Intelligent control method of skid-mounted detection rack for engine performance test

Through big data analysis and working condition enhancement technology, an enhanced test working condition series is generated, which solves the problems of high engine performance testing costs and long cycles, and realizes efficient test control and data acquisition.

CN119086072AActive Publication Date: 2024-12-06NANTONG LIDA ENVIRONMENTAL PROTECTION EQUIP CO LTD
View PDF 6 Cites 0 Cited by

Patent Information

Application Number
CN202411588550.X
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-11-08
Publication Date
2024-12-06
Estimated Expiration
2044-11-08

AI Technical Summary

Technical Problem

Existing engine performance testing costs are high and the test cycle is long, making it difficult to effectively shorten.

Method used

The operating condition data set of the target engine is extracted through big data, and the calibration operating condition set is determined using statistical analysis methods, a working condition enhancement channel is constructed, an enhanced test operating condition series is generated, and synchronized to the detection bench for testing control to obtain the test data set.

Benefits of technology

It shortens the test cycle, reduces the test cost, and improves the test efficiency and effect.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN119086072A_ABST
    Figure CN119086072A_ABST
Patent Text Reader

Abstract

The invention discloses a skid-mounted detection rack intelligent control method for engine performance testing, and relates to the technical field of engine testing, and the method comprises the steps: extracting a working condition data set of a target engine based on big data, and determining a calibration working condition set through a statistical analysis method; acquiring a test demand scene, and setting a standard test condition series in combination with the calibration condition set; constructing a working condition enhancement channel, taking the standard test working condition series and the test scene features as input, activating the working condition enhancement channel to perform working condition enhancement, and generating an enhanced test working condition series; and synchronously enhancing the test condition series to the target detection rack for test control, and obtaining a test data set. The technical problems of high test cost and long test period in the prior art are solved, and the technical effects of shortening the test period and reducing the test cost are achieved.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The invention relates to the technical field of engine testing, and in particular to an intelligent control method of a skid-mounted testing bench for engine performance testing. Background Art

[0002] Engine performance testing is a key link in the process of automobile manufacturing and R&D, and is designed to evaluate the performance and reliability of the engine under various working conditions. Current engine performance testing mainly relies on standard test procedures. Standard tests include steady-state operating condition tests and dynamic operating condition tests, which are used to evaluate the stability and response performance of the engine respectively. These tests are usually time-consuming and costly. There are technical problems such as high testing costs and long testing cycles. Summary of the invention

[0003] The present invention provides an intelligent control method for a skid-mounted test bench for engine performance testing, so as to solve the technical problems of high testing cost and long testing cycle in the prior art, and achieve the technical effect of shortening the testing cycle and reducing the testing cost.

[0004] The present invention provides a skid-mounted test bench intelligent control method for engine performance testing, wherein the method comprises: Based on big data, the operating condition data set of the target engine is extracted, and the calibration operating condition set is determined through statistical analysis methods; the test demand scenario is obtained, and the standard test condition series is set in combination with the calibration operating condition set; a condition enhancement channel is constructed, and the standard test condition series and the test scenario characteristics are used as input to activate the condition enhancement channel for condition enhancement and generate an enhanced test condition series; the enhanced test condition series is synchronized to the target detection bench for test control, and the test data set is obtained.

[0005] In a feasible implementation, based on big data, the operating condition data set of the target engine is extracted, and a calibration operating condition set is determined through a statistical analysis method, including: Constructing a working condition mapping space, wherein a plurality of coordinate axes of the working condition mapping space correspond to a plurality of working condition parameters of a target engine; parsing the working condition data set and mapping a plurality of working condition data into a plurality of working condition scatter points in the working condition mapping space to form a working condition scatter point density map; constructing a digital hill based on the working condition scatter point density map, wherein the digital hill includes a plurality of digital peaks; simulating watering of the digital hill, and performing adaptive working condition division to determine a plurality of calibration working conditions, which are stored as the calibration working condition set.

[0006] In a feasible implementation, a digital hill is constructed based on the operating condition scatter density map, wherein the digital hill includes a plurality of digital peaks, including: Based on the working condition scatter point density map, the probability density of multiple scattered points is calculated, and the working condition probability density map is obtained by fitting; according to the working condition probability density map, the digital hill is formed by taking the probability density value as the digital surface altitude and the working condition parameter as the digital surface plane coordinate.

