A seedling avoidance and steering method for paddy field weeders based on satellite and vision navigation

Abstract

The present invention discloses a seedling avoidance and steering method for a paddy field weeder based on satellite and visual navigation. During the steering process, the weeder may crush some seedlings, thereby affecting the growth of crops. The present invention first selects the optimal steering method, then obtains the position information of the seedlings in the world coordinate system, adopts the artificial potential field method, establishes a gravitational potential field based on the coordinate points that will crush the seedlings, and establishes a repulsive potential field based on the position of the seedlings to calculate the resultant force. A new path is planned based on the resultant force, and the original path coordinate points that are judged to crush the seedlings are replaced with these new path coordinate points to complete the local fine-tuning of the original path; then, the preview distance of the paddy field weeder in the current posture is calculated, and based on the geometric relationship between the new path and the position and posture of the paddy field weeder, an improved pure tracking algorithm is used to calculate the target turning angle of the front wheel of the paddy field weeder, so that the paddy field weeder can track the new path when turning at the ridge head.

Description

Technical Field

[0001] The invention belongs to the field of agricultural machinery, and in particular relates to a seedling avoidance and steering method for a paddy field weeder based on satellite and visual navigation. Background Art

[0002] With the continuous development of agricultural robotics, robots are increasingly being used in agricultural production, such as unmanned tractors and automated pesticide sprayers. The application of satellite-based vision-based navigation methods can enhance the intelligence of agricultural robots, thereby better meeting the needs of agricultural production. Paddy field environments are complex, and single-sensor navigation methods are susceptible to external influences and have poor navigation accuracy. Therefore, most field navigation methods use multi-sensor fusion to achieve autonomous navigation. Paddy field weeders frequently need to turn the ridge head during field operations. Currently, there are two mainstream methods for turning the ridge head: reserving space at the ridge head for the weeder to turn, which wastes a lot of paddy field space; and turning directly in the field. While this method saves space, the weeder inevitably crushes some seedlings during the turning process, affecting crop growth. Therefore, ridge head turning is a very important technical issue for paddy field weeders and requires improvement. Summary of the Invention

[0003] The purpose of the present invention is to address the problems existing in the field operation of unmanned paddy field weeders and propose a seedling avoidance and steering method for paddy field weeders based on satellite and visual navigation to ensure the efficiency and quality of the operation.

[0004] To achieve the above object, the present invention adopts the following technical solutions:

[0005] The present invention provides a method for avoiding seedlings and steering a paddy field weeder based on satellite and visual navigation, which is specifically as follows:

[0006] Step 1. First, based on the comparison between the minimum steering radius of the paddy field weeder and half of the steering width, eliminate inappropriate steering modes from several steering modes, including U-shaped steering, smooth steering, bulb-shaped steering, hook-shaped steering and fishtail-shaped steering; then, use the hierarchical analysis method to establish a hierarchical structure model and select the optimal steering mode.

[0007] Step 2: Use the visual navigation information of the industrial camera to obtain the position information of the seedlings in the visual image. At the same time, according to the installation position and posture information of the industrial camera, the mapping relationship between the camera coordinate system and the world coordinate system is obtained to obtain the position information of the seedlings in the world coordinate system.

[0008] Step three, make the following plans for each coordinate point on the path of the optimal steering method, starting from the first coordinate point to the preset number of coordinate points thereafter: select the seedlings within the radius range of ρ0 with the coordinate point as the center, and calculate that when the distance between the coordinate point and the selected i-th seedling position point is less than the threshold value L1, it is considered that the rear wheel of the paddy field weeder will crush the i-th seedling when it continues to travel along the original path to the coordinate point; from the first coordinate point that is judged to crush the seedling to the last coordinate point that will crush the seedling, re-plan the new path according to the resultant force calculated by the artificial potential field method, and each coordinate point on the original path that will crush the seedling is replaced by a coordinate point on the new path that will not crush the seedling. When the rear wheel of the paddy field weeder is traveling along the re-planned new path, the above-mentioned planning is continuously performed on the unplanned coordinate points on the new path, thereby continuously updating the subsequent path, so that the paddy field weeder can travel along the real-time updated path to achieve seedling avoidance and steering; wherein the seedling influence radius W s is the row spacing of seedlings, L s The spacing between seedlings.

