A method, apparatus, and storage medium for generating test standards

Abstract

The present invention provides a method, apparatus, and storage medium for generating test standards. The method includes: obtaining a combination of operating condition data when multiple testers conduct tests, where the combination of operating condition data includes at least one data feature and a time series of feature values corresponding to each data feature; determining a specific combination of operating condition data that meets a preset evaluation standard among all the combinations of operating condition data, inputting all the data features in the specific combination of operating condition data into a trained machine learning model to determine specific data features related to the operating habits of the testers; determining a value range of the specific data feature according to the time series of feature values corresponding to the specific data feature in the specific combination of operating condition data, and using the specific data feature and its value range as test standards. The technical solution of the present invention can reduce the influence of the operating habits of testers on test results.

Description

Technical Field

[0001] The present invention relates to the technical field of tests, and in particular, to a test standard generation method, device, and storage medium. Background Art

[0002] Testing equipment is an important measure to verify the reliability and safety of each component of the equipment. Its test items are usually fixed. For example, the test items of a crane test include static load test and dynamic load test. However, due to different operating habits of different testers, the test results obtained by different testers when performing the same test item may vary. For example, the test efficiency and test fuel consumption of different testers when performing the same crane test item are different, resulting in poor consistency of test results when different testers perform the same test. Summary of the Invention

[0003] The problem solved by the present invention is how to reduce the influence of the operating habits of testers on test results, so as to improve the consistency of test results when different testers perform the same test.

[0004] To solve the above problems, the present invention provides a test standard generation method, device, and storage medium.

[0005] In a first aspect, the present invention provides a test standard generation method, including:

[0006] Obtaining a combination of working condition data of multiple testers performing tests, where the combination of working condition data includes at least one data feature and a time series of feature values corresponding to each data feature;

[0007] Determining a specific combination of working condition data that meets a preset evaluation criterion among all the combinations of working condition data, and inputting all the data features in the specific combination of working condition data into a trained machine learning model to determine specific data features related to the operating habits of the testers;

[0008] Determining a value range of the specific data feature according to the time series of feature values corresponding to the specific data feature in the specific combination of working condition data, and using the specific data feature and its value range as a test standard.

[0009] Optionally, the number of the specific combinations of working condition data and the number of the specific data features are both at least one. After determining the specific data features related to the operating habits of the testers, the method further includes:

[0010] Calculating the coincidence degree between the time series of feature values of the same specific data feature in each specific combination of working condition data;

[0011] Compare the coincidence degrees of the respective specific data features with a preset threshold, and retain the specific data features whose coincidence degrees are greater than or equal to the preset threshold according to the comparison results.

[0012] Optionally, calculating the coincidence degree between the eigenvalue time series of the same specific data feature in each of the specific working condition data combinations includes:

[0013] For any one of the specific data features, respectively determine the first variance of the eigenvalue time series of the specific data feature in each of the specific working condition data combinations;

[0014] Calculate the second variance of all the first variances corresponding to the specific data feature, and the second variance is used to characterize the coincidence degree corresponding to the specific data feature.

[0015] Optionally, determining the value range of the specific data feature according to the eigenvalue time series corresponding to the specific data feature in the specific working condition data combination includes:

[0016] Determine the confidence interval of the specific data feature according to all the first variances of the specific data feature;

[0017] Determine the value range of the specific data feature according to the interval values of the confidence interval.

[0018] Optionally, the test standard generation method further includes: obtaining the test data of each of the testers for the test, and the test data includes the test duration and / or the test fuel consumption;

[0019] The specific working condition data combination includes a recommended data combination and / or a non-recommended data combination, and determining the specific working condition data combination that meets the preset evaluation criteria among all the working condition data combinations includes:

[0020] Among all the working condition data combinations, determine that the working condition data combination with the test duration less than or equal to the first preset duration threshold and / or the test fuel consumption less than or equal to the first preset fuel consumption threshold is the recommended data combination;

[0021] And / or,

[0022] Among all the working condition data combinations, determine that the working condition data combination with the test duration greater than or equal to the second preset duration threshold and / or the test fuel consumption greater than or equal to the second preset fuel consumption threshold is the non-recommended data combination.

