluation results.

[0055] Specifically, as Figure 2As shown, the response feature analysis network includes an input processing layer and a feature extraction layer, and the specific steps of obtaining the response feature data of the scientific research achievement are as follows: in the input processing layer of the response feature analysis network, the response delay measured value of the prototype of the scientific research achievement in each function trigger test and the output error measured value in each function running process are standardized (i.e. normalized, missing value completion and time axis alignment processing), and the standard processing response data is obtained, which is specifically: the response delay measured value in each function trigger test is normalized, the maximum value and the minimum value of all response delay values are calculated, each response delay value is subtracted by the minimum value and divided by the difference between the maximum value and the minimum value, and the standardized response delay data is obtained; the output error measured value in each function running process is processed by z-score standardization, the mean and standard deviation of the output error are calculated, each output error value is subtracted by the mean and divided by the standard deviation, and the standardized error data is obtained; the timestamp data is sorted and aligned to ensure that the response delay and output error data are arranged in the same time sequence, and the time-aligned response data is obtained;

[0056] In the feature extraction layer of the response feature analysis network, the standard processing response data is statistically analyzed, the response feature data is extracted, and the output processing is performed, which is specifically: the average response time of the prototype is calculated, the average value of all response delay measured values is taken; the maximum response time of the prototype is calculated, the maximum value in the response delay measured value is found; the response time fluctuation amplitude of the prototype is calculated, the difference between the maximum value and the minimum value in the response delay measured value is taken; the maximum error value of the prototype is calculated, the maximum value in the output error measured value is found; the error standard deviation of the prototype is calculated, the standard deviation of the output error measured value is calculated; the error offset of the prototype is calculated, the average absolute deviation of the output error measured value is calculated, and the error offset is obtained.

[0057] In this embodiment, the response feature analysis network provides a more scientific and efficient evaluation method for the transformation of scientific research achievements through accurate data preprocessing and feature extraction. First, the input processing layer standardizes the response delay and output error of the prototype in the function trigger test, ensuring that all data is unified and comparable. The response delay data is processed by normalization to eliminate the interference of extreme values, allowing comparison on the same scale; the output error is processed by the z-score standardization method to ensure that the data dimension is unified, avoiding data errors caused by inconsistent dimensions, and the time axis alignment ensures the synchronization of different data to improve the accuracy of the data. In the feature extraction layer, the calculated response feature data, such as average response time, maximum response time, fluctuation amplitude, error standard deviation and error offset, can comprehensively reflect the performance of the prototype under different test conditions. This process provides accurate and reliable data support for subsequent risk assessment of scientific research achievement transformation, helping decision-makers better understand the stability and adaptability of scientific research achievements.

[0058] Specifically, the application adaptation data includes a prototype rated operating voltage, a prototype installation volume, a prototype communication rate measured value in each communication session, and a system timestamp value of the prototype at each fault event occurrence.

[0059] The specific steps of analyzing the adaptation capability index of the scientific research achievement are as follows: based on the pre-trained performance feature analysis network, the communication rate measured value of the prototype of the scientific research achievement in each communication session and the system timestamp value of the prototype at each fault event occurrence are analyzed to obtain performance feature data of the scientific research achievement, including prototype average communication rate, prototype maximum communication rate, prototype communication rate fluctuation amplitude, prototype fault event occurrence frequency, and prototype average fault interval length; the prototype rated operating voltage and the prototype installation volume of the scientific research achievement are analyzed comprehensively in combination with the adaptation capability feature data to obtain the adaptation capability index of the scientific research achievement.

