Inertial navigation

Abstract

To provide an inertial navigation of high precision to trace a route from a current position to a target position.SOLUTION: The present invention can draw (detect) a locus drawn by a wheel (tire) on a ground surface, from a wheel speed of the wheel, thus being capable of calculating a curvature radius of the locus. The present invention relates to a control system for acquiring a correlation expression between a curvature (curvature of route) of the locus and a handle angle at that time to trace the route, using the correlation expression as a control formula. The locus varies by not only a handle but also a crossing gradient (cant) of a road, so that an influence of the cant appears in the curvature of the wheel locus. Further, the locus varies also by a variation of a vehicle total weight accompanying the number of passengers and a change in loading, and in vehicle state quantity such as positions of a center of gravity. That is, the present invention has a configuration having the control formula where a vehicle route change based on the handle, the cant, and a loading state is captured by a wheel speed to perform vehicle route control.SELECTED DRAWING: Figure 1

Description

[Technical Field]

[0001] The present invention relates to an inertial navigation system that detects the position and speed of a vehicle using an on-board sensor.

[0002] Automobile steering systems have what is called steering wheel play (plasticity characteristics in the propagation process from steering wheel angle to wheel speed). This steering wheel play is effective when used in manual driving situations such as counter-steering, but in automated driving, it can cause a decrease in the accuracy of course control. This invention relates to an inertial navigation method that obtains the desired course control accuracy while ignoring the reality (existence) of steering wheel play. [Background technology]

[0003] A car's steering wheel has play. The actual steering angle of the front axle changes depending on the steering angle. When the actual steering angle of the front axle changes, a lateral force is generated in the front axle tires, causing the front axle to displace laterally. This lateral displacement of the front axle is propagated through the front axle suspension, through the body, and through the rear axle suspension, resulting in a change in rear axle alignment.

[0004] This propagation process involves the steering power steering, front wheel alignment, front axle alignment, front axle suspension lateral displacement characteristics (front axle lateral force compliance), and rear axle suspension lateral displacement characteristics (rear axle lateral force compliance). The plastic elements involved in this propagation process are factors that cause the aforementioned "play." Design know-how is incorporated into each of these play factors, and vehicle performance (maneuverability and stability) is designed accordingly.

[0005] The power steering incorporates design know-how that reduces the rigidity of the steering system in the extreme center range to achieve disturbance-suppressing forward flow compliance steer, and increases the rigidity outside the extreme center range to achieve high steering responsiveness (Non-Patent Document 1).

[0006] Front wheel alignment involves tilting the kingpin axis to generate kingpin lift torque due to gravity on the left and right wheels, and incorporating other design know-how to determine the steering neutral position from the left and right balance (Non-Patent Document 1).

[0007] The front axle is equipped with forward flow compliance steering, and the rear axle is equipped with reverse flow compliance steering, incorporating design know-how to improve straight-line and cornering stability (Non-Patent Document 2).

[0008] Steering play exhibits hysteresis, with the low frequency of human steering superimposed on the high unsprung frequency.

[0009] When determining a target destination (movement target point), calculating (detecting) the curvature of the target route to get there, and determining the steering angle to follow that route curvature, even if the steering angle is determined, the route curvature will not be stable due to changes in the steering wheel corresponding to hysteresis. If a person were driving, they would perceive the corresponding changes visually, etc., and make appropriate corrective steering.

[0010] In this regard, Patent Document 1 proposes solving the problem of difficulty in determining steering, particularly in small steering angle ranges, due to a change in the ratio of steering angle divided by actual steering angle caused by hysteresis, and of dealing with changes in steering effectiveness caused by changes in vehicle weight due to loading, and proposing a method for controlling the steering angle to plot a target course by setting a target destination point ahead in the traveling direction, determining the curvature of the path to this target destination point from the forward distance, lateral deviation distance, course angle, and vehicle body sideslip angle of the vehicle, determining the actual steering angle for tracing the path from the determined curvature of the path, and multiplying the determined actual steering angle by the diagonal gear ratio to calculate the required steering angle and control the steering motor.