[0007] In a feasible implementation, the digital hill is simulated for irrigation, and adaptive working condition division is performed to determine a plurality of calibration working conditions, which are stored as the calibration working condition set, including: The digital surface altitude zero point of the digital hill is used as the starting point for simulated watering to raise the water level; when the water level passes through the first digital peak, it is used as the first calibration center; the water level continues to be raised, and the digital peaks that have passed through are recorded respectively, and the digital peaks are merged with the first calibration center to obtain a calibration center set; the calibration center set is used as a cluster center set, a plurality of the operating condition scatter points are clustered and optimized, and the cluster center set is updated, and the updated cluster center set is output as the calibration operating condition set.

[0008] In a feasible implementation, a working condition enhancement channel is constructed, and the standard test working condition series and the test scenario characteristics are used as input, and the working condition enhancement channel is activated to perform working condition enhancement, and an enhanced test working condition series is generated, which includes: Based on the calibration working condition set, a reference working condition is determined; in combination with the test target of the test requirement scenario, the calibration working condition set is calibrated with an equivalent coefficient, and stored as an equivalent mapping matrix.

[0009] In a feasible implementation, a working condition enhancement channel is constructed, and the standard test working condition series and the test scenario characteristics are used as input, and the working condition enhancement channel is activated to perform working condition enhancement, and an enhanced test working condition series is generated, including: Based on the calibration condition set, a calibration condition space is generated with the proportion of multiple calibration conditions as the multi-dimensional coordinate axis; in combination with the test scenario characteristics, random selection is performed in the calibration condition space to construct and initialize a swarm of candidate condition particles; an enhanced evaluation function is constructed, and the enhanced evaluation function is combined with the equivalent mapping matrix to iteratively optimize the swarm of candidate condition particles to obtain the optimal candidate condition particles, which are output as the enhanced test condition series.

[0010] In a feasible implementation, constructing a reinforcement evaluation function also includes: The enhanced evaluation function expression is as follows: ; in, Design time for the series of standard test conditions; , Respectively characterize the lowest equivalent coefficient value of the particles of the alternative working condition and the standard test working condition series; Characterize the proportion of working conditions corresponding to the working condition with the lowest value of the intermediate generation coefficient of particles in the alternative working conditions; Characterizes the proportion of working conditions corresponding to the working condition with the lowest value of the generation coefficient in the series of standard test conditions; , Respectively represent the types of working conditions in the calibration working condition set and the standard test working condition series; n represents the types of working conditions in the alternative working condition particles; , The test energy consumption corresponding to the condition with the lowest generation coefficient value in the alternative working condition particles and the test energy consumption corresponding to the condition with the lowest generation coefficient value in the standard test working condition series are respectively characterized.

[0011] In a feasible implementation, the method further includes: A control output series of the test bench is extracted based on the test results; the control output series is compared with the enhanced test condition series to obtain condition control differences; and control feedback optimization of the test bench is performed based on the condition control differences.

[0012] The present invention discloses an intelligent control method for a skid-mounted test bench for engine performance testing, including: extracting a target engine's operating condition data set based on big data, and determining a calibration operating condition set through a statistical analysis method; obtaining a test requirement scenario, and setting a standard test operating condition series in combination with the calibration operating condition set; constructing an operating condition enhancement channel, using the standard test operating condition series and test scenario features as input, activating the operating condition enhancement channel for operating condition enhancement, and generating an enhanced test operating condition series; synchronizing the enhanced test operating condition series to the target test bench for test control, and obtaining a test data set. The intelligent control method for a skid-mounted test bench for engine performance testing disclosed by the present invention solves the technical problems of high test cost and long test cycle, and achieves the technical effect of shortening the test cycle and reducing test cost. BRIEF DESCRIPTION OF THE DRAWINGS

[0013] Figure 1 It is a schematic flow chart of the intelligent control method of the skid-mounted test bench for engine performance testing of the present invention; Figure 2 The present invention is a schematic diagram of a flow chart for determining a calibration condition set in an intelligent control method for a skid-mounted test bench for engine performance testing. DETAILED DESCRIPTION

[0014] The technical solution provided in the embodiments of the present invention is to solve the technical problems of high testing cost and long testing cycle in the prior art. The overall idea adopted is as follows: Firstly, based on big data, the operating condition data set of the target engine is extracted, and the calibration operating condition set is determined through statistical analysis methods; then, the test demand scenario is obtained, and the standard test condition series is set in combination with the calibration operating condition set; then, a condition enhancement channel is constructed, and the standard test condition series and the test scenario characteristics are used as input, the condition enhancement channel is activated to perform condition enhancement, and an enhanced test condition series is generated; finally, the enhanced test condition series is synchronized to the target detection bench for test control, and the test data set is obtained.