[0060] The pre-training steps of the performance feature analysis network are as follows:

[0061] In the pre-training process of the performance feature analysis network, first, the communication rate measured value in each communication session of the prototype and the system timestamp value at each fault event occurrence are preprocessed. First, all the communication rate measured values are standardized, the mean and standard deviation are calculated, and the z-score standardization method is used to convert them into standardized communication rate data. Then, all the system timestamp values are sorted in chronological order, the time interval between adjacent two fault event timestamps is calculated to obtain the fault interval length sequence, and the sequence is normalized to calculate the maximum and minimum values. Each interval value is subtracted by the minimum value and divided by the difference to obtain the normalized fault interval data. Next, data augmentation techniques are used to expand the original data to generate additional sample data under different test conditions to ensure that the network can learn more input features.

[0062] After data preprocessing and enhancement, the network is trained in the training stage using a supervised learning method. The standardized communication rate sequence and the fault interval sequence are used as input, and the known adaptation performance evaluation label of the corresponding prototype is set as the target output. The mean square error is used as the loss function in the training, the weights are updated by back propagation, and cross-validation is performed during the training process. The data set is divided into a training set and a validation set, and the training is iterated until convergence. Finally, the trained performance feature analysis network model is obtained.

[0063] In the calculation of the adaptation capability index of the scientific research achievement, the prototype rated operating voltage, the prototype installation volume, the prototype average communication rate, the prototype maximum communication rate, the prototype communication rate fluctuation amplitude, and the prototype average fault interval length need to be unit-free.

[0064] The specific formula for calculating the adaptation capability index of scientific research results is as follows: ; wherein, , , , , , , , The adaptation capability index of scientific research results, the average communication rate of the prototype, the communication rate fluctuation amplitude of the prototype, the maximum communication rate of the prototype, the frequency of failure events of the prototype, the average failure interval length of the prototype, the rated working voltage of the prototype, and the installation volume of the prototype, is a natural constant, and in this embodiment, the value is 2.71, , , The communication control coefficient, the failure control coefficient, and the configuration control coefficient stored in the database are in turn.

[0065] It needs to be explained that the specific acquisition steps of the communication control coefficient , the failure control coefficient , and the configuration control coefficient stored in the database are as follows:

[0066] The calculation result of the unit-processed (prototype average communication rate / (1+prototype communication rate fluctuation amplitude))×(prototype maximum communication rate / (1+prototype failure event occurrence frequency)) of scientific research results is taken as the corresponding communication sum value;

[0067] The calculation result of the unit-processed (1+(prototype average failure interval length / (prototype rated working voltage×prototype installation volume))) of scientific research results is taken as the corresponding failure sum value;

[0068] The calculation result of (2.71^(unit-processed prototype rated working voltage×prototype installation volume) of scientific research results) is taken as the corresponding configuration sum value;

[0069] The communication sum value, the failure sum value, and the configuration sum value of scientific research results are summed up for proportion analysis, and the proportion analysis result is the corresponding coefficient.

[0070] In this embodiment, by introducing a performance characteristic analysis network based on big data analysis, the adaptation capability of scientific research achievements in the transformation process can be comprehensively analyzed. Through standardized processing and preprocessing of key data such as the communication rate, fault event frequency, and fault interval length of the prototype, the consistency and comparability of the input data are ensured, the errors caused by data differences are reduced, the sample data is expanded using data enhancement technology, the model can learn more features, and the accuracy of the evaluation is improved. When calculating the adaptation capability index, by comprehensively analyzing the communication performance, fault frequency, power consumption, and other factors of the prototype, and combining with the communication control coefficient, fault control coefficient, and configuration control coefficient stored in the database, the adaptability and market application potential of scientific research achievements can be comprehensively evaluated. This method not only provides more scientific and accurate evaluation results, but also dynamically reflects the adaptation of scientific research achievements in the transformation process, thereby helping decision-makers make better technical adjustments and optimization strategies, and improving the market adaptability and transformation efficiency of scientific research achievements.