[0011] Roads are equipped with drainage gradients and super-slope gradients to mitigate centrifugal force on curves (Article 16 of the Road Structure Ordinance). Due to this super-slope gradient, or cant, vehicles traveling on the road are subjected to a lateral force due to the balance between the lateral component of the acceleration of gravity multiplied by the cant angle and the centrifugal force caused by the curvature of the road.

[0012] Driving along the lane depends on the balance between the lateral cant force acting from the road surface, the lateral road friction force, and the lateral vehicle yaw force (centrifugal force). Steering is performed to determine the course while applying "counter steering" to counter the change in lateral cant force that accompanies changes in the curvature of the road.

[0013] Taking left-hand traffic as an example, cant on a straight road or left curved road is left cant, and on a right curved road it is right cant, so on straight roads and left curved roads, course control steering in a "right counter steering" state is required to correspond to the left cant, and on a right curved road, course control steering in a "left counter steering" state is required. A "right counter steering" state corresponds to a state in which there is a change in front wheel steering that balances the left cant lateral force and a change in rear axle alignment that balances the left cant, and the opposite corresponds to a "left counter steering" state.

[0014] In the process of changing from "right counter rudder" to "left counter rudder" or from "left counter rudder" to "right counter rudder," it becomes necessary to control the course while adapting to both the change in curvature and the change in cant angle. A control law that can adapt to these changes is required. In the case of a right cant when driving on the right side of the road, the cant counter rudder is the opposite of the left cant when driving on the left side of the road.

[0015] Non-Patent Document 2 points out the problem of low-frequency lateral vehicle movement occurring even without steering input, which flows downwards due to road surface cant, and derives a design measure that incorporates positive compliance steer characteristics into the front axle and adverse compliance steer characteristics into the rear axle, verifying the validity (effectiveness) of this design method in Non-Patent Document 1. Compliance here refers to a certain type of hysteresis required as know-how in vehicle suspension system design, and autonomous driving control must accept the existence of this compliance.

[0016] Non-Patent Document 3 introduces a control model that receives "curvature and cant" as road information and "vehicle speed" as vehicle information, but there remains a problem in dealing with the hysteresis that occurs when "curvature and cant" switches from turning left to turning right.

[0017] Non-Patent Document 4 describes how to determine a target path as follows: "All that is needed is to detect the lateral deviation of the front and rear ends of the vehicle from the desired path using an appropriate method, and perform steering that is sensitive to the sum and difference of the two. This is because the sum of the two is proportional to the lateral deviation of the vehicle's center of gravity from the desired path, and the difference between the two is proportional to the angular deviation of the vehicle's center of gravity from the desired path, and thus shows the principle of steering for a desired straight path. There is no mention of curved paths."

[0018] Furthermore, Non-Patent Document 5 states as follows: "Because a driver can sense changes in the current position and speed direction by the rotation of his field of vision and by receiving lateral acceleration, it is believed that it is easy to predict the position the vehicle will reach if it continues traveling at the current speed. This explains the steering model, which includes predictions for following a target curved course. It also states that it is necessary to speed up the response, such as by applying counter-steering, at the start of a curve, or to make appropriate changes taking into account the impact of changes in the curvature of the course, and proposes a steering model based on steering force, which includes secondary predictions of future position."

[0019] Regarding how to deal with steering hysteresis, Non-Patent Document 6 states that the following equation is used for the command angle to the steering motor.

number

[0020] However, this is based on the hysteresis being a parallelogram and the slope of the diagonal being the gear ratio, and is data obtained on a turning radius gauge, not data obtained during actual driving when lateral tire force is applied, nor is it based on the actual curvature of the route and steering angle.

[0021] Non-Patent Document 6 discusses the issue of balancing power performance and handling stability of automobiles, stating, "The problem of balancing power performance and handling stability arises because road friction decreases as the speed increases. This problem becomes more pronounced as the vehicle size increases, which is also specific to automobiles." Fig. 4 (attached to this application as Fig. 8) shows a distribution diagram of tire cornering power with respect to total vehicle weight, and states that this problem becomes more pronounced as the vehicle size increases.