[0023] Optionally, the test standard generation method further includes: obtaining the test data of each of the testers for the test, and the test data includes the test duration and / or the test fuel consumption;

[0024] The specific operating condition data combination includes a recommended data combination and / or a non-recommended data combination. Determining the specific operating condition data combination that meets the preset evaluation criteria among all the operating condition data combinations includes:

[0025] Selecting the first preset number of the operating condition data combinations in ascending order of the test duration as the recommended data combination, and / or selecting the second preset number of the operating condition data combinations in ascending order of the test fuel consumption as the recommended data combination;

[0026] and / or

[0027] Selecting the third preset number of the operating condition data combinations in descending order of the test duration as the non-recommended data combination, and / or selecting the fourth preset number of the operating condition data combinations in descending order of the test fuel consumption as the non-recommended data combination.

[0028] Optionally, before inputting all the data features in the specific operating condition data combination into the trained machine learning model, it further includes:

[0029] Obtaining a plurality of different labeled data features, where the labels include those related to the operator's operation habits and those not related to the operator's operation habits;

[0030] Training a pre-established machine learning model with all the labeled data features to obtain the trained machine learning model.

[0031] In a second aspect, the present invention provides a test standard generation device, including:

[0032] An acquisition module for acquiring the operating condition data combinations of multiple testers during the test, where the operating condition data combination includes at least one data feature and the feature value time series corresponding to each data feature;

[0033] A processing module for determining the specific operating condition data combination that meets the preset evaluation criteria among all the operating condition data combinations, inputting all the data features in the specific operating condition data combination into the trained machine learning model, and determining the specific data features related to the operator's operation habits;

[0034] A generation module for determining the value range of the specific data feature according to the feature value time series corresponding to the specific data feature in the specific operating condition data combination, and using the specific data feature and its value range as the test standard.

[0035] In a third aspect, the present invention provides a computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, the test standard generation method described in any one of the first aspects is implemented.

[0036] In a fourth aspect, the present invention provides a crane test monitoring platform, including a memory and a processor;

[0037] The memory is used to store a computer program;

[0038] The processor is used to implement the test standard generation method described in any one of the first aspects when executing the computer program.

[0039] The beneficial effects of the test standard generation method, device and storage medium of the present invention are as follows: Obtain the combined working condition data obtained by multiple testers conducting tests respectively. Each tester corresponds to a combined working condition data, and the combined working condition data includes at least one data feature and its feature value time series. Determine specific combined working condition data that meet the preset evaluation criteria among all the combined working condition data. The preset evaluation criteria can be specifically set according to the optimization objectives of the test standard. For example, the combined working condition data with the highest test efficiency. Screen out specific data features related to the operation habits of the testers from all the data features through a trained machine learning model, and determine the value range of the specific data features based on the feature values at different time points in the specific combined working condition data that meet the preset evaluation criteria. Use the specific data feature and its value range as the test standard. When this test standard is used to guide the testers to conduct tests, it can make the test results close to the preset evaluation criteria. The test standard generated by the present invention is used to guide different testers to make the feature values of specific data features fall within the corresponding value ranges when conducting tests, which can reduce the influence of the operation habits of the testers on the test results and improve the consistency of the test results when different testers conduct the same test. Description of the Drawings

[0040] Figure 1 It is a flowchart showing the test standard generation method according to an embodiment of the present invention;

[0041] Figure 2 It is a data feature curve graph when tester A conducts a test according to an embodiment of the present invention;

[0042] Figure 3 It is a data feature curve graph when tester B conducts a test according to an embodiment of the present invention;

[0043] Figure 4 It is a structural schematic diagram of a test standard generation device according to another embodiment of the present invention. Detailed Embodiments

[0044] To make the above objects, features, and advantages of the present invention more apparent and understandable, the following provides a detailed description of specific embodiments of the present invention with reference to the accompanying drawings. Although some embodiments of the present invention are shown in the drawings, it should be understood that the present invention can be implemented in various forms and should not be construed as limited to the embodiments set forth herein. Instead, these embodiments are provided to more thoroughly and completely understand the present invention. It should be understood that the drawings and embodiments of the present invention are only for exemplary purposes and are not used to limit the protection scope of the present invention.

[0045] It should be understood that the various steps recorded in the method embodiments of the present invention can be executed in different orders and / or executed in parallel. In addition, the method embodiments may include additional steps and / or omit the steps shown. The scope of the present invention is not limited in this regard.