[0071] Specifically, the performance characteristic analysis network includes a data preprocessing layer and a performance analysis layer. The specific steps of obtaining the performance characteristic data of scientific research achievements are as follows: in the data preprocessing layer of the performance characteristic analysis network, the measured value of the communication rate of the prototype of scientific research achievements in each communication session and the system timestamp value at the occurrence of each fault event are standardized (i.e. normalized, missing value completion and time axis alignment processing), to obtain standard processing performance characteristic data, which is specifically: the measured value of the communication rate in each communication session is subjected to z-score standardization processing, the mean and standard deviation of the communication rate are calculated, each communication rate value is subtracted from the mean and divided by the standard deviation to obtain the standardized communication rate data; the system timestamp data of each fault event is sorted and aligned, the time difference between adjacent two fault events is calculated to obtain the fault interval length sequence, the maximum and minimum values are calculated, and the standardized fault interval data is obtained by normalization method;

[0072] In the performance analysis layer of the performance characteristic analysis network, the standard processing performance characteristic data is subjected to statistical analysis, the performance characteristic data is extracted, and output processing is performed, which is specifically: the average communication rate of the prototype is calculated, and the average value of the measured value of the communication rate is taken; the maximum communication rate of the prototype is calculated, and the maximum value in the measured value of the communication rate is found; the communication rate fluctuation amplitude of the prototype is calculated, and the difference between the maximum and minimum values in the measured value of the communication rate is taken; the fault event frequency of the prototype is calculated, and the ratio of the number of faults occurring in the test to the total number of tests is calculated; the average fault interval length of the prototype is calculated, and the average value of the fault interval sequence is taken.

[0073] In this embodiment, the performance feature analysis network improves the accuracy of scientific research achievement transformation risk assessment through accurate data preprocessing and performance analysis. First, the data preprocessing layer ensures the consistency and comparability of the data by standardizing and aligning the timestamp data of the communication rate and failure events. The communication rate is standardized by z-score to ensure the dimensionality of different test data, avoiding errors caused by data differences. At the same time, the failure interval data is normalized to make the failure frequency data more comparable. The performance analysis layer further analyzes the processed data to extract key performance indicators such as average communication rate, maximum communication rate, communication fluctuation amplitude, and failure event frequency. These indicators can comprehensively reflect the performance and stability of the prototype in actual application.

[0074] Specifically, the prototype test data includes prototype single function test time consumption, prototype cumulative integration test times, prototype manufacturing qualified rate, and prototype function test pass rate.

[0075] The specific steps of analyzing the transformation maturity index of scientific research achievements are as follows: reading the prototype single function test time consumption of scientific research achievements and performing standardization processing (i.e., unit removal processing); combining the standardized prototype single function test time consumption of scientific research achievements with the corresponding prototype cumulative integration test times, prototype manufacturing qualified rate, and prototype function test pass rate for comprehensive analysis to obtain the transformation maturity index of scientific research achievements.

[0076] The specific formula for calculating the transformation maturity index of scientific research achievements is as follows: ; wherein, , , , The transformation maturity index of scientific research achievements, prototype cumulative integration test times, prototype manufacturing qualified rate, and prototype function test pass rate are in turn, The standardized prototype single function test time consumption of scientific research achievements, is a natural constant, which is taken as 2.71 in this embodiment, , , The test control coefficient, qualified control coefficient, and mature control coefficient stored in the database are in turn.

[0077] It needs to be explained that the specific acquisition steps of the test control coefficient , qualified control coefficient , and mature control coefficient stored in the database are as follows:

[0078] The calculation result of (prototype cumulative integrated test times / (1 + prototype single function test time after unit processing)) of the scientific research achievement is taken as the corresponding test and value;

[0079] The calculation result of ((prototype manufacturing yield rate x prototype function test pass rate) ^ (1 / 2)) of the scientific research achievement is taken as the corresponding qualified and value;

[0080] The calculation result of (1 - (2.71 ^ (prototype function test pass rate of scientific research achievement))) is taken as the corresponding mature and value;

[0081] The test and value, the qualified and value, and the mature and value of the scientific research achievement are summed up for proportion analysis, and the proportion analysis result is the corresponding coefficient.