[0015] The above technical solution will be described in detail below in conjunction with the accompanying drawings and specific implementation methods of the specification to better understand the above technical solution. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all of the embodiments of the present invention. It should be understood that the present invention is not limited to the example embodiments used only to explain the present invention. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present invention. In addition, it should be noted that, for the convenience of description, only the parts related to the present invention are shown in the drawings, rather than all of them.

[0016] Example Figure 1 The figure is a flow chart of an intelligent control method of a skid-mounted test bench for engine performance testing according to the present invention, wherein the method comprises: Based on big data, the operating condition data set of the target engine is extracted, and the calibration operating condition set is determined through statistical analysis methods.

[0017] Specifically, the operating condition data set of the target engine acquired by the interactive big data includes multiple operating condition data of the target engine and engines with the same purpose as the target engine. The operating condition data set reflects all working scenarios of the planned operation of the target engine, exemplarily including starting conditions, idling conditions, acceleration conditions, constant speed conditions, deceleration conditions, etc., and different working conditions have multiple different load conditions.

[0018] Among them, when the target engine is a new product or is not widely used, an engine with the same purpose as the target engine, that is, the same source model of the target engine, is selected to obtain the operating condition data set. The engine with the same purpose or source model has similar design goals or application scenarios as the target engine, and the obtained operating condition data set can also reflect all working scenarios of the target engine's planned work.

[0019] Specifically, the calibration condition set is a collection of a series of operating point data, which includes all important working scenarios of the target engine and is the basis for selecting test points for subsequent performance tests. Through this calibration condition set, the performance characteristics of the target engine can be better reflected, thereby improving the confidence and representativeness of subsequent performance tests.

[0020] In some embodiments, Figure 2 As shown, based on big data, the operating condition data set of the target engine is extracted, and the calibration operating condition set is determined through statistical analysis methods, including: A working condition mapping space is constructed, wherein a plurality of coordinate axes of the working condition mapping space correspond to a plurality of working condition parameters of the target engine.

[0021] The operating condition data set is parsed and a plurality of operating condition data are mapped into a plurality of operating condition scatter points in the operating condition mapping space to form an operating condition scatter point density map.

[0022] A digital hill is constructed based on the operating condition scatter density map, wherein the digital hill includes a plurality of digital peaks.

[0023] The digital hill is watered by simulation, and adaptive working condition division is performed to determine a plurality of calibration working conditions, which are stored as the calibration working condition set.

[0024] Specifically, based on the operating parameters of the target engine, an operating condition mapping space is constructed, including using multiple operating parameters as multiple coordinate axes, for example, using the speed of the target engine as the first coordinate axis and the torque of the target engine as the second coordinate axis, to construct an orthogonal two-dimensional plane coordinate system to obtain the operating condition mapping space.

[0025] Furthermore, the operating condition data set is cleaned and standardized to remove outliers and noise to ensure data consistency and accuracy. Exemplarily, this includes removing operating condition data that clearly exceeds the usage scenario defined for the target engine. Next, multiple operating condition parameters in each group of operating condition samples are extracted from the processed operating condition data set, and each group of operating condition samples is mapped to the operating condition mapping space based on the multiple groups of operating condition parameters obtained, multiple operating condition scatter points are obtained, and an operating condition scatter point density map is formed.

[0026] Specifically, the density analysis of the operating condition scatter density map is performed to identify high-density areas (digital peaks) and low-density areas (valleys), and then a digital hill model is constructed to show the distribution of different density areas. The digital hill model is used to intuitively reflect the operating condition probability density distribution of the target engine, thereby helping to determine which operating condition points are highly representative. Among them, each digital peak represents a specific operating condition area or state, and the highest point of each digital peak is the possible typical operating condition of the operating condition area or state.

[0027] Optionally, based on visualization tools such as Matplotlib and Plotly, digital peaks and density maps are combined to generate three-dimensional visualization charts to obtain digital hills. Then, multiple digital peaks in the digital hills are labeled to indicate the working conditions and important parameters corresponding to the peaks.

[0028] Specifically, by simulating the process of water gradually pouring into the digital hill from low to high, different working condition areas are adaptively divided according to the water level changes during the water filling process. For example, first, starting from the lowest point of the digital hill, the water level is gradually increased, the submerged areas under each water level are recorded, and the adaptive working condition division boundaries and the working conditions corresponding to these areas are determined. Then, based on the working condition area identification results, representative working conditions are selected as calibration working conditions. The characteristic data of the calibration working conditions (such as torque, speed, etc.) are stored in the database to form a calibration working condition set for easy management and maintenance.