[0046] As used herein, the term "including" and its variants are open-ended, that is, "including but not limited to". The term "based on" means "at least partially based on". The term "one embodiment" means "at least one embodiment"; the term "another embodiment" means "at least one additional embodiment"; the term "some embodiments" means "at least some embodiments"; the term "optionally" means "optional embodiments". The relevant definitions of other terms will be given in the following description. It should be noted that the concepts such as "first" and "second" mentioned in the present invention are only used to distinguish different devices, modules, or units, and are not used to limit the order of functions performed by these devices, modules, or units or their interdependent relationships.

[0047] It should be noted that the modifications of "one" and "multiple" mentioned in the present invention are illustrative rather than restrictive. Those skilled in the art should understand that, unless otherwise clearly indicated in the context, it should be understood as "one or more".

[0048] The names of the messages or information exchanged between multiple devices in the embodiments of the present invention are only for illustrative purposes and are not used to limit the scope of these messages or information.

[0049] In the prior art, multiple testers often conduct tests on the equipment separately, and then analyze all the test results. The test result with the highest test efficiency or the lowest test fuel consumption among all the test results obtained by each tester is determined as the final test result. However, for each piece of equipment, multiple testers are required to conduct tests, resulting in low test efficiency and high test costs.

[0050] As Figure 1 shown, a test standard generation method provided by the present invention includes:

[0051] Step S100, obtain the working condition data combinations of multiple testers during the test. The working condition data combinations include at least one data feature and the eigenvalue time series corresponding to each data feature.

[0052] Specifically, when each tester conducts the test, they respectively record the eigenvalues of each data feature during the test process, and arrange the eigenvalues of the same data feature in chronological order during the test process to obtain the eigenvalue time series of this data feature when this tester conducts the test. When obtaining the working condition data combinations of multiple testers during the test, the test data obtained by each tester during the test can also be obtained. The test data includes the test duration and / or the test fuel consumption, and one test data corresponds to the working condition data combination obtained by one tester during the test.

[0053] It should be noted that in this embodiment, the test standard generation method is specifically described by taking the crane test as an example, but the test standard generation method of the present invention is not limited to being applied in the crane test, and can also be applied to generate test standards for other equipment.

[0054] Taking the crane test as an example, the data features include at least one of the slewing angle, the crane boom length, the crane speed, the torque percentage, the opening and closing amplitudes of the handle on the X and Y axes, the throttle pedal amplitude, etc. The values of these features can be obtained by statistically analyzing the data transmitted back by the vehicle-end T-BOX; the data features can also include the failure rate, etc., and the value of this feature can be obtained through the test monitoring platform.

[0055] Exemplarily, for tester A, assuming that the data features include data feature 1, data feature 2, and data feature 3, respectively record the eigenvalues of data feature 1, data feature 2, and data feature 3 at each time point during the crane test by tester A, and arrange all the eigenvalues of data feature 1 in chronological order to obtain the eigenvalue time series of data feature 1 when tester A conducts the crane test; based on the same processing process, the eigenvalue time series of data feature 2 and data feature 3 when tester A conducts the crane test can be obtained. Display the eigenvalue time series of data feature 1, data feature 2, and data feature 3 of tester A on the same graph, and the data feature curve graph as shown in Figure 2 can be obtained.

[0056] Based on the same processing process, the eigenvalue time series of data feature 1, data feature 2, and data feature 3 when tester B conducts the crane test can be obtained respectively. Display the eigenvalue time series of data feature 1, data feature 2, and data feature 3 of tester B on the same graph, and the data feature curve graph as shown in Figure 3 can be obtained.

[0057] Step S200, determine a specific combination of operating conditions data that meets the preset evaluation criteria among all the combinations of operating conditions data, input all the data features in the specific combination of operating conditions data into the trained machine learning model, and determine the specific data features related to the operator's operating habits.

[0058] Specifically, the preset evaluation criteria can be specifically set according to actual needs. For example, the preset evaluation criteria can be set as the N combinations of operating conditions data with the minimum test fuel consumption, or the M combinations of operating conditions data with the shortest test duration, etc., which are not limited here.

[0059] The machine learning model is used to classify the input data features, and machine learning models that can implement classification functions such as decision tree models, random forest models, naive Bayes models, and support vector machines can be used.

