Heavy-duty commercial vehicle auxiliary braking working condition coverage analysis method

Abstract

The invention discloses a heavy-duty commercial vehicle auxiliary braking condition coverage analysis method, which comprises the following steps: collecting vehicle speed, gradient, braking time and braking switch signals during vehicle operation, and cleaning and standardizing data by using a big data technology; adopting a machine learning regression algorithm to construct a vehicle speed-slope relation model; analyzing braking force and braking power of engine braking and retarder braking based on the model, and converting the braking force and the braking power into a vehicle speed-slope fitting curve; and by comparing actual braking demand curves, the braking coverage and the combined coverage of the two are evaluated, and a braking system optimization suggestion is proposed. According to the method, based on big data and a machine learning regression algorithm, the braking performance of the vehicle can be accurately evaluated, the configuration of a braking system is optimized, and the safety and reliability of the vehicle under various working conditions are ensured.

Description

Technical Field

[0001] The present invention relates to the technical field of automobile braking systems, and in particular to a method for analyzing auxiliary braking conditions coverage of heavy-duty commercial vehicles. Background Art

[0002] The auxiliary braking system of heavy-duty commercial vehicles includes engine braking and retarder braking. In the extraction and analysis of auxiliary braking conditions of commercial vehicles, the existing technology relies on laboratory simulation, which has problems such as data being out of touch with reality, high cost, and long cycle. Summary of the invention

[0003] The present invention provides a heavy-duty commercial vehicle auxiliary braking condition coverage analysis method based on big data and machine learning regression algorithm, which solves the limitations of traditional methods in terms of high number of experiments and high labor costs. By combining actual operating data with a machine learning algorithm, a relationship model between vehicle speed and slope is constructed to verify the matching of the vehicle braking system. By calculating the braking condition coverage, a parameter basis is provided for the optimization of the braking system.

[0004] The present invention is achieved by adopting the following technical solutions:

[0005] A heavy-duty commercial vehicle auxiliary braking condition coverage analysis method based on big data and machine learning regression algorithm, including:

[0006] Collect vehicle and brake system related operating data, including switch signal data, vehicle speed data, altitude data, and calculate the slope based on the altitude data;

[0007] The random forest regression algorithm is used to construct the relationship model between vehicle speed and slope;

[0008] Based on the relationship model, the braking force and braking power of the engine braking and the retarder braking are calculated and converted into a vehicle speed-slope scatter plot;

[0009] The vehicle speed-slope scatter plot is subjected to boundary fitting to obtain a fitting curve, which is compared with the actual braking demand curve to evaluate the coverage of the braking conditions by engine braking or retarder braking, as well as the coverage of the braking conditions combined with the two. Based on the analysis results, an optimization strategy for the braking system of heavy commercial vehicles is generated.

[0010] According to a further explanation of the present invention, the method described comprises the following specific steps:

[0011] S1. Data collection and preprocessing: Collect switch signal data of engine braking and retarder braking, vehicle speed data, and altitude data, and remove abnormal values ​​through big data filtering technology; filter the altitude data, calculate the slope, and average the vehicle load according to the operation start and stop;

[0012] S2. Braking condition extraction: extract the braking working status of the engine and retarder of the vehicle under different working conditions according to the brake switch signal data, and classify and sort the extracted braking conditions in chronological order;

[0013] S3. Calculation of braking force and braking power: Calculate the braking force and braking power under engine braking and retarder braking using the power balance formula based on vehicle mass, speed, and acceleration parameters;

[0014] S4. Relationship model construction: With vehicle speed as input and slope as output, the model is trained using the random forest regression algorithm, and the top 10% high slope data are selected to fit the maximum coverage curve;

[0015] S5. Braking coverage analysis: The braking power of the engine and retarder to be compared and analyzed is converted into a vehicle speed-slope scatter plot, and a fitting curve is obtained by fitting the maximum value of the upper boundary of the scatter plot, and compared with the actual braking demand curve to calculate the coverage ratio;

[0016] S6. Application of analysis results: Adjust braking parameters according to coverage results, optimize braking strategy, and verify whether it is necessary to match the retarder or adjust the engine braking function.