[0029] Through digital hills and simulated watering technology, the distribution of operating condition data is more intuitive and easier to understand, thereby more accurately identifying and dividing operating areas, providing reliable basic data support for subsequent performance tests.

[0030] In some implementations, a digital hill is constructed based on the operating condition scatter density map, wherein the digital hill includes a plurality of digital peaks, including: Based on the operating condition scatter point density map, the probability density of multiple scatter points is calculated, and the operating condition probability density map is obtained by fitting.

[0031] According to the operating condition probability density map, the digital hills are formed by taking the probability density value as the digital surface altitude and taking the operating condition parameters as the digital surface plane coordinates.

[0032] Specifically, the probability density distribution diagram calculated based on the scattered data of the working condition reflects the probability density under different working condition parameter combinations. For example, a suitable density estimation method, such as kernel density estimation (KDE), is selected to calculate the probability density of multiple scattered points. Then, the calculated probability density values ​​are fitted using interpolation techniques (such as spline interpolation) or other smoothing algorithms to ensure the smoothness and continuity of the probability density diagram and generate a continuous working condition probability density diagram.

[0033] Specifically, the operating condition parameters are used as the plane coordinates of the digital surface to establish a two-dimensional coordinate system. For example, the torque and speed are used as the X-axis and Y-axis coordinates respectively. Then, the probability density value in the operating condition probability density map is mapped to the altitude of the digital surface. The higher the probability density value, the higher the corresponding altitude, forming the ups and downs of the digital hills. Finally, the plane coordinate system and the altitude are combined to form a three-dimensional digital hill model.

[0034] Optionally, use a three-dimensional modeling tool (such as MATLAB, Python's matplotlib library, etc.) to visualize the formed digital hills.

[0035] The above method steps more accurately reflect the distribution of the working condition data through probability density calculation and fitting, and then construct a digital hill, making the distribution and density changes of the working condition data more intuitive and easy to understand.

[0036] In some implementations, simulating watering of the digital hill and performing adaptive working condition division to determine a plurality of calibration working conditions are stored as the calibration working condition set, including: The digital surface altitude zero point of the digital hill is used as a starting point for simulated watering to raise the water level.

[0037] When the horizontal plane passes through the first digital peak, it is taken as the first calibration center.

[0038] Continue to increase the horizontal plane height, and record the digital peaks that have not been passed, and merge them with the first calibration center to obtain a calibration center set.

[0039] The calibration center set is used as the cluster center set, cluster optimization is performed on the plurality of operating condition scattered points, and the cluster center set is updated, and the updated cluster center set is output as the calibration operating condition set.

[0040] Among them, the simulated water filling based on digital hills is a process of identifying different working condition areas by simulating the process of water gradually filling the digital hills from low to high. Specifically, first, starting from the lowest point of the digital hills (zero altitude), the simulated water level is gradually increased. Including using image processing technology (such as flood filling algorithm) or through three-dimensional modeling simulation to simulate the expansion process of water. Then, in the process of simulating water filling, the position of the first digital peak flooded by water is recorded. And the highest point of the digital peak is used as the first calibration center, and its coordinates and related working condition parameters are recorded. Then, continue to increase the water level, record the positions of subsequent flooded digital peaks, and merge these positions with the first calibration center to form an initial calibration center set. This calibration center set represents the set of working condition points with the highest frequency of occurrence determined based on probability.

[0041] Furthermore, clustering algorithms (such as K-means, DBSCAN, etc.) are used to cluster all the scattered points of working conditions with the calibration center set as the initial cluster center. The location and number of cluster centers are adjusted through iterative optimization to ensure that each cluster center can represent an independent working condition area. The cluster center set after cluster optimization is then used as the final calibration working condition set.

[0042] The above method realizes the precise division and optimization of the two-step working condition data through simulated watering, clustering algorithm and data analysis technology, improves the representativeness and selection efficiency of calibration working condition points, and enhances the practical value of the obtained calibration working condition set.

[0043] Obtain the test requirement scenario, and set a standard test condition series in combination with the calibration condition set.

[0044] Optionally, a test requirement scenario refers to an application or usage scenario corresponding to the performance test of the target engine. The test requirement scenario reflects the test requirements under a specific application scenario and is used to define the types and conditions of operating conditions that need to be tested.

[0045] Specifically, first determine the test requirement scenario, including the test purpose, test environment, test conditions, etc. For example, the car's urban commuting, cross-city commuting, high-speed travel, mixed road and other scenarios. Optionally, the test requirement scenario also includes specific test environment parameters, such as load conditions, ambient temperature, ambient weather, etc.

