Method for estimating the mass of a road freight vehicle based on a lane keeping assist system

Abstract

The present invention discloses a method for estimating the mass of a road freight vehicle based on a lane keeping assist system. During the entire operation of the vehicle, the vehicle's overall mass estimation system retrieves the lane keeping pictures generated in the lane keeping assist system. By capturing the position where the lower edge of the lane image taken by the camera occupies the lane keeping pictures generated by the lane keeping assist system, and using the mapping relationship between Mi and Si stored in the vehicle's overall mass estimation system, the overall mass of the current vehicle can be obtained. The structure of the present invention is simple, without the need to install other sensors, and it is convenient to use. It can be mainly applied to road freight vehicles. By processing the data of the lane keeping assist system, the overall mass of the freight vehicle can be obtained in real time. On the one hand, using the overall mass of the freight vehicle as the input of the vehicle stability control system can effectively improve the running safety of the vehicle; on the other hand, it can determine whether the vehicle is overloaded and is used for the statistics and analysis of road freight volume.

Description

Technical Field

[0001] The present invention belongs to the technical field of vehicle body safety, and relates to a method for estimating the total vehicle mass of a freight truck, in particular to a method for estimating the total vehicle mass based on a lane keeping assist system. Background Art

[0002] The information of the total vehicle mass (weight) of a freight vehicle can serve to assist the driver in safe driving, achieve accurate control of vehicle braking and steering safety, and thus improve the technical level of vehicle operation safety. At the same time, obtaining the vehicle mass information can solve the overloading problem from the source and is also an effective technical means for controlling overloading. Therefore, it is of great significance to carry out the estimation of the total vehicle mass of a freight vehicle.

[0003] At present, there are various on-vehicle methods for estimating the total vehicle mass at home and abroad. These include: 1) a method for estimating the total vehicle mass based on suspension displacement or axle strain calibration; 2) a method for estimating the total vehicle mass based on vehicle dynamics. The method based on the change of vehicle suspension or axle displacement mainly realizes it through sensors and calibration. The method for estimating the total vehicle mass based on vehicle dynamics is an algorithm for estimating the total vehicle mass based on vehicle suspension dynamics, lateral / yaw dynamics, and longitudinal dynamics. By establishing the input-output relationship of the vehicle dynamics model and combining the real-time operation data of the vehicle (from the vehicle itself or additional sensors), methods such as Kalman filtering and recursive least squares are used to continuously update the total vehicle mass until convergence. Almost all of these estimation methods require installing corresponding sensing devices on the vehicle, which makes it difficult to promote and apply in terms of cost.

[0004] With the development of vehicle intelligent technology, the images and video data collected by the vehicle can be applied. For example, the currently highly concerned lane keeping assist system, the popularization and installation of which have become a development trend. It identifies the positions of lane lines and road edges through a camera installed in the cab and judges the distance of the vehicle relative to the lane lines. As one of the advanced driver assistance systems, it is also an essential function for realizing autonomous driving, and it can effectively avoid traffic accidents caused by the vehicle deviating from the normal driving lane. Summary of the Invention

[0005] In order to solve the above problems, the present invention designs a method for estimating the mass of a road freight vehicle based on a lane keeping assist system. For the estimation of the total vehicle mass, through the relationship between the parallel auxiliary horizontal line in the total vehicle mass scale model and the total vehicle mass, the estimation of the total vehicle mass can be realized without additional installation of sensing devices. Only the total vehicle mass estimation system connected to the lane keeping assist system needs to be installed and the data is analyzed and applied, reducing the installation and cost investment of existing on-vehicle total vehicle mass products.

[0006] The technical solution adopted by the present invention is a method for estimating the mass of a road freight vehicle based on a lane keeping assist system. The lane keeping assist system is realized through software processing of a camera installed in the cab of the freight vehicle. The key lies in that: in the lane keeping picture generated by the lane keeping assist system of the freight vehicle, a set of equally spaced parallel auxiliary horizontal lines are added to form a vehicle mass scale model, and the vehicle mass of the freight vehicle is determined according to the position where the lower edge of the lane image captured by the camera occupies the lane keeping picture generated by the lane keeping assist system. The specific steps include:

[0007] a. Add a vehicle mass estimation system. The vehicle mass estimation system retrieves the lane keeping picture generated by the lane keeping assist system, divides the lane keeping picture with a set of equally spaced parallel auxiliary horizontal lines, and generates a vehicle mass scale model;

[0008] b. Calibrate the vehicle, record the positions where the lower edge of the lane image captured by the camera occupies the lane keeping picture generated by the lane keeping assist system when the vehicle is from no-load to full-load and overloaded, form the mapping relationship between Mi and Si as the vehicle weight calibration, and store it in the vehicle mass estimation system, where Mi represents the vehicle mass and Si represents the i-th parallel auxiliary horizontal line from bottom to top in the vehicle mass scale model, and i is an integer from 0 to 200, 0 represents no-load, 50 represents half-load, 100 represents full-load, 150 represents 50% overload, and 200 represents 100% overload;

[0009] c. On this basis, during the whole process of vehicle operation, the vehicle mass estimation system retrieves the lane keeping picture generated by the lane keeping assist system, and through the position where the lower edge of the lane image captured by the camera occupies the lane keeping picture generated by the lane keeping assist system, and the mapping relationship between Mi and Si stored in the vehicle mass estimation system, the vehicle mass of the current vehicle can be obtained.