[0017] Preferably, in step S1, the altitude data is smoothed using a moving average filtering method to remove abnormal values.

[0018] Preferably, in step S1, the slope calculation formula is:

[0019]

[0020] Where α is the slope, h is the altitude difference, and d is the horizontal distance, which is calculated by the mileage difference c and the altitude difference h:

[0021]

[0022] Preferably, in step S1, the vehicle load data is segmented according to the start and stop speed, and the load data of each segment is combined with the operating mileage ratio to obtain the average load m of the vehicle start and stop segment.

[0023] Preferably, in step S3, the braking force is calculated based on the power balance formula:

[0024]

[0025] Among them, F 制动力 is the braking force, m is the average load of the vehicle during the start-stop phase, g is the acceleration of gravity 9.8N / kg, v is the vehicle speed, a is the acceleration, C D , A, and f are respectively the air resistance coefficient, the frontal area, and the rolling resistance coefficient;

[0026] Calculate the braking power:

[0027] W=F 制动力 ×v;

[0028] Among them, W is the braking power and v is the vehicle speed.

[0029] Preferably, in step S4, the hyperparameters of the random forest regression model are optimized by grid search, and the feature importance ranking is used to explain the slope-vehicle speed correlation. The risk of overfitting of the model to the training data is reduced by automatically constraining the parameters, and the generalization ability is enhanced to traverse the parameter combination, avoiding the blindness of manual parameter adjustment, thereby ensuring that the model achieves the optimal prediction accuracy in complex data.

[0030] Preferably, in step S5,

[0031] In step S5, the coverage calculation formula is:

[0032]

[0033] Compared with the prior art, the present invention has the following beneficial effects:

[0034] 1. Big data driven: The present invention is based on big data analysis and can make full use of actual operating data to achieve accurate analysis of braking conditions.

[0035] 2. Improve the matching degree of the braking system: Through the machine learning regression algorithm, the present invention can construct a relationship model between vehicle speed and slope, realize the verification of the matching of the vehicle braking system, and provide parameter basis for the optimization of the braking system by calculating the coverage of braking conditions.

[0036] 3. Reduce R&D costs: The present invention utilizes big data and machine learning technology to reduce the number of experiments and labor costs, thereby reducing R&D costs.

[0037] 4. The method of the present invention can utilize big data to extract and analyze the engine braking and retarder braking conditions of a large number of operating vehicles, and combine it with a machine learning regression algorithm to construct a relationship model to obtain the braking condition requirements of the customer's vehicle during actual operation. Combined with the analysis of the braking capacity of the engine and retarder, the coverage of the engine and retarder braking conditions is obtained, providing a parameter basis for the matching of vehicle engine braking and retarder braking. BRIEF DESCRIPTION OF THE DRAWINGS

[0038] Attached Figure 1 The figure is a flow chart of the auxiliary braking condition coverage analysis method for heavy-duty commercial vehicles based on big data and machine learning regression algorithm. DETAILED DESCRIPTION

[0039] In order to enable those skilled in the art to better understand the technical solutions in the present invention, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work should fall within the scope of protection of the present invention.

[0040] like Figure 1 As shown, the heavy-duty commercial vehicle auxiliary braking condition coverage analysis method based on big data and machine learning regression algorithm of the present invention has an overall framework of six core modules: data collection and preprocessing, braking condition extraction, braking force parameter output, relationship model construction, coverage analysis, and analysis result application. It mainly extracts and analyzes the braking conditions of massive operational data through big data technology, and constructs a speed-slope relationship model through the machine learning regression algorithm random forest. Based on the constructed model, combined with the braking condition coverage, optimization and improvement suggestions for the braking system of heavy-duty commercial vehicles are proposed.