[0046] Specifically, analyze the working condition parameters of each calibration center in the calibration working condition set, establish and analyze the corresponding relationship between the calibration working condition set and the working condition type and conditions. Then, match the test demand scenario with the calibration working condition set, and select the appropriate calibration center as the test working condition point. For example, the urban driving scenario needs to select a low-speed, frequent start-stop working condition point, while the high-speed driving scenario needs to select a high-speed, stable working condition point.

[0047] Specifically, according to the test requirement scenario, a standard test condition series is set. Each test condition series includes multiple test condition sections, each section corresponds to one or more calibration condition points, and the proportion and order of each section are set. Exemplarily, for the urban driving test series, the proportion of the test condition series is: low-speed driving (30%), frequent start and stop (40%), medium-speed driving (30%). For the high-speed driving test series, the proportion is high-speed stable driving (70%), high-speed acceleration (15%), and high-speed deceleration (15%).

[0048] Specifically, a specific test condition sequence is generated according to the set test condition segment ratio and sequence. For example, the test condition sequence for urban driving is: low-speed driving-frequent start-stop-medium-speed driving-low-speed driving-frequent start-stop-medium-speed driving. The test condition sequence for high-speed driving is: high-speed stable driving-high-speed acceleration-high-speed deceleration-high-speed stable driving-high-speed acceleration-high-speed deceleration. The setting of the standard test condition series makes the test process more standardized and systematic, which helps to improve the test efficiency and effect.

[0049] By analyzing the test demand scenarios, obtaining the actual test requirements, and adjusting the proportion of each test condition, the set standard test condition series can better reflect the actual usage, thereby ensuring the representativeness and effectiveness of the test results.

[0050] Construct a working condition enhancement channel, take the standard test working condition series and test scenario characteristics as input, activate the working condition enhancement channel to perform working condition enhancement, and generate an enhanced test working condition series.

[0051] Among them, the working condition enhancement channel is used to optimize and enhance the test conditions to improve the test efficiency and effect. The test scenario characteristics refer to the characteristic parameters of the specific test scenario, including the performance parameters of the test bench, such as bench power, torque monitoring range, temperature monitoring range, pressure monitoring range, cooling power, dynamometer power, etc.

[0052] Specifically, the enhanced test condition series is an optimized test condition sequence generated by the condition enhancement channel. It is an enhanced test condition sequence with a higher load level than the standard test condition series, and can complete the test in a shorter test time while having the same test effectiveness.

[0053] In some embodiments, a working condition enhancement channel is constructed, the standard test working condition series and the test scenario characteristics are used as input, the working condition enhancement channel is activated to perform working condition enhancement, and an enhanced test working condition series is generated, which includes: Based on the calibration operating condition set, a reference operating condition is determined.

[0054] Combined with the test target of the test requirement scenario, the calibration condition set is calibrated with equivalent coefficients and stored as an equivalent mapping matrix.

[0055] Specifically, the benchmark operating condition is a standard operating condition used to measure the relative strength or intensity of other operating conditions. Preferably, the benchmark operating condition is the most common or representative operating condition. Exemplarily, a medium-speed, stable driving condition such as urban driving is selected from the calibration operating condition set as the benchmark operating condition.

[0056] Specifically, the above-mentioned equivalence coefficient is used to reflect the numerical relationship of the relative strength or severity of different working conditions, and is used for conversion between working conditions. The equivalence coefficient is calculated based on the characteristic parameters of the benchmark working condition and other working conditions in the test items (such as fuel consumption, wear, load, etc.). For example, if the benchmark working condition of 1 hour is equivalent to the second working condition of 15 minutes, the equivalence coefficient of the second working condition can be 0.25 (or 4), where 0.25 means that the benchmark working condition of 1 unit time can be converted to the second working condition of 0.25 unit time, and 4 means that the strength or severity of the second working condition under the current test item is 4 times that of the benchmark working condition.

[0057] Furthermore, the calibration result of the equivalent coefficient is stored as an equivalent mapping matrix, which includes the equivalent coefficients of different working conditions, so as to facilitate the conversion between the equivalent coefficient calls and working conditions.

[0058] Optionally, in the equivalent mapping matrix, the rows and columns of the matrix represent different working conditions, and the elements in the matrix represent the equivalent coefficients between the corresponding working conditions, thereby helping to improve the efficiency of calling the equivalent coefficients in the working condition conversion.

[0059] In some embodiments, a working condition enhancement channel is constructed, and the standard test working condition series and the test scenario characteristics are used as input, and the working condition enhancement channel is activated to perform working condition enhancement, and an enhanced test working condition series is generated, including: Based on the calibration condition set, a calibration condition space is generated with the proportions of multiple calibration conditions as multi-dimensional coordinate axes.