[0022] As is clear from Figure 8 above, the tire cornering power relative to engine output is significantly smaller for large vehicles such as trucks than for passenger cars. In addition, there are changes in steering response due to changes in passenger numbers, load changes, and the resulting changes in total vehicle weight, vehicle center of gravity position, and tire characteristics, and in addition to responding to these changes, specific processing is required for large vehicles such as trucks when it comes to processing steering play in autonomous driving. [Prior art documents] [Patent documents]

[0023] [Patent Document 1] Patent No. 6202700 [Non-patent literature]

[0024] [Non-Patent Document 1] Fujio Momiyama et al.; Improvement of Straight-Line Driving Performance of Large Vehicles by Active Use of Lateral Force Compliance Steer (Part 2), Transactions of the Society of Automotive Engineers of Japan, Vol. 34, No. 2, April 2003, No. 20034187, pp. 89-94. [Non-patent document 2] Fujio Momiyama et al.; Improvement of Straight-Line Driving Performance of Large Vehicles by Active Use of Lateral Force Compliance Steer (First Report), Transactions of the Society of Automotive Engineers of Japan, Vol. 34, No. 2, April 2003, No. 20034187, pp. 83-88. [Non-patent document 3] Fujio Momiyama et al.; Cant Lane Keeping Analysis and Modeling of Autonomous Trucks, Transactions of the Society of Automotive Engineers of Japan, Vol. 45 No. 6 November 2014, No. 20144800, pp. 1027-1034. [Non-patent document 4] Masaichi Kondo, Basic Automotive Engineering, Second Edition, Yokendo Co., Ltd., page 21. [Non-Patent Document 5] Kenichi Yoshimoto: Simulation of the human-vehicle system using a steering model including prediction, Japan Society of Mechanical Engineers, Vol. 71, No. 596, pp. 13-18. [Non-patent document 6] Fujio Momiyama; Interest in and Technology of Dynamics of Heavy Vehicles, [No. 930-81] Proceedings of the 2nd Transportation and Logistics Conference of the Japan Society of Mechanical Engineers [1993-12.6-9. Kawasaki] pp.19-26. [Non-Patent Document 7] Lemniscate curve, Automotive Technology Handbook (Volume 7), Testing and Evaluation (Vehicles), p. 113, Society of Automotive Engineers of Japan Summary of the Invention [Problem to be solved by the invention]

[0025] None of the above prior art describes a solution for the counter-steering control law that addresses steering wheel play. The present invention is an autonomous vehicle equipped with an autonomous driving system that takes steering wheel play into consideration. The autonomous driving system acquires xy data at the actual vehicle site, with the steering wheel angle as the y value and the route curvature as the x value, identifies a control equation of y=f(x), and uses this control equation to achieve control that "determines a target destination point (movement target point), calculates (detects) the curvature of the target route to reach that point, and determines the steering wheel angle that follows that route curvature." [Means for solving the problem]

[0026] To solve the above problems, the present invention can detect the path traced by a wheel (tire) on the ground from the wheel speed of the wheel and calculate the radius of curvature of that path. The control system obtains a correlation equation between the curvature of the path (route curvature) and the steering angle at that time, and uses this as a control equation to follow the path. The path changes not only depending on the steering but also on the cross-slope (cant) of the road, so the influence of cant is also reflected in the curvature of the wheel path. Furthermore, the path also changes depending on changes in vehicle state variables such as the number of passengers and the total vehicle weight and center of gravity position associated with changes in load. In other words, this invention is configured to "control the path of a vehicle using a control equation that captures changes in the vehicle's path from wheel speed" based on the steering, cant, and load condition. [Effects of the Invention]