[0082] In the embodiment, the evaluation precision of the conversion maturity of scientific research achievements is significantly improved through accurate prototype test data processing and comprehensive analysis. First, the single function test time of the prototype is standardized processed to ensure the consistency and comparability of the data and eliminate the influence of unit differences. Then, the maturity of the scientific research achievements is comprehensively evaluated by combining the cumulative integrated test times, the manufacturing yield rate and the function test pass rate of the prototype. The conversion maturity index of the scientific research achievements is obtained by calculating the corresponding test, qualified and mature coefficients and performing proportion analysis on the characteristic data. The comprehensive analysis method can accurately quantify the maturity and market adaptability of the scientific research achievements, and help researchers and decision makers evaluate the actual conversion potential of the achievements.

[0083] Please refer to Figure 3 The embodiment of the present application provides a technical solution: a scientific research achievement conversion risk assessment system based on big data analysis, comprising: a conversion data acquisition module for acquiring conversion characteristic data of scientific research achievements, including structural performance data, application adaptation data and prototype test data; a conversion characteristic acquisition module for analyzing a conversion characteristic index set of scientific research achievements based on the conversion characteristic data, including a structural stability index, an adaptation capability index and a conversion maturity index; a conversion risk assessment module for analyzing a conversion risk assessment index of scientific research achievements based on the conversion characteristic index set, and performing judgment analysis with a plurality of preset conversion risk assessment intervals, and each conversion risk assessment interval corresponds to a risk level.

[0084] Although the preferred embodiments of the present application have been described, those skilled in the art can make further changes and modifications to the embodiments once they know the basic inventive concept. Therefore, the appended claims are intended to be interpreted as including all changes and modifications falling within the scope of the present application.

[0085] Obviously, many modifications and variations of the present application are possible in light of the above teachings. It is, therefore, to be understood that within the scope of the appended claims and their equivalents, the application can be practiced otherwise than as specifically described.

Claims

1. A risk assessment method for scientific research results transformation based on big data analysis, characterized in that: The following steps are involved: Obtain transformation characteristic data of scientific research results, including structural performance data, application adaptation data and prototype test data; Based on the transformation characteristic data, the transformation characteristic index set of scientific research results is analyzed, including the structural stability index, adaptability index, and transformation maturity index; Based on the transformation characteristic index set, the transformation risk assessment index of scientific research results is analyzed and compared with several preset transformation risk assessment intervals, and each transformation risk assessment interval corresponds to a risk level; The specific formula for calculating the transformation risk assessment index of scientific research results is as follows: ； in, 、 、 、 They are the transformation risk assessment index, structural stability index, adaptability index, and transformation maturity index of scientific research results. 、 、 They are the structural risk assessment coefficient, adaptation risk assessment coefficient, and maturity risk assessment coefficient stored in the database respectively.

2. The method for risk assessment of scientific research achievement transformation based on big data analysis according to claim 1, characterized in that: The structural performance data includes the maximum power consumption value of the prototype, the minimum operating temperature value of the prototype, the maximum operating temperature value of the prototype, the measured value of the response delay of the prototype in each function trigger test, and the measured value of the output error of the prototype during each function operation.

3. The method for risk assessment of scientific research achievement transformation based on big data analysis according to claim 2, characterized in that: The specific steps for analyzing the structural stability index of scientific research results are as follows: Based on the pre-trained response feature analysis network, the measured response delay values ​​of the prototype of the scientific research results in each function trigger test and the measured output error values ​​of the prototype during each function operation are analyzed to obtain the response feature data of the scientific research results, including the average response time of the prototype, the maximum response time of the prototype, the fluctuation range of the prototype response time, the maximum error value of the prototype, the standard deviation of the prototype error, and the prototype error offset; The maximum power consumption value, minimum operating temperature value and maximum operating temperature value of the prototype of the scientific research results are comprehensively analyzed in combination with the response characteristic data to obtain the structural stability index of the scientific research results.