[0060] In combination with the test scenario characteristics, random selection is performed in the calibration working condition space to construct and initialize a particle swarm of candidate working conditions.

[0061] An enhanced evaluation function is constructed, and the enhanced evaluation function is combined with the equal generation mapping matrix to perform iterative optimization of the candidate working condition particle group to obtain the optimal candidate working condition particles, and output them as the enhanced test condition series.

[0062] Specifically, the proportion of multiple working conditions in the calibration working condition is used as multiple coordinate axes to construct a multidimensional space to form the above-mentioned calibration working condition space. In other words, each calibration working condition corresponds to a dimension, and the value of the dimension represents the proportion of the working condition in the entire test. Furthermore, the points in the calibration space represent a scheme of a test working condition sequence.

[0063] Specifically, according to the specific test scenario characteristics, random selection is performed in the calibration condition space to generate an initial candidate condition particle group. The above test scenario characteristics are selection constraints for the calibration condition space, and the test scenario characteristics specify a reasonable range of the test condition sequence scheme, that is, the particles in the candidate condition particle group randomly selected and generated under the constraints of the test scenario characteristics are executable in the target test scenario, that is, they can be executed by the target detection bench.

[0064] Furthermore, an enhanced evaluation function is constructed to evaluate the quality of each candidate working condition particle. Specifically, the enhanced evaluation function can comprehensively consider multiple factors, such as test coverage, test efficiency, resource consumption, etc. The value of the enhanced evaluation function reflects the performance of the candidate working condition particle.

[0065] Specifically, the particle swarm of alternative working conditions is iteratively optimized using the enhanced evaluation function and the equal-generation mapping matrix until the optimal alternative working condition particle is found, including calling the equal-generation mapping matrix to convert between different working conditions to ensure that the working condition combination can be reasonably adjusted during the optimization process; the particle swarm optimization algorithm (PSO) or other optimization algorithms are used to update the coordinate position of the particle according to the score of the current particle and the score of the global optimal particle, and the particle speed is updated according to the change in the particle position to find the optimal working condition combination.

[0066] Specifically, through multiple iterations of optimization, the particle with the highest score is the optimal candidate working condition particle, which is output as an enhanced test working condition series. Through the above steps, the constructed working condition enhancement channel can optimize the test working condition series, thereby improving the test efficiency and effect.

[0067] In some implementations, constructing a reinforcement evaluation function further includes: The enhanced evaluation function expression is as follows: ; in, Design time for the series of standard test conditions; , Respectively characterize the lowest equivalent coefficient value of the particles of the alternative working condition and the standard test working condition series; Characterize the proportion of working conditions corresponding to the working condition with the lowest value of the intermediate generation coefficient of particles in the alternative working conditions; Characterizes the proportion of working conditions corresponding to the working condition with the lowest value of the generation coefficient in the series of standard test conditions; , Respectively represent the types of working conditions in the calibration working condition set and the standard test working condition series; n represents the types of working conditions in the alternative working condition particles; , The test energy consumption corresponding to the condition with the lowest generation coefficient value in the alternative working condition particles and the test energy consumption corresponding to the condition with the lowest generation coefficient value in the standard test working condition series are respectively characterized.

[0068] Specifically, the enhanced evaluation function is used to evaluate and quantify the pros and cons of each alternative working condition particle by comprehensively considering factors such as design time, test energy consumption, operating condition proportion and operating condition type, so as to find an optimized working condition combination to ensure high test efficiency, low energy consumption and comprehensive coverage.

[0069] in, is the test duration of the enhanced standard working condition in the alternative working condition particles. The test duration is determined based on the equivalent coefficients of the two standard working conditions before and after the enhancement of the corresponding part. For example, if the corresponding lowest equivalent coefficient value in the standard test condition series is 1.2, and the corresponding lowest equivalent coefficient value in the alternative working condition particles is 3, then the test duration of this part after enhancement is shortened to 0.4 times the original test duration, and the corresponding shortened duration is 0.4 .and then, It is the total test duration of the particles under alternative working conditions.

[0070] in, It represents the degree of deviation of the proportion of the enhanced standard working condition in the alternative working condition particle from the proportion of the standard working condition in the corresponding position in the standard test working condition series. The smaller the deviation, the closer the proportion of the working condition of the alternative working condition particle is to that of the standard test working condition series, and the corresponding alternative working condition particle can be better equivalent to the standard test working condition series. It represents the overall difference in the proportion of the alternative working condition particles and the standard test condition series.