[0010] During the whole process of vehicle operation, the vehicle mass is estimated by repeating at intervals of a period of time and then averaging to obtain an accurate vehicle mass.

[0011] In step b, the accuracy of vehicle weight calibration is inversely proportional to the spacing between the parallel auxiliary horizontal lines.

[0012] When the vehicle is driving on a curve, the lane keeping assist system first fits the curve into a straight road, and then calculates the vehicle mass through step c.

[0013] The technical principle of the method of the present invention is that, for road freight vehicles, especially heavy vehicles, due to the elasticity of tires and suspensions under different loading conditions, the actual height of the camera installed in the cab lane keeping assist system from the ground is different. Therefore, the relationship between the vehicle's total mass Mi corresponding to different loading levels (such as 0%, 50%, 100%, 150%) and the pixel Si of the closest connection line from the bottom of the camera to the lane lines on both sides of the lane can be calibrated and established. Therefore, during the actual driving of the vehicle, by combining the vehicle lane image collected by the lane keeping system with the established relationship above, the vehicle's total mass can be calculated.

[0014] The beneficial effects of the present invention are as follows: the structure is simple, no other sensors need to be installed, it is easy to use, and it can be mainly applied to road freight vehicles. By processing the data of the lane keeping assist system, the vehicle's total mass can be obtained in real time. On the one hand, taking the vehicle's total mass as the input of the vehicle stability control system can effectively improve the vehicle operation safety; on the other hand, it can determine whether the vehicle is overloaded and is used for the statistics and analysis of road freight volume. With the development of vehicle intelligent technology in recent years, the application of the vehicle's total mass estimation method of the present invention will be more and more extensive. Description of the Drawings

[0015] Figure 1 It is an embodiment of a lane keeping picture of a truck in an empty load state with a vehicle total mass scale.

[0016] Figure 2 It is an embodiment of a lane keeping picture of a truck in a half load state with a vehicle total mass scale.

[0017] Figure 3 It is an embodiment of a lane keeping picture of a truck in a full load state with a vehicle total mass scale.

[0018] Figure 4 It is an embodiment of a lane keeping picture of a truck in an overloaded state with a vehicle total mass scale.

[0019] Figure 5 It is a schematic diagram of the mass estimation principle of the present invention.

[0020] Figure 6 It is a schematic diagram of the mass estimation principle illustrated by a lane keeping picture.

[0021] In the drawings, 1 represents the camera, and 2 represents the parallel auxiliary horizontal line. Detailed Embodiments

[0022] When the method of the present invention is specifically implemented, refer to Figure 5, the core technical principle of the present invention is that due to the large load capacity of freight vehicles, there are relatively large changes in the springs and tires of the vehicle chassis, resulting in changes in the height of the camera 1 in the cab. Therefore, the images captured by the camera 1 are different. Suppose Figure 5 the solid line captured by the camera 1 in Figure 5 represents the field of view when the vehicle is unloaded. Then L1 represents the distance captured by the camera 1 in the unloaded state. When the vehicle is fully loaded, the vehicle chassis springs and tires have large deformations, causing the position of the camera 1 to decrease. Refer to the dotted line part. Since the shooting angle of the camera 1 is fixed, the distance captured by the camera 1 when fully loaded is L2. ΔL represents the difference area of the images that the camera 1 can capture when the vehicle is unloaded and fully loaded. If the difference area can be calibrated, then the calculation of the vehicle's total mass can be achieved.

[0023] The calibration of the difference area is carried out by Figure 6 the method. Refer to Figure 6 . Since the overall picture size of the lane keeping assist system is fixed, if the change in the vehicle's total mass is small, such as adding one or two more people, the position of the camera can be almost ignored. But when it comes to freight vehicles with a large load capacity, the position of the camera has a big difference. Figure 6 The solid line lane part in Figure 6 is the road condition captured by the camera 1 when the vehicle is unloaded, and the dotted line lane part is the road condition captured by the camera 1 when the vehicle is fully loaded. It can be seen that Figure 6 ΔL in Figure 6 is the difference in the road conditions captured by the vehicle when unloaded and fully loaded. For example, the unladen mass of a heavy truck is 14 tons, and the fully loaded mass is 100 tons. There are 10 parallel auxiliary horizontal lines 2 in the middle, indicating that each parallel auxiliary horizontal line 2 represents a weight of 8.6 tons. That is, the more the number of parallel auxiliary horizontal lines 2 and the smaller the spacing, the more accurate the calculation of the vehicle's total mass.