[0060] It can be understood that a specific combination of operating conditions data includes multiple data features. Among them, some data features may be related to the operator's operating habits, and some data features may be unrelated to the operator's operating habits. The trained machine learning model is used to screen out the specific data features related to the operator's operating habits.

[0061] Taking the crane test as an example, the specific data features related to the operator's operating habits include the opening and closing amplitude of the handle, the amplitude of the accelerator pedal, the torque percentage, the speed, the boom length, the slewing angle, etc.

[0062] Step S300, determine the value range of the specific data feature according to the time series of the feature values corresponding to the specific data feature in the specific combination of operating conditions data, and use the specific data feature and its value range as the test standard.

[0063] Specifically, the operation standard includes a recommended operation standard and a non-recommended operation standard. Using the specific data feature and its value range as the test standard means that every time a test is carried out, it is recommended that the feature value of the specific data feature be within the value range corresponding to the recommended operation standard, and it is not recommended that the feature value of the specific data feature be within the value range corresponding to the non-recommended operation standard.

[0064] For a specific data feature, the maximum value and the minimum value of the specific data feature can be determined according to the time series of the feature values. If the number of specific combinations of operating conditions data is multiple, then according to the time series of the corresponding feature values in each specific combination of operating conditions data, the local maximum value and the local minimum value of the specific data feature in each specific combination of operating conditions data are determined respectively, and then the local maximum values and the local minimum values corresponding to each specific combination of operating conditions data are compared to determine the global maximum value and the global minimum value of the specific data feature. The numerical range between the global minimum value and the global maximum value is the value range of the specific data feature.

[0065] Alternatively, for a specific data feature, based on the time series of the feature values, the values of the feature data feature at each time point can be determined, and based on these values, the value combinations of the data feature can be determined. If the number of specific working condition data combinations is multiple, then the value combinations of the specific data feature corresponding to each working condition data combination are merged to obtain the final value combination of the specific data feature, which serves as the value range of the specific data feature.

[0066] In this embodiment, the working condition data combinations obtained by multiple testers conducting tests are acquired. Each tester corresponds to a working condition data combination, and the working condition data combination includes at least one data feature and its time series of feature values. Specific working condition data combinations that meet the preset evaluation criteria are determined from all the working condition data combinations. The preset evaluation criteria can be specifically set according to the optimization objectives of the test standard. For example, the working condition data combination with the highest test efficiency. The specific data features related to the operating habits of the testers are screened out from all the data features through the trained machine learning model. The value range of the specific data feature is determined based on the feature values at different time points in the specific working condition data combinations that meet the preset evaluation criteria. Using the specific data feature and its value range as the test standard, when this test standard guides the testers to conduct tests, it can make the test results close to the preset evaluation criteria. The test standard generated in this embodiment is used to guide different testers to make the feature values of the specific data feature fall within the corresponding value ranges during the test, which can reduce the influence of the operating habits of the testers on the test results and improve the consistency of the test results when different testers conduct the same test.

[0067] Exemplarily, compared with the prior art, for equipment of the same model, the present invention only requires multiple testers to conduct test operations when generating the operation standard. After the operation standard is generated, the operation standard can guide any tester to complete a test that meets the preset evaluation criteria. When testing equipment of the same model, only one tester is required to operate, effectively improving the test efficiency and reducing the test cost.

[0068] Optionally, the method for generating the test standard further includes: acquiring the test data of each of the testers conducting the test, where the test data includes the test duration and / or the test fuel consumption.

[0069] The specific working condition data combination includes a recommended data combination and / or a non-recommended data combination. Determining the specific working condition data combination that meets the preset evaluation criteria from all the working condition data combinations includes:

[0070] Determining the working condition data combinations with the test duration less than or equal to the first preset duration threshold and / or the test fuel consumption less than or equal to the first preset fuel consumption threshold among all the working condition data combinations as the recommended data combinations;

[0071] And / or,

[0072] Among all the combinations of the working condition data, determine that the combination of the working condition data with the test duration greater than or equal to the second preset duration threshold and / or the test fuel consumption greater than or equal to the second preset fuel consumption threshold is the non-recommended data combination.