[0071] The enhanced test condition series is synchronized to the target detection bench for test control, and a test data set is obtained.

[0072] Optionally, after the enhanced test condition series is generated, the conditions in the enhanced test condition series are synchronized to the target detection bench and test control is performed.

[0073] Specifically, first, the enhanced test condition series is transmitted to the control system of the test bench, and the format is converted, such as from CSV format to a format supported by the test bench control system. Ensure that the data format is compatible with the control system of the target test bench. Then, based on the enhanced test condition series, the relevant test control parameters are set on the target test bench to ensure that the test is carried out in accordance with the enhanced test condition series. Exemplary include condition parameters (including speed, torque, acceleration, load, etc.), test time (test time and order of each condition), and environmental conditions (such as temperature, humidity, etc.).

[0074] Furthermore, the target detection bench is started, and the detection bench automatically performs the test according to the preset working condition parameters. During the test, the operating status of the detection bench is monitored in real time, and the test data is collected and saved. By synchronizing the enhanced test condition series to the target detection bench, and performing test control and data collection, a comprehensive test data set can be obtained.

[0075] Furthermore, the method further comprises: A series of control outputs of the test bench is extracted based on the test results.

[0076] The control output series is compared with the enhanced test condition series to obtain condition control differences, and control feedback optimization of the test bench is performed based on the condition control differences.

[0077] Specifically, the control output series of the test bench is extracted from the test data set. The control output series is the actual output data generated by the test bench during the test process, including load, torque, temperature, etc.

[0078] Specifically, the stored test data is parsed to extract the control output parameters at each time point. Then, the control output data is synchronized with the enhanced test condition series in time for accurate comparison.

[0079] Furthermore, the extracted control output series is compared with the original enhanced test condition series to identify the control difference of the condition. For example, the difference between the control output series and the enhanced test condition series at each time point is calculated, including absolute error, relative error, etc. Then, the distribution and trend of the difference are analyzed to identify significant control deviations and abnormal points.

[0080] Specifically, based on the comparison results, the working condition control difference is obtained. The working condition control difference refers to the deviation between the actual output of the test bench and the expected working condition. Optionally, the working condition control difference is represented by a difference matrix, which records the difference values ​​of each working condition parameter at different time points.

[0081] Specifically, the control feedback optimization of the test bench is carried out based on the operating condition control difference. First, a difference feedback mechanism is established to feed back the operating condition control difference to the control system of the test bench. Then, optimization algorithms such as PID adjustment and fuzzy control are used to tune the control parameters according to the real-time operating condition control difference, thereby improving the test accuracy and effect.

[0082] In summary, the intelligent control method of the skid-mounted test bench for engine performance testing provided by the present invention has the following technical effects: Through big data, the target engine's operating condition data set is extracted, and the calibration operating condition set is determined through statistical analysis methods; the test demand scenario is obtained, and the standard test condition series is set in combination with the calibration operating condition set; the operating condition enhancement channel is constructed, and the standard test condition series and test scenario characteristics are used as input to activate the operating condition enhancement channel for operating condition enhancement and generate an enhanced test condition series; the enhanced test condition series is synchronized to the target detection bench for test control and the test data set is obtained. This achieves the technical effect of shortening the test cycle and reducing the test cost.

[0083] It should be understood that the embodiments disclosed in the present invention and the above description can enable those skilled in the art to use the present invention to implement the present invention. At the same time, the present invention is not limited to the above-mentioned embodiments. It should be understood that those skilled in the art can still modify the technical solutions recorded in the above-mentioned embodiments, or replace some of the technical features therein by equivalents; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of the present invention, and should be included in the protection scope of the present invention.

Claims

1. An intelligent control method for a skid-mounted test bench for engine performance testing, characterized in that: The method comprises: Based on big data, the target engine's operating condition data set is extracted, and the calibration operating condition set is determined through statistical analysis methods; Obtain the test requirement scenario, and set a standard test condition series in combination with the calibration condition set; Constructing a working condition enhancement channel, taking the standard test working condition series and the test scenario characteristics as input, activating the working condition enhancement channel to perform working condition enhancement, and generating an enhanced test working condition series; The enhanced test condition series is synchronized to the target detection bench for test control, and a test data set is obtained.

2. The intelligent control method for a skid-mounted test bench for engine performance testing according to claim 1, characterized in that: Based on big data, the target engine's operating condition data set is extracted, and the calibration operating condition set is determined through statistical analysis methods, including: Constructing an operating condition mapping space, wherein a plurality of coordinate axes of the operating condition mapping space correspond to a plurality of operating condition parameters of the target engine; Parsing the operating condition data set and mapping a plurality of operating condition data into a plurality of operating condition scatter points in the operating condition mapping space to form an operating condition scatter point density map; Using the working condition scatter density map as a base, constructing a digital hill, wherein the digital hill includes a plurality of digital peaks; The digital hill is watered by simulation, and adaptive working condition division is performed to determine a plurality of calibration working conditions, which are stored as the calibration working condition set.