4. The method for risk assessment of scientific research achievement transformation based on big data analysis according to claim 3 is characterized in that: The response feature analysis network includes an input processing layer and a feature extraction layer. The specific steps for obtaining the response feature data of scientific research results are as follows: In the input processing layer of the response characteristic analysis network, the measured response delay values ​​of the prototype of the scientific research results in each function trigger test and the measured output error values ​​during each function operation are standardized to obtain standard processed response data; In the feature extraction layer of the response feature analysis network, statistical analysis is performed on the standard processing response data, response feature data is extracted, and output processing is performed.

5. The method for risk assessment of scientific research achievement transformation based on big data analysis according to claim 1, characterized in that: The application adaptation data includes the rated operating voltage of the prototype, the installation volume of the prototype, the measured communication rate value of the prototype in each communication session, and the system timestamp value of the prototype when each fault event occurs.

6. The method for risk assessment of scientific research achievement transformation based on big data analysis according to claim 5, characterized in that: The specific steps for analyzing the adaptability index of scientific research results are as follows: Based on the pre-trained performance characteristic analysis network, the measured communication rate of the prototype of the scientific research results in each communication session and the system timestamp value of the prototype when each failure event occurs are analyzed to obtain the performance characteristic data of the scientific research results, including the average communication rate of the prototype, the maximum communication rate of the prototype, the fluctuation range of the communication rate of the prototype, the frequency of prototype failure events, and the average failure interval length of the prototype; The rated working voltage and installation volume of the prototype of the scientific research results are comprehensively analyzed in combination with the adaptability characteristic data to obtain the adaptability index of the scientific research results.

7. The method for risk assessment of scientific research achievement transformation based on big data analysis according to claim 6, characterized in that: The performance characteristic analysis network includes a data preprocessing layer and a performance analysis layer. The specific steps for obtaining the performance characteristic data of scientific research results are as follows: In the data preprocessing layer of the performance characteristic analysis network, the measured communication rate values ​​of the prototype of the scientific research results in each communication session and the system timestamp value when each fault event occurs are standardized to obtain standard processing performance characteristic data; In the performance analysis layer of the performance characteristic analysis network, statistical analysis is performed on the standard processing performance characteristic data, the performance characteristic data is extracted, and output processing is performed.

8. The method for risk assessment of scientific research achievement transformation based on big data analysis according to claim 1, characterized in that: The prototype test data includes the time consumed for a single functional test of the prototype, the cumulative number of integration tests of the prototype, the prototype manufacturing qualification rate, and the prototype functional test pass rate.

9. The method for risk assessment of scientific research achievement transformation based on big data analysis according to claim 8, characterized in that: The specific steps for analyzing the transformation maturity index of scientific research results are as follows: Reading the scientific research results of the prototype single functional test is time-consuming and requires standardization; The time consumption of a single functional test of a standardized prototype of scientific research results is comprehensively analyzed in combination with the corresponding cumulative number of prototype integration tests, prototype manufacturing qualification rate, and prototype functional test pass rate to obtain the transformation maturity index of scientific research results.

10. A scientific research achievement transformation risk assessment system based on big data analysis, applying the scientific research achievement transformation risk assessment method based on big data analysis according to any one of claims 1 to 9, characterized in that: include: The transformation data acquisition module is used to obtain the transformation characteristic data of scientific research results, including structural performance data, application adaptation data and prototype test data; The transformation feature acquisition module is used to analyze the transformation feature index set of scientific research results based on the transformation feature data, including the structural stability index, adaptability index, and transformation maturity index; The transformation risk assessment module is used to analyze the transformation risk assessment index of scientific research results based on the transformation characteristic index set, and perform judgment analysis with several preset transformation risk assessment intervals, and each transformation risk assessment interval corresponds to a risk level.

Images