[0073] Specifically, the first preset duration threshold, the first preset fuel consumption threshold, the second preset duration threshold, and the second preset fuel consumption threshold can be specifically set according to the actual situation and are not limited here. For a combination of working condition data, if the corresponding test duration is less than or equal to the first preset duration, it means that the tester corresponding to this combination of working condition data has higher efficiency during the test; on the contrary, if the corresponding test duration is greater than or equal to the second preset duration threshold, and the second preset duration threshold can be greater than or equal to the first preset duration threshold, it means that the tester corresponding to this combination of working condition data has lower efficiency during the test.

[0074] For a combination of working condition data, if the corresponding test fuel consumption is less than or equal to the first preset fuel consumption threshold, it means that the tester corresponding to this combination of working condition data has lower energy consumption and is more energy-efficient during the test; on the contrary, if the corresponding test fuel consumption is greater than or equal to the second preset fuel consumption threshold, and the second preset fuel consumption threshold can be greater than or equal to the first preset fuel consumption threshold, it means that the tester corresponding to this combination of working condition data has higher energy consumption during the test.

[0075] Optionally, the test standard generation method further includes: obtaining the test data of each of the testers during the test, where the test data includes the test duration and / or the test fuel consumption.

[0076] The specific combination of working condition data includes the recommended data combination and / or the non-recommended data combination. Determining the specific combination of working condition data that meets the preset evaluation criteria among all the combinations of working condition data includes:

[0077] Select the first preset number of the combinations of working condition data in ascending order of the test duration as the recommended data combination, and / or select the second preset number of the combinations of working condition data in ascending order of the test fuel consumption as the recommended data combination;

[0078] and / or,

[0079] Select the third preset number of the combinations of working condition data in descending order of the test duration as the non-recommended data combination, and / or select the fourth preset number of the combinations of working condition data in descending order of the test fuel consumption as the non-recommended data combination.

[0080] Specifically, the first preset quantity, the second preset quantity, the third preset quantity, and the fourth preset quantity can be specifically set according to the actual situation and are not limited herein. Sort the combinations of working condition data in ascending order of the test duration. For the combinations of working condition data in the top N positions in terms of the test duration, it indicates that the corresponding test operator has higher efficiency during the test; conversely, for the combinations of working condition data in the bottom M positions in terms of the test duration, it indicates that the corresponding test operator has lower efficiency during the test, where N and M are greater than or equal to 1.

[0081] For a combination of working condition data, sort the combinations of working condition data in ascending order of the test fuel consumption. For the combinations of working condition data in the top P positions in terms of the test fuel consumption, it indicates that the corresponding test operator has lower energy consumption and is more energy-efficient during the test; conversely, for the combinations of working condition data in the bottom Q positions in terms of the test fuel consumption, it indicates that the corresponding test operator has higher energy consumption during the test, where P and Q are greater than or equal to 1.

[0082] In this alternative embodiment, the recommended value range of a specific data feature can be determined based on the time series of feature values corresponding to the specific data feature in the recommended data combination as the recommended standard for the test; the non-recommended value range of the specific data feature can be determined based on the time series of feature values corresponding to the specific data feature in the non-recommended data combination as the non-recommended standard for the test. By mining the recommended value range and non-recommended value range of the data features related to the test operator's operation habits from the combinations of working condition data of a large number of test operators for generating the test standard, it is possible to reduce the influence of different test operators' operation habits on the test efficiency and test fuel consumption and improve the consistency of test results when different test operators conduct the same test.

[0083] Optionally, the quantity of the specific combinations of working condition data and the quantity of the specific data features are both at least one. After determining the specific data features related to the test operator's operation habits, it further includes:

[0084] Calculate the coincidence degree between the time series of feature values of the same specific data feature in each specific combination of working condition data.

[0085] Exemplarily, assume that data feature 1 is a specific data feature, and combination of working condition data 1, combination of working condition data 2, and combination of working condition data 3 are all specific combinations of working condition data. Then calculate the coincidence degree between the time series of feature values of data feature 1 in combination of working condition data 1, the time series of feature values of data feature 1 in combination of working condition data 2, and the time series of feature values of data feature 1 in combination of working condition data 3.

[0086] Specifically, trajectory similarity algorithms such as the DTW algorithm and the LCSS algorithm can be used to calculate the coincidence degree between different time series of feature values. The specific calculation process is prior art and will not be elaborated herein.