[0009] Preferably, a hierarchical structure model is established by using the hierarchical analysis method to select the optimal steering method, specifically: ① Design the target layer, the intermediate layer and the scheme layer. The target layer is the selection of the optimal steering method. The intermediate layer is the factors that need to be considered in selecting the optimal steering method, including the steering seedling pressing amount A1, the steering time A2, the lateral offset A3 of the steering final position and the heading angle A4 of the steering final position. The scheme layer includes various steering methods left after excluding inappropriate steering methods; ② Construct a judgment matrix A = (a ij ) 4×4 , a ij for a i Factor and a j The relative importance of factors to the target is divided into five levels: equally important, slightly important, obviously important, strongly important, and extremely important. The judgment matrix A is tested for consistency, and the eigenvector corresponding to the maximum eigenvalue of the judgment matrix A is calculated and normalized to obtain the weight of each factor to the target W = [w1w2w3w4], w i , i=1,2,3,4 is the weight of the i-th factor to the target; ③ Construct the judgment matrix of each plan relative to each factor n is the number of options, and each judgment matrix Through consistency detection, calculate each judgment matrix The eigenvector corresponding to the maximum eigenvalue is normalized to obtain the weight of each scheme on the factor c ki, i=1,2,3,4 is the weight of the k-th solution on the i-th factor, k=1,2,…,n;④ Perform the hierarchical total sorting consistency test. If the hierarchical total sorting consistency test fails, modify the judgment matrix A and the judgment matrix Re-execute steps ② and ③, and re-test the hierarchical total order consistency until the test passes. Calculate the total weight matrix and select the largest weight d in the total weight matrix. k The corresponding solution is the optimal steering method.

[0010] Preferably, the paddy field weeder is simplified into a bicycle model, and the calculation formula for the turning radius of the paddy field weeder is R=L / tanδ, where R is the turning radius of the paddy field weeder, L is the wheelbase between the front and rear wheels at the bottom of the paddy field weeder, and δ is the target turning angle of the front wheel; when the target turning angle of the front wheel is the largest, the minimum turning radius of the paddy field weeder is calculated.

[0011] More preferably, the consistency index CI is defined as (λ-m) / (m-1), where λ is the maximum eigenvalue of the judgment matrix A and m is the order of the judgment matrix; the random consistency index RI is introduced, the average random consistency index table is checked, the 4th-order judgment matrix RI is determined to be 0.90, and the consistency ratio CR is defined as CI / RI. When the consistency ratio CR is <0.1, it is considered that the inconsistency degree of A is within the allowable range.

[0012] More preferably, in the hierarchical total order consistency test, the hierarchical total order consistency ratio is set Where CI i is the consistency index of the solution layer relative to the i-th factor in the factor layer, RI i is the corresponding random consistency index, if the hierarchical total ranking consistency ratio CR T <0.1, the total hierarchical order consistency is acceptable. The total weight matrix D = [d1, d2, ..., d n ], d k represents the weight of the k-th solution to the target, Select the largest weight d in the total weight matrix k The corresponding solution is the optimal steering method.

[0013] Preferably, step 2 is as follows:

[0014] The visual images collected by the industrial camera are converted to grayscale using 2G-RB color images and then binarized using Ostu. Morphological operations are then performed on the binary images. The radius range is set, and the minEnclosingCircle function is used to find the minimum circumscribed circle that encloses the seedlings in the image. The center of the circle is the position coordinate P of the seedling in the image. uv ; At the same time, the intrinsic parameter matrix K of the industrial camera is obtained according to the industrial camera calibration m, the external parameter matrix R of the industrial camera is obtained from the installation position and posture information of the industrial camera m , then the mapping relationship between the camera coordinate system and the world coordinate system is P uv =K m R m P W According to this relationship, the position coordinates P of the seedling in the world coordinate system can be obtained W .