[0041] Auxiliary braking condition coverage analysis method for heavy-duty commercial vehicles based on big data and machine learning regression algorithm

[0042] S1. Data collection and preprocessing:

[0043] Collect data related to the braking system from massive amounts of operating vehicle data, namely, collect switch signal data of engine braking and retarder braking, vehicle speed data, altitude data, and load data; remove abnormal values ​​through big data filtering technology, such as altitude less than 0, vehicle speed greater than 120km / h, etc.

[0044] The moving average filtering method is used to smooth the altitude data and remove outliers. After the altitude data is filtered, the mileage data and altitude data are used to differentiate the data to calculate the horizontal distance. The slope is then calculated based on the ratio of the altitude difference to the horizontal distance. The formula is:

[0045]

[0046] Where α is the slope, h is the altitude difference, and d is the horizontal distance, which is calculated by the mileage difference c and the altitude difference h:

[0047]

[0048] In step S1, the vehicle load data is segmented according to the start and stop speed, and the load data of each segment is combined with the operating mileage ratio to obtain the average load m of the vehicle start and stop segment, which is applied to the calculation of the braking force formula in S3.

[0049] S2. Braking condition extraction:

[0050] Based on the collected data, the braking status of the vehicle under different working conditions is extracted, including the working status of engine braking and retarder braking, and the extracted braking conditions are classified and sorted in chronological order to facilitate subsequent analysis and calculation;

[0051] S3. Calculation of braking force and braking power:

[0052] Using the power balance formula, the braking force and braking power under engine braking and retarder braking are calculated based on the vehicle mass, speed, and acceleration parameters;

[0053] In step S3, the braking force is calculated based on the power balance formula:

[0054]

[0055] Among them, F 制动力 is the braking force, m is the average load of the vehicle during the start-stop phase, g is the acceleration of gravity 9.8N / kg, v is the vehicle speed, a is the acceleration, C D , A, and f are respectively the air resistance coefficient, the frontal area, and the rolling resistance coefficient;

[0056] Calculate the braking power:

[0057] W=F 制动力 ×v;

[0058] Among them, W is the braking power and v is the vehicle speed.

[0059] S4. Construction of the speed and slope relationship model:

[0060] Based on the collected vehicle speed and slope data of engine braking and retarder braking under actual operating conditions, the relationship model is trained using the random forest regression algorithm, and the top 10% high slope data is selected for fitting processing to obtain the maximum coverage curve with the X-axis as the vehicle speed and the Y-axis as the slope;

[0061] In step S4, the hyperparameters of the random forest regression model are optimized through grid search, and the feature importance ranking is used to explain the slope-vehicle speed correlation. The risk of overfitting the model to the training data is reduced by automatically constraining the parameters, and the generalization ability is enhanced to traverse the parameter combination, avoiding the blindness of manual parameter adjustment, ensuring that the model achieves the optimal prediction accuracy in complex data.

[0062] S5. Braking coverage analysis:

[0063] The braking power of the engine and retarder that need to be compared and analyzed is converted into a speed slope scatter plot, and the maximum value of the upper boundary of the scatter plot is fitted to obtain a fitting curve, which is compared with the actual braking condition demand curve. By comparing the parameters of the two curves, the braking coverage of the engine to be analyzed and the braking coverage of the retarder to be analyzed, as well as the combined braking condition coverage of the two, can be calculated.

[0064] In step S5, the coverage calculation formula is:

[0065]

[0066] S6. Application of analysis results:

[0067] The results of the brake coverage analysis can be used to adjust the parameter settings of engine braking and retarder braking and optimize the braking strategy of the brake system. R&D personnel can verify whether the vehicle needs to match the retarder and engine braking functions based on the coverage analysis results, and provide parameter support for vehicle brake system matching.

[0068] The present invention uses a heavy-duty commercial vehicle auxiliary braking condition coverage analysis method based on big data and machine learning regression algorithm, which can accurately evaluate the vehicle's braking performance and optimize the braking system configuration to ensure the safety and reliability of the vehicle under various working conditions.