[0087] The overlap degree of each of the specific data features is compared with a preset threshold value, and the specific data features whose overlap degree is greater than or equal to the preset threshold value are retained according to the comparison result.

[0088] Specifically, for a specific data feature, when its overlap is greater than or equal to a preset threshold, it means that the specific data feature is highly consistent with the operating habits of different experimenters, and the specific data feature is retained; when its overlap is less than the preset threshold, it means that the characteristic value volatility of the specific data feature is high and may be greatly affected by other factors, and the specific data feature is discarded.

[0089] It can be understood that the specific data features related to the operating habits of the tester in the recommended data combination and the non-recommended data combination may be the same or different.

[0090] In this optional embodiment, by calculating the overlap between the time series of each characteristic value of a specific data feature, the overlap can reflect whether the specific data feature is mainly affected by the operating habits of the tester. The higher the overlap, the more affected the specific data feature is by the operating habits of the tester. Retaining specific data features with a higher overlap means mining out specific data features with a higher correlation with the operating habits of the tester. Generating operating standards based on the retained specific data features can reduce the impact of different testers' operating habits on test results such as test efficiency and test fuel consumption.

[0091] Optionally, the calculating the overlap between the feature value time series of the same specific data feature in each of the specific operating condition data combinations includes:

[0092] For any of the specific data features, the first variance of the feature value time series of the specific data feature in each of the specific operating condition data combinations is determined respectively.

[0093] Exemplarily, for data feature 1, the first variance of the feature value time series corresponding to data feature 1 for experimenter A is determined, the first variance of the feature value time series corresponding to data feature 1 for experimenter B is determined, and so on.

[0094] A second variance of all the first variances corresponding to the specific data feature is calculated, where the second variance is used to characterize the degree of overlap corresponding to the specific data feature.

[0095] Specifically, for a specific data feature, using all the first variances of the specific data feature as data samples, calculate the second variance, which is used to characterize the coincidence degree of the specific data feature. Among them, the larger the second variance, the lower the coincidence degree; the smaller the second variance, the higher the coincidence degree. The coincidence degree can be set as the reciprocal of the second variance for screening specific data features, or the second variance can be compared with a preset threshold, and the specific data features with the second variance less than or equal to the preset threshold are retained.

[0096] In this optional embodiment, the first variance reflects the degree of dispersion of the time series of each eigenvalue of the specific data feature, and the second variance reflects the degree of dispersion of all the first variances of the specific data feature. The larger the second variance, the higher the degree of dispersion of all the first variances of the specific data feature, that is, some first variances are large and some first variances are small. That is to say, the degree of dispersion of the time series of some eigenvalues corresponding to the specific data feature is high, and the degree of dispersion of the time series of some eigenvalues is low, and the coincidence degree between the corresponding time series of each eigenvalue is obviously low; on the contrary, the smaller the second variance, the lower the degree of dispersion of all the first variances of the specific data feature, and the degree of dispersion of each eigenvalue time series is similar, indicating that the coincidence degree between the time series of each eigenvalue is high. By calculating the first variance of each eigenvalue time series and judging the coincidence degree corresponding to the specific data feature according to the second variance of all the first variances, the calculation process is simple, the calculation amount is small, and the calculation efficiency is effectively improved.

[0097] Optionally, the determining the value range of the specific data feature according to the time series of eigenvalues corresponding to the specific data feature in the specific working condition data combination includes:

[0098] Determine the confidence interval of the specific data feature according to all the first variances of the specific data feature.

[0099] Specifically, the average value of all the first variances of the specific data feature can be calculated, and the confidence interval of the specific data feature is determined according to the calculated average variance. Among them, the process of determining the confidence interval according to the variance is a prior art and will not be elaborated here.

[0100] Determine the value range of the specific data feature according to the interval values of the confidence interval.

[0101] Specifically, the value interval of the specific data feature can be determined according to the maximum interval value and the minimum interval value in the confidence interval as its value range; or, all the interval values within the confidence interval can be used as the value combination, and this value combination is used as the value range of the specific data feature.

[0102] For the recommended data combinations, the confidence intervals of specific data features are determined according to the first variances corresponding to the respective recommended data combinations, and the recommended value ranges of the specific data features are determined according to the interval values of the confidence intervals.