[0027] Conventional autonomous driving simply involves manipulating the steering wheel angle to ensure the vehicle follows a route. However, because there is play in the steering wheel, the actual steering is unstable, resulting in delayed response and a lack of precision. According to the inertial navigation method of the present invention, the trajectory is determined from the actual wheel speed, which is already affected by the steering wheel play, and the path radius curvature is determined from this trajectory, and the correlation equation between this path radius curvature and the steering wheel angle is obtained. Since the trajectory is the actual trajectory, the correlation equation between the path radius curvature and the steering wheel angle is not affected by the steering wheel play. [Brief explanation of the drawings]

[0028] [Figure 1] An explanatory diagram of the correlation between the steering angle (y) and the path curvature (x) and the method for obtaining the vehicle side slip angle (β) and the path curvature (x) [Figure 2] An explanatory diagram of how to calculate the correlation equation y=f(x) between the steering angle (y) and the path curvature (x) [Figure 3] An explanatory diagram of how to obtain the correlation equation between the steering angle (y) and the path curvature (x) in the actual work site, and how to correct the steering neutral position. [Figure 4]An explanatory diagram of how to obtain the correlation equation between the steering angle (y) and the path curvature (x) in the actual work site, and how to correct the steering neutral position. [Figure 5] An explanatory diagram of a method (control formula) for controlling the steering angle (y) using the correlation formula y=f(x). [Figure 6] Illustrative diagram of adding the curvature conversion value of the cant angle to the target route curvature [Figure 7] An explanatory diagram of how to determine the steering angle by finding the curvature of the path leading to the point of interest (target point of movement). [Figure 8] An explanatory diagram of a method for detecting a change in the position of the center of gravity of a vehicle. [Figure 9] FIG. 10 is a diagram showing the difference in tire output relative to engine output between a large vehicle and a passenger car, as described in Non-Patent Document 6. DETAILED DESCRIPTION OF THE INVENTION

[0029] To determine a target destination (movement target point), calculate the curvature of the target route to get there, and determine the steering angle to follow that route curvature, first, the relationship between the steering angle (y) and the route curvature (x) and the vehicle side slip angle are required.

[0030] FIG. 1 explains how to obtain the steering angle (y), path curvature (x), and vehicle side slip angle (β). The left side of the figure shows the process leading up to obtaining the correlation between the steering angle and the route curvature, and the right side shows an explanatory diagram corresponding to that process. This diagram shows the vehicle coordinates (x, y coordinates) on the map coordinate system (XY coordinates) with the vehicle's center of gravity as the origin. In Figure 1, (1) is the longitudinal axis X of the Earth coordinate system and the absolute coordinate system, (2) is the horizontal axis (Y) of the Earth coordinate system and the absolute coordinate system, (3) is the longitudinal axis of the vehicle coordinate system, (4) is the horizontal axis of the vehicle coordinate system, (5) is the steering angle (δh), and (6) is the left rear wheel speed (v rl ), (7) is the rear right wheel speed (v rr ), (8) is the longitudinal velocity (v x ), (9) is the rear wheelbase (Tr), (10) is the yaw rate (γ), (11) is the path curvature (current curvature) (ρ G), (12) is the path radius (rG), (13) is the distance between the rear axle and the center of gravity (lr), (14) is the lateral speed (vy), and (15) is the vehicle side slip angle (β).

[0031] Here, the steering angle (δ h ), rear axle left wheel speed (v rl ), rear axle right wheel speed (v rr ) can be acquired in real time from the vehicle LAN, and from these data, the steering angle (y), i.e., δ h and the path curvature x, i.e., ρ G Correlation is calculated in real time.

[0032] First, in step 1, the rear left wheel speed (v rl ) and rear right wheel speed (v rr ) and calculate the longitudinal velocity (v x ) is obtained.

[0033] Next, in step 2, the rear left wheel speed (v rl ) and rear right wheel speed (v rr ) difference between the rear wheelbase (T r ) to find the yaw rate (γ).

[0034] In step 3, the yaw rate (γ) is calculated based on the longitudinal velocity (v x ) to find the curvature (ρ G ) is obtained.