[0024] Before the calculation of the vehicle total mass estimation system, it is certainly necessary to perform mass calibration first. The calibration work of the vehicle's total mass is completed in advance. Each type of vehicle only needs to be calibrated once according to the accuracy of load measurement. That is, calibrate from unladen to fully loaded and overloaded. It can be calibrated once for every 1 ton increase, once for every 2 ton increase, or once for every 5 ton increase. The calibration interval can be calibrated according to different actual test scenarios. Each mass calibration is also achieved by means of the vehicle total mass estimation system, including the mapping relationship between Mi and Si.

[0025] Due to the method of estimating the vehicle's total mass based on the lane keeping assist system, the evaluation of the vehicle's total mass during the entire transportation process of the vehicle is realized. The vehicle total mass data can be used for overloading, improving vehicle stability control, freight volume statistics, government department supervision, etc., which were not achievable by previous weighing methods.

[0026] Of course, the lane keeping assist system is a relatively mature technology in the prior art. The vehicle mass in this solution is mainly obtained by processing and calculating the data with the vehicle mass estimation system by leveraging the data of the lane keeping assist system. It does not upgrade the lane keeping system. For example, regarding how to estimate the mass in a curve, the lane keeping system itself can display the curve as a straight line through pre-simulation, and then the calculation can be carried out through this solution.

[0027] To avoid the position change of the camera 1 during vehicle bumping or going up and down slopes, which may ultimately lead to inaccurate calculation, the method of taking the average of multiple measurements can be adopted to achieve a relatively more accurate mass estimation.

Claims

1. A method for estimating the mass of a road freight vehicle based on a lane keeping assist system, where the lane keeping assist system is implemented through software processing of a camera (1) installed in the cab of the freight vehicle, characterized in that: The method described is to add a set of equally spaced parallel auxiliary horizontal lines (2) to the lane-keeping picture generated by the lane-keeping assistance system of a freight vehicle to form a vehicle gross mass scale model, and determine the vehicle gross mass of the freight vehicle according to the position where the lower edge of the lane image captured by the camera (1) occupies the lane-keeping picture generated by the lane-keeping assistance system. The specific steps include: a. Add a vehicle gross mass estimation system. The vehicle gross mass estimation system retrieves the lane-keeping picture generated in the lane-keeping assistance system and divides the lane-keeping picture with a set of equally spaced parallel auxiliary horizontal lines (2) to generate a vehicle gross mass scale model; b. Calibrate the vehicle. Record the positions where the lower edge of the lane image captured by the camera (1) occupies the lane-keeping picture generated by the lane-keeping assistance system when the vehicle is from no-load to full-load and overloaded, and form the mapping relationship between Mi and Si as the vehicle weight calibration, and store it in the vehicle gross mass estimation system, where Mi represents the vehicle gross mass, Si represents the i-th parallel auxiliary horizontal line (2) from bottom to top in the vehicle gross mass scale model, and i is an integer from 0 to 200, 0 represents no-load, 50 represents half-load, 100 represents full-load, 150 represents 50% overload, and 200 represents 100% overload; c. On this basis, during the whole process of vehicle operation, the vehicle gross mass estimation system retrieves the lane-keeping picture generated in the lane-keeping assistance system. By the position where the lower edge of the lane image captured by the camera (1) occupies the lane-keeping picture generated by the lane-keeping assistance system and the mapping relationship between Mi and Si stored in the vehicle gross mass estimation system, the vehicle gross mass of the current vehicle can be obtained. The lower edge of the lane image described is the parallel horizontal line formed by the intersection of the boundary lines of both sides of the road and the picture.

2. The method for estimating the mass of a road freight vehicle based on a lane keeping assist system according to claim 1, wherein: During the whole process of vehicle operation, the vehicle gross mass is estimated by repeating every other period of time and then averaging to obtain the accurate vehicle gross mass.

3. The method for estimating the mass of a road freight vehicle based on a lane keeping assist system according to claim 1, characterized in that: In the step b described, the accuracy of the vehicle weight calibration is inversely proportional to the spacing between the parallel auxiliary horizontal lines (2).

4. The method for estimating the mass of a road freight vehicle based on a lane keeping assist system according to claim 1, wherein: When the vehicle is driving on a curve, the lane-keeping assistance system first fits the curve into a straight road, and then calculates the vehicle gross mass through step c.

Images