[0103] For the non-recommended data combinations, the confidence intervals of specific data features are determined according to the first variances corresponding to the respective non-recommended data combinations, and the non-recommended value ranges of the specific data features are determined according to the interval values of the confidence intervals.

[0104] In this optional embodiment, by determining the confidence intervals of specific data features, determining the value ranges of specific data features according to the interval values of the confidence intervals, and removing the feature values outside the confidence intervals, that is, the values with poor confidence, the accuracy of the generated operation standards can be improved, and the influence of the operation habits of the testers on the test results such as test efficiency and test fuel consumption can be effectively reduced.

[0105] Optionally, before inputting all the data features in the specific working condition data combination into the trained machine learning model, it further includes:

[0106] Obtaining a plurality of different labeled data features, where the labels include those related to the operation habits of the testers and those unrelated to the operation habits of the testers;

[0107] Training a pre-established machine learning model with all the labeled data features to obtain the trained machine learning model.

[0108] Specifically, a plurality of data features can be obtained in advance, the data features related to the operation habits of the testers are manually labeled, and then the labeled data features are used to train the machine learning model to obtain the trained machine learning model. The specific training process of the machine learning model is a prior art and will not be elaborated here.

[0109] As Figure 4 shown, the present invention provides a test standard generation device, including:

[0110] An acquisition module, configured to acquire the working condition data combinations of multiple testers during the test, where the working condition data combinations include at least one data feature and the feature value time series corresponding to each data feature;

[0111] A processing module, configured to determine specific working condition data combinations that meet the preset evaluation criteria among all the working condition data combinations, input all the data features in the specific working condition data combinations into the trained machine learning model, and determine specific data features related to the operation habits of the testers;

[0112] A generation module, configured to determine a value range of the specific data feature according to the eigenvalue time series corresponding to the specific data feature in the specific working condition data combination, where the specific data feature and its value range are used as a test standard.

[0113] The test standard generation device of this embodiment is used to implement the test standard generation method as described above, and its beneficial effects correspond to those of the test standard generation method as described above, and will not be elaborated here.

[0114] In a third aspect, the present invention provides a computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, the test standard generation method as described in any item of the first aspect is implemented.

[0115] In a fourth aspect, the present invention provides a crane test monitoring platform, including a memory and a processor;

[0116] The memory is used to store a computer program;

[0117] The processor is configured to implement the test standard generation method as described above when executing the computer program.

[0118] Those of ordinary skill in the art can understand that all or part of the processes in the methods of the above embodiments can be completed by instructing relevant hardware through a computer program. The program can be stored in a computer-readable storage medium. When the program is executed, it can include the processes of the embodiments of the above methods. Among them, the storage medium can be a magnetic disk, an optical disk, a read-only memory (ROM), or a random access memory (RAM), etc. In this application, the units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they can be located in one place, or they can be distributed to multiple network units. Part or all of the units can be selected according to actual needs to achieve the purpose of the solution of the embodiments of the present invention. In addition, the functional units in each embodiment of the present invention can be integrated into one processing unit, or each unit can exist physically alone, or two or more units can be integrated into one unit. The above integrated units can be implemented in the form of hardware or in the form of software functional units.

[0119] Although the present invention is disclosed as above, the protection scope of the present invention is not limited thereto. Those skilled in the art can make various changes and modifications without departing from the spirit and scope of the present invention, and these changes and modifications will all fall within the protection scope of the present invention.

Claims

1. A method for generating a test standard, characterized in that, Including: Obtain the working condition data combinations of multiple testers during the test. The working condition data combinations include at least one data feature and the eigenvalue time series corresponding to each data feature. At least some of the at least one data feature are related to the operation habits of the testers; Determine specific working condition data combinations that meet the preset evaluation criteria among all the working condition data combinations. Input all the data features in the specific working condition data combinations into a trained machine learning model to determine specific data features related to the operation habits of the testers. The trained machine learning model is trained with multiple labeled data features, and the labels include related to the operation habits of the testers and not related to the operation habits of the testers; Determine the value range of the specific data feature according to the eigenvalue time series corresponding to the specific data feature in the specific working condition data combination. The specific data feature and its value range are used as the test standard.