[0035] In step 4, the yaw rate (γ) is calculated based on the distance between the rear axle and the center of gravity (l r ) to find the lateral velocity (v y ) is calculated. Here, the distance between the rear axle and the center of gravity (l r ) will be described later with reference to FIG.

[0036] In step 5, the lateral velocity (v y ) to the longitudinal velocity (v x ) to find the vehicle side slip angle (β).

[0037] In step 6, the steering angle y, i.e., δ h and the path curvature x, i.e., 1 / rG That is, ρ G Obtain the correlation of When driving a lemniscate curve (see Reference 7), commonly known as a figure-eight turn, and carrying out steps 1 to 6 described above, the steering angle δ h Let y be the path curvature x, that is, 1 / r G That is, ρ G A correlation diagram can be obtained where x is the curvature of the figure eight. The lemniscate is a curve in which the curvature increases continuously from the central intersection of the figure eight, where the curvature is zero, to the apex of the circular arc, and then decreases continuously from that apex before returning to the central intersection. Therefore, the relationship between the steering angle (y) and the path curvature (x), y=f(x), is a linear relationship, y=ax. The graph (actual diagram) shows the equivalent curvature due to road cant and the compliance steer of the steering and suspension system (see References 1 and 2) superimposed on y=ax.

[0038] In step 7, the vehicle side slip angle (y, i.e., β) and the path curvature (x, i.e., ρ G This β is a variable when calculating the curvature of the path leading to the gaze point (movement target point) in FIG. 9, which will be described later.

[0039] We will now explain the diagram illustrating how to find the correlation equation y=f(x) between the steering angle (y) and the path curvature (x) in Figure 2. The upper right side of the diagram shows a correlation diagram between the steering angle (y) and the path curvature (x) obtained by performing steps 1 to 6 in Figure 1, and steps 1 to 3 are shown for applying the least squares method to the data in the correlation diagram to obtain the correlation equations (a), (b), and (c).

[0040] In step 1, the graph in the upper right of Figure 2 is obtained by running along the lemniscate curve and executing steps 1 to 6 in Figure 1. In step 2, steps (1), (2), and (3) shown at the bottom of Figure 2 are executed. That is, the correlation diagram data in the upper right graph is converted to absolute values, and y = ax + b' is obtained by the least squares method. In step 3, b is doubled from b' in y = ax + b',

[0041]

number

[0042] The upper right diagram can be interpreted as follows. In other words, it is data containing fluctuations in the direction of decreasing curvature relative to the straight line y = ax...(a) passing through the origin. Fluctuations in the direction of decreasing curvature are a phenomenon caused by the vehicle's understeer characteristics. The relationship between the current route curvature x and the current steering angle is captured, and it can be interpreted that turning at steering angle y will result in a curvature oscillation of Δx in curvature x. Therefore, by detecting the curvature of the target route in some way, the steering angle relative to that target curvature can be determined using this equation. The target route can be traced by using y = ax...(a) as the control equation for steering back, and y = ax + b...(b) and y = ax - b...(c) as the control equations for steering forward. In other words, equations (a), (b), and (c) in Figure 2 can be used as control equations.

[0043] A method for obtaining the correlation equation between the steering angle (y) and the path curvature (x) in the actual operation site and a method for correcting the steering neutral position will be explained below. The "method of finding the correlation equation y=f(x) between the steering angle (y) and the route curvature (x)" in Figure 2 requires an open space where the vehicle can make lemniscate turns. Instead of lemniscate turns, methods of finding these at the actual site where the vehicle is in operation, such as circular or right-angle turns, are explained below using the diagrams (1) to (5) in the bottom section of the figure, along with <Step 1> to <Step 6> shown in the top section of the figure.

[0044] <Process 1> Data group (1) is acquired by making a left steady partial circle turn, and data group (2) is acquired by making a right steady partial circle turn. (The data groups are data with a width in the curvature direction at a constant steering angle.)