[0069] The method of the present invention can utilize big data to extract and analyze the braking conditions of massive operating vehicles, and combine with machine learning regression algorithm to construct a relationship model between vehicle speed and slope, so as to obtain the braking condition requirements of customer vehicles during actual operation, and combine with the braking capacity analysis of the engine and retarder to obtain the braking condition coverage, thereby providing parameter basis for matching vehicle engine braking and retarder braking.

[0070] The above description is only an example of the implementation of the present invention and does not limit the present invention in any form. The protection scope of the present invention shall be based on the claims and is not limited by the above specific embodiments. Any simple modification or equivalent changes and modifications made to the above implementation according to the technical essence of the present invention shall fall within the protection scope of the present invention.

Claims

1. A heavy-duty commercial vehicle auxiliary braking condition coverage analysis method, characterized in that: The following steps are involved: Collect vehicle and brake system related operating data, including switch signal data, vehicle speed data, altitude data, and calculate the slope based on the altitude data; The random forest regression algorithm is used to construct the relationship model between vehicle speed and slope; Based on the relationship model, the braking force and braking power of the engine braking and the retarder braking are calculated and converted into a vehicle speed-slope scatter plot; The vehicle speed-slope scatter plot is subjected to boundary fitting to obtain a fitting curve, which is compared with the actual braking demand curve to evaluate the coverage of the braking conditions by engine braking or retarder braking, as well as the coverage of the braking conditions combined with the two. Based on the analysis results, an optimization strategy for the braking system of heavy commercial vehicles is generated.

2. The method according to claim 1, characterized in that The specific steps include: S1. Data collection and preprocessing: Collect switch signal data of engine braking and retarder braking, vehicle speed data, and altitude data, and remove abnormal values ​​through big data filtering technology; filter the altitude data, calculate the slope, and average the vehicle load according to the operation start and stop; S2. Braking condition extraction: extract the braking working status of the engine and retarder of the vehicle under different working conditions according to the brake switch signal data, and classify and sort the extracted braking conditions in chronological order; S3. Calculation of braking force and braking power: Calculate the braking force and braking power under engine braking and retarder braking using the power balance formula based on vehicle mass, speed, and acceleration parameters; S4. Relationship model construction: With vehicle speed as input and slope as output, the model is trained using the random forest regression algorithm, and the top 10% high slope data are selected to fit the maximum coverage curve; S5. Braking coverage analysis: The braking power of the engine and retarder to be compared and analyzed is converted into a vehicle speed-slope scatter plot, and a fitting curve is obtained by fitting the maximum value of the upper boundary of the scatter plot, and compared with the actual braking demand curve to calculate the coverage ratio; S6. Application of analysis results: Adjust braking parameters according to coverage results, optimize braking strategy, and verify whether it is necessary to match the retarder or adjust the engine braking function.

3. The method according to claim 2, characterized in that: In step S1, the altitude data is smoothed using a moving average filtering method to remove abnormal values.

4. The method according to claim 2, characterized in that: In step S1, the slope calculation formula is: Where α is the slope, h is the altitude difference, and d is the horizontal distance, which is calculated by the mileage difference c and the altitude difference h:

5. The method according to claim 2, characterized in that: In step S1, the vehicle load data is segmented according to the start and stop speed, and the load data of each segment is combined with the operating mileage ratio to obtain the average load m of the vehicle start and stop segment.

6. The method according to claim 5, characterized in that: In step S3, the braking force is calculated based on the power balance formula: Among them, F 制动力 is the braking force, m is the average load of the vehicle during the start-stop phase, g is the acceleration of gravity 9.8N / kg, v is the vehicle speed, a is the acceleration, C D , A, and f are respectively the air resistance coefficient, the frontal area, and the rolling resistance coefficient; Calculate the braking power: W=F 制动力 ×v; Among them, W is the braking power and v is the vehicle speed.

7. The method according to claim 2, characterized in that: In step S4, the hyperparameters of the random forest regression model are optimized by grid search, and feature importance ranking is used to explain the slope-vehicle speed correlation.

8. The method according to claim 2, characterized in that: In step S5, the coverage calculation formula is:

Images