2. The test standard generation method according to claim 1, characterized in that The number of the specific working condition data combinations and the number of the specific data features are both at least one. After determining the specific data features related to the operation habits of the testers, it further includes: Calculate the coincidence degree between the eigenvalue time series of the same specific data feature in each specific working condition data combination; Compare the coincidence degree of each specific data feature with a preset threshold respectively, and retain the specific data features whose coincidence degree is greater than or equal to the preset threshold according to the comparison result.

3. The test standard generation method according to claim 2, wherein The calculating the coincidence degree between the eigenvalue time series of the same specific data feature in each specific working condition data combination includes: For any specific data feature, respectively determine the first variance of the eigenvalue time series of the specific data feature in each specific working condition data combination; Calculate the second variance of all the first variances corresponding to the specific data feature. The second variance is used to characterize the coincidence degree corresponding to the specific data feature.

4. The test standard generation method according to claim 3, characterized in that The determining the value range of the specific data feature according to the eigenvalue time series corresponding to the specific data feature in the specific working condition data combination includes: Determine the confidence interval of the specific data feature according to all the first variances of the specific data feature; Determine the value range of the specific data feature according to the interval values of the confidence interval.

5. The method for generating a test standard according to any one of claims 1 to 4, characterized in that, It further includes: Obtain the test data of each tester during the test. The test data includes the test duration and / or the test fuel consumption; The specific working condition data combination includes a recommended data combination and / or a non-recommended data combination. The determining the specific working condition data combination that meets the preset evaluation criteria among all the working condition data combinations includes: Determine that the working condition data combination with the test duration less than or equal to the first preset duration threshold and / or the test fuel consumption less than or equal to the first preset fuel consumption threshold among all the working condition data combinations is the recommended data combination; And / or, Among all the combinations of the working condition data, determine that the combination of the working condition data with the test duration greater than or equal to the second preset duration threshold and / or the test fuel consumption greater than or equal to the second preset fuel consumption threshold is the non-recommended data combination.

6. The method for generating a test standard according to any one of claims 1 to 4, characterized in that It further includes: Obtain the test data of each test operator, where the test data includes the test duration and / or the test fuel consumption; The specific combination of the working condition data includes the recommended data combination and / or the non-recommended data combination. Determining the specific combination of the working condition data that meets the preset evaluation criteria among all the combinations of the working condition data includes: Select the first preset number of the combinations of the working condition data in ascending order of the test duration as the recommended data combination, and / or select the second preset number of the combinations of the working condition data in ascending order of the test fuel consumption as the recommended data combination; and / or Select the third preset number of the combinations of the working condition data in descending order of the test duration as the non-recommended data combination, and / or select the fourth preset number of the combinations of the working condition data in descending order of the test fuel consumption as the non-recommended data combination.

7. The method for generating a test standard according to any one of claims 1 to 4, characterized in that, Before inputting all the data features in the specific combination of the working condition data into the trained machine learning model, it further includes: Obtain a plurality of different labeled data features, where the labels include those related to the operation habits of the test operator and those not related to the operation habits of the test operator; Train the pre-established machine learning model with all the labeled data features to obtain the trained machine learning model.

8. An experimental standard generation device, characterized in that, It includes: An acquisition module, configured to acquire the combinations of the working condition data of multiple test operators. The combination of the working condition data includes at least one data feature and the time series of the feature values corresponding to each data feature. At least some of the at least one data feature are related to the operation habits of the test operator; A processing module, configured to determine the specific combination of the working condition data that meets the preset evaluation criteria among all the combinations of the working condition data, input all the data features in the specific combination of the working condition data into the trained machine learning model, and determine the specific data features related to the operation habits of the test operator. The trained machine learning model is trained with a plurality of labeled data features, and the labels include those related to the operation habits of the test operator and those not related to the operation habits of the test operator; A generation module, configured to determine the value range of the specific data feature according to the time series of the feature values corresponding to the specific data feature in the specific combination of the working condition data. The specific data feature and its value range are used as the test standard.

9. A computer-readable storage medium, characterized in that, A computer program is stored on the storage medium. When the computer program is executed by a processor, the test standard generation method according to any one of claims 1 to 7 is implemented.

10. A crane test monitoring platform, characterized in that, It includes a memory and a processor; The memory is used to store the computer program; The processor is configured to implement the test standard generation method according to any one of claims 1 to 7 when executing the computer program.

Images