[0045] <Process 2> Determine the coordinates of the maximum inflection point. Maximum absolute coordinate (x1, y1) of data group (1) Maximum absolute coordinate (x2, y2) of data group (2) Press down.

[0046] <Process 3> The maximum absolute coordinates (x1, y1) of data group (1) The maximum absolute coordinate (x2, y2) of data group (2) Obtain the equation of the straight line (a) connecting the two points.

number

[0047] <Step 4> The maximum absolute coordinate (x3, y1) of the data group (1) is obtained, and the c1 value is obtained, and the linear equation (b) is obtained.

number

[0048] <Process 5> The maximum absolute coordinate (x4, y2) of the data group (2) is determined to obtain the value of c1, and the linear equation (c) is obtained.

number

[0049] <Process 6> The b0 value in steps (3), (4), and (5) is the error in the neutral position of the steering wheel. Therefore, when calibration is performed to set y=b0 to the neutral position of the steering wheel, equations (a), (b), and (c) are rewritten as follows, which are used as control equations.

number

[0050] A method for obtaining the correlation equation y=f(x) between the steering angle (y) and the path curvature (x) in the actual operation site and a method for correcting the steering neutral position will be explained below. This diagram is composed of four boxes surrounded by dotted lines. The upper left box shows "symlink (A) for acquiring data group (1) of step (1) in Figure 3," the lower left box shows "symlink (B) for acquiring data group (2) of step (1) in Figure 3," the lower right box shows "the calculation process for acquiring the correlation equation between the steering angle (y) and the path curvature (x)," and the upper right box shows "the steering neutral position detection process."

[0051] The symlink (A) that acquires data group (1) takes in the path curvature x during a turn while the steering angle is held at y1, and calculates the maximum value x1, minimum value x3, and the difference c1 between the maximum and minimum values of x from the symlink composed of a Switch block and a Memory block.

[0052] The symlink (B) that acquires data group (2) takes in the path curvature x during a turn while the steering angle is held at y2, and calculates the maximum value x2, minimum value x4, and the difference c2 between the maximum and minimum values of x from the symlink composed of a Switch block and a Memory block.

[0053] In the calculation process to obtain the correlation equation between the steering angle (y) and the path curvature (x), the outputs from Symlink A and Symlink B are received, and the correlation equations for the return equation y=ax, the left cutting equation y=ax+b1, and the right cutting equation y=ax-b2 are generated and used for control. In the steering wheel neutral position detection process, a neutral error b0 of the steering wheel angle is detected, and the neutral position of the steering wheel angle is corrected.

[0054] The method (control formula) for controlling the steering angle (y) using the correlation equation y = f(x) in Figure 5 is explained below. How to use the control formulas (a), (b), and (c) obtained in Figure 2 is explained using <Step 1>, <Step 2>, and <Step 3> shown on the left side of the diagram. A symlink diagram is shown at the bottom.

[0055] <Step 1> Compare the current curvature with the target curvature. If the target curvature is greater and the target curvature x is "positive", calculate the control handle angle y using the cutting equation y=ax+b···(b). <Step 2> Compare the current curvature with the target curvature. If the target curvature is greater and the target curvature x is negative, calculate the control handle angle y using the cutting equation y=ax-b···(c). <Step 3> The absolute values of the current curvature and the target curvature are compared, and if the target curvature is smaller, the control handle angle y is calculated by y=ax...(a).

[0056] We will now explain the diagram of adding the curvature conversion value of the cant angle to the target curvature in Figure 6. The right frame of the figure shows a photograph of a truck traveling on a bank (a curved section with a cant) on a circular test course. On a canted curve, the centrifugal acceleration resulting from the relationship between the lane curvature and the vehicle speed balances with the centripetal acceleration obtained by multiplying the acceleration of gravity by the cant angle, resulting in a turn.