3. The intelligent control method for a skid-mounted test bench for engine performance testing according to claim 2, characterized in that: A digital hill is constructed based on the operating condition scatter density map, wherein the digital hill includes a plurality of digital peaks, including: Based on the operating condition scatter point density map, the probability density of multiple scatter points is calculated, and the operating condition probability density map is obtained by fitting; According to the operating condition probability density map, the digital hills are formed by taking the probability density value as the digital surface altitude and taking the operating condition parameters as the digital surface plane coordinates.

4. The intelligent control method for a skid-mounted test bench for engine performance testing according to claim 3, characterized in that: The digital hill is watered by simulation, and adaptive working condition division is performed to determine a plurality of calibration working conditions, which are stored as the calibration working condition set, including: Taking the digital surface altitude zero point of the digital hill as the starting point, simulated watering is performed to raise the water level; When the horizontal plane passes through the first digital mountain, it is taken as the first calibration center; Continue to increase the horizontal plane height, and respectively record the digital peaks that have not been passed, and merge them with the first calibration center to obtain a calibration center set; The calibration center set is used as the cluster center set, cluster optimization is performed on the plurality of operating condition scattered points, and the cluster center set is updated, and the updated cluster center set is output as the calibration operating condition set.

5. The intelligent control method for a skid-mounted test bench for engine performance testing according to claim 1, characterized in that: Constructing a working condition enhancement channel, taking the standard test working condition series and the test scenario characteristics as input, activating the working condition enhancement channel to enhance the working condition, and generating an enhanced test working condition series, previously including: Determining a reference operating condition based on the calibration operating condition set; Combined with the test target of the test requirement scenario, the calibration condition set is calibrated with equivalent coefficients and stored as an equivalent mapping matrix.

6. The intelligent control method for a skid-mounted test bench for engine performance testing according to claim 5, characterized in that: Constructing a working condition enhancement channel, taking the standard test working condition series and the test scenario characteristics as input, activating the working condition enhancement channel to perform working condition enhancement, and generating an enhanced test working condition series, including: Based on the calibration condition set, a calibration condition space is generated with the proportions of a plurality of the calibration conditions as multi-dimensional coordinate axes; In combination with the test scenario characteristics, randomly select in the calibration working condition space to construct and initialize the candidate working condition particle swarm; An enhanced evaluation function is constructed, and the enhanced evaluation function is combined with the equal generation mapping matrix to perform iterative optimization of the candidate working condition particle group to obtain the optimal candidate working condition particles, and output them as the enhanced test condition series.

7. The intelligent control method for a skid-mounted test bench for engine performance testing according to claim 6, characterized in that: Constructing a reinforcement evaluation function also includes: The enhanced evaluation function expression is as follows: ; in, Design time for the series of standard test conditions; , Respectively characterize the lowest equivalent coefficient value corresponding to the particles of the alternative working condition and the standard test working condition series; Characterize the proportion of working conditions corresponding to the working condition with the lowest value of the intermediate generation coefficient of particles in the alternative working conditions; Characterizes the proportion of working conditions corresponding to the working condition with the lowest value of the generation coefficient in the series of standard test conditions; , Respectively represent the types of working conditions in the calibration working condition set and the standard test working condition series; n represents the types of working conditions in the alternative working condition particles; , The test energy consumption corresponding to the condition with the lowest generation coefficient value in the alternative working condition particles and the test energy consumption corresponding to the condition with the lowest generation coefficient value in the standard test working condition series are respectively characterized.

8. The intelligent control method for a skid-mounted test bench for engine performance testing according to claim 1, characterized in that: The method further comprises: Extracting a control output series of the test bench based on the test results; The control output series is compared with the enhanced test condition series to obtain condition control differences, and control feedback optimization of the test bench is performed based on the condition control differences.

Citation Information

Patent Citations

  • Data visualization method and data visualization system based on hierarchical model

    CN105912562A

  • Driving motor steady state reliability test condition construction method and system

    CN116756986A

  • Engine calibration method and device, electronic equipment and storage medium

    CN117147166A

  • Intelligent automobile standard test scene generation method based on inverse reinforcement learning

    CN118228591A

  • Vehicle emission evaluation method and system, electronic equipment and storage medium

    CN118603578A