[0057] If the vehicle speed is low, the vehicle will drift down the bank (downhill), and if the vehicle speed is high, it will drift up the bank (uphill). Driving down the bank can be suppressed by steering upward on the bank, and driving up the bank can be suppressed by steering downward on the bank. This is what is known as "steering." The relationship is shown in equation (d). In other words, the cant angle is the sum of the "measured lateral acceleration" and the "value obtained by dividing the product of the vehicle speed and yaw rate by the acceleration due to gravity" or the "curvature of the curved section" multiplied by the "vehicle speed squared" and divided by the acceleration due to gravity. The "-" sign is used to match the acceleration output in SAE coordinates with the equation written in ISO coordinates.

[0058] Expressing equation (d) using symmetry to find the cant angle and multiplying it by the acceleration of gravity gives the lateral acceleration due to cant, and dividing this by the square of the vehicle speed gives the cant equivalent curvature. Calculating the difference between this cant equivalent curvature and the path curvature calculated from Figure 1 above gives the curvature equivalent to "counter steering." Substituting this into equation (a) above to calculate the "counter steering" steering angle (y), and by controlling the counter steering, the path can be traced.

[0059] The method of determining the steering angle by finding the curvature of the path leading to the point of interest (movement target point) in Figure 7 will be explained. A visual sensor is provided at the front of the vehicle described in Figure 1 above. The visual sensor detects the point of interest (P T The target route is followed by determining the target point (movement target point), detecting the lateral deviation (e2) and angular deviation or azimuth angle (e3) of the target route at that position, and reaching the target point (movement target point). This process is explained in steps 1 to 5 shown on the left side of the figure.

[0060] In step 1, the velocity at the center of gravity of the vehicle (v x ) and sideslip angle (β) can be constantly recognized by the above-mentioned Figure 1. In step 2, the coordinates (x1, y1) and direction (e3) of the gaze point (movement target point) are detected by a visual sensor such as a camera or a lidar. In step 3, the current rotation center (O now ) on the extension line of the gaze point P t From the angle deviation e3, the side (P t- O target ) and the side (Cg) extending perpendicular to the side slip angle (β) from the center of gravity (Cg - O target ) and determine the isosceles triangle ΔO yaw P T ,P G Decide. In step 4, the isosceles triangle ΔO yaw P T ,P G Isosceles of (P t- O target ), (Cg - O target ) to the target radius (r t ), target curvature (ρ G ) is found. In step 5, the curvature obtained by adding the "curvature equivalent to the applied rudder" in Figure 6 to the path curvature in step 4 is applied to the step in Figure 5 to determine the control handle angle.

[0061] Point of interest (P t) is set to a position several meters ahead, but taking into consideration the need to anticipate changes in the curvature of the route and capture the area ahead, the initial value is set to counter rudder left cant if on a straight or left-hand turning route, and right cant if turning right, and control is considered to replace this with the measured cant.

[0062] The method for detecting changes in the position of the center of gravity of the vehicle is explained in the diagram of Figure 8. The upper left side of the diagram shows a plan view of the vehicle, the lower left side shows a side view of the vehicle, and the right side shows an air piping diagram of the pneumatic suspension of the rear axle. The change in air pressure (Δp L , Δp R ) to rear axle load change (ΔN r ) and calculate the axle load N after the vehicle is empty. Er Add to the loaded axle load N Lr The detected rear axle load change is subtracted from the change in vehicle weight to calculate the change in front axle load, and this is added to the unladen front axle load to obtain the loaded front axle load. The wheelbase (l) is multiplied by the ratio of rear axle load to vehicle weight to obtain the distance from the front axle to the center of gravity (l when unladen). Ef and when loaded Lf ) can be seen. Lf and when empty Ef The change in the center of gravity can be determined from the difference.

[0063]

number

[0064] Here, Δp L is the change in air pressure due to the load change on the left air spring, Δp R is the change in air pressure due to a change in load on the right air spring, ΔNr is the change in rear axle load, ΔN f is the change in front axle load, N Er is the rear axle empty load, N Ef is the front axle empty load, N Lr is the rear axle load, N Lf is the front axle load, GVW is the gross vehicle weight (weight when loaded), VW is the vehicle weight (weight when unladen), l is the wheelbase, l Efis the distance from the front axle to the center of gravity of the empty vehicle, l Lf is the distance from the front axle to the center of gravity of the loaded vehicle, and Δl is the distance the center of gravity changes due to the loaded vehicle.

[0065] We will explain the air piping diagram for the rear axle air suspension on the right side of Figure 8. In the diagram, R1RF refers to the right front air spring on the first rear axle, R1RR refers to the right rear air spring on the first rear axle, R1LF refers to the left front air spring on the first rear axle, R1LR refers to the left rear air spring on the first rear axle, R2RF refers to the right front air spring on the second rear axle, R2RR refers to the right front and rear air springs on the second rear axle, R2LF refers to the left front air spring on the second rear axle, and R2LR refers to the left rear air spring on the second rear axle.

[0066] The right front and rear air pipes of the first rear axle and the right front and rear air pipes of the second rear axle are connected via a joint, and similarly the left front and rear air pipes of the first rear axle and the left front and rear air pipes of the second rear axle are connected via a joint, and air is supplied from the rear air tank to each of the right and left connecting pipes via a rear axle height adjustment solenoid valve. The rear axle height adjustment solenoid valve receives a vehicle height signal from a vehicle height stroke sensor (not shown) installed on the right and left sides of the first or second rear axle, and supplies air from the rear air tank to the right and left connecting pipes to set the vehicle height to a set value. An air pressure sensor is connected to each of the right and left connecting pipes to detect changes in air pressure (Δp L ,Δp R ) and detects the change in center of gravity position due to the change in weight using the above equations (11) to (15).

Claims

1. This is an inertial navigation system that controls the course of a vehicle by capturing the path traced by the wheels (tires) on the road surface using the wheel speed, obtaining the path curvature and sideslip angle of the vehicle's center of gravity, and corresponding to the steering play, according to a "path curvature based on wheel speed and steering angle control formula for turning in and turning back"; Rear axle left wheel speed (v ｒl ) and rear right wheel speed (v ｒｒ ) and calculate the longitudinal velocity (v ｘ ) and then calculate the rear left wheel speed (v ｒl ) and rear right wheel speed (v ｒｒ ) difference between the rear wheelbase (T ｒ ) to obtain the yaw rate (γ), and then use that yaw rate (γ) to calculate the longitudinal speed (v ｘ ) to find the path curvature (ρ Ｇ ) in search of Steering wheel angle and route curvature (ρ Ｇ ) is a variable in the cutting and return control formula, The yaw rate (γ) is divided by the distance between the rear axle and the center of gravity (lr) to obtain the lateral speed (v ｙ ) and calculate the lateral velocity (v y ) to the longitudinal speed (v ｘ ) to determine the vehicle side slip angle (β), and then a visual sensor is used to determine the target route curvature to the forward gaze point. The target route curvature is then added to the curvature equivalent to cant detected by the lateral acceleration sensor and yaw rate sensor, and the steering angle is controlled using a control formula based on the route curvature and steering angle determined by the wheel speed.

2. In the inertial navigation system according to claim 1, the yaw rate (γ) is divided by the distance between the rear axle and the center of gravity (lr) to calculate the lateral velocity (v ｙ ) and calculate the lateral velocity (v y ) to the longitudinal speed (v ｘ ) to detect the vehicle side slip angle (β).

3. 2. The inertial navigation system according to claim 1, wherein a correlation equation corresponding to a plastic characteristic (so-called steering play) of a propagation process from a steering angle to a wheel speed is derived.

4. 2. The inertial navigation system according to claim 1, wherein air pressure sensors are provided in the air circuit of the right air spring and the air circuit of the left air spring of the rear air spring, and the position of the center of gravity is detected from the detected air pressures.

5. 2. The inertial navigation system according to claim 1, further comprising: a calculation step for obtaining a correlation equation between a steering angle (y) and a route curvature (x); and a steering neutral position detection step.

Images