[0015] Preferably, the artificial potential field method is used and the resultant force is calculated as follows:

[0016]

[0017] Among them, k0 is the gravitational gain coefficient, k i is the repulsion gain coefficient of the selected i-th seedling, n1 is the number of repulsions, F att (q) is the gravitational force obtained by establishing the gravitational potential field based on the coordinate points on the path, is the repulsive force obtained by establishing the repulsive potential field based on the selected position of the i-th seedling;

[0018] Assume that the rear wheel position of the paddy field weeder is q(x,y), and the coordinate point on the path that will crush the seedlings is q g (g x ,g y ), gravitational potential field U att (q) and the gravitational force F att The expressions of (q) are:

[0019]

[0020] Among them, η is the proportional gain coefficient, ρ(q,q g ) is the Euclidean distance between the rear wheel driving position of the paddy field weeder and the coordinate point on the path that will crush the seedlings, and the direction is defined as the direction from the rear wheel driving position of the paddy field weeder to the coordinate point on the path.

[0021] The repulsive potential field U established based on the selected seedling position rep The expression is:

[0022]

[0023] γ is the proportional gain coefficient, q0 is the seedling position point, and the corresponding repulsive force is the negative gradient of the repulsive potential field, expressed as:

[0024]

[0025] Preferably, when the rear wheels of the paddy field weeder travel along the re-planned new path, the satellite navigation information of the dual satellite antennas is used to obtain the position information and posture information of the paddy field weeder, and the angle sensor at the front wheel is used to obtain the current turning angle information of the front wheel. According to the geometric relationship between the re-planned new path and the position and posture of the paddy field weeder, an improved pure tracking algorithm is used to calculate the target turning angle of the front wheel of the paddy field weeder, and the current turning angle of the front wheel is subtracted from the front wheel target turning angle δ to obtain the required turning angle of the front wheel, wherein the required turning angle of the front wheel is positive when it indicates a right turn, and negative when it indicates a left turn. The four-wheel steering mechanism is driven by a motor to drive the front wheel to rotate the required turning angle of the front wheel, or to drive the front wheel to rotate half of the required turning angle of the front wheel, and the rear wheel rotates an angle of the same size and opposite direction as the front wheel, so that the paddy field weeder can track the re-planned new path when turning at the ridge head.

[0026] More preferably, the target turning angle of the front wheel of the paddy field weeder is calculated as follows:

[0027] Take the rear wheel driving position as the tangent point and the longitudinal direction of the paddy field weeder as the tangent line, according to the preview distance L d Get the preview point on the new path after replanning (g x ,g y ), the following formula is derived using the law of sine:

[0028]

[0029] In formula (1), α is the heading angle, which is the current posture of the paddy field weeder and the preview point (g x ,g y ) and the angle between the line connecting the rear wheel driving position points; preview distance L d =0.55v+L0, where L0 is the initial value of the preview distance;

[0030] Simplified:

[0031]

[0032] Substituting formula (2) into R = L / tanδ, we can obtain the target turning angle of the front wheel of the paddy field weeder:

[0033]

[0034] L d =0.55v+L0 Substituting into formula (3), we get:

[0035]

[0036] Preferably, the four-wheel steering mechanism is composed of a front steering mechanism and a rear steering mechanism with the same structure and symmetrical arrangement; the front steering mechanism is composed of a front wheel, a front axle, a front steering knuckle, a front steering knuckle arm, a front steering ball head and a front connecting rod; the cylinder body of the front steering oil cylinder is fixed to the bottom of the lawn mower body through the front axle, and the piston rod of the front steering oil cylinder is hinged through the front steering ball head and one end of the front connecting rod and one end of a front steering knuckle arm; the other end of the front connecting rod is hinged to one end of the other front steering knuckle arm; the other ends of the two front steering knuckle arms are respectively connected through the front steering knuckle It is hinged to a front wheel; the rear steering mechanism consists of a rear wheel, a rear axle, a rear steering knuckle, a rear steering knuckle arm, a rear steering ball head and a rear connecting rod; the cylinder body of the rear steering cylinder is fixed to the bottom of the lawn mower body through the rear axle, and the piston rod of the rear steering cylinder is hinged through the rear steering ball head and one end of the rear connecting rod and one end of a rear steering knuckle arm; the other end of the rear connecting rod is hinged to one end of the other rear steering knuckle arm; the other ends of the two rear steering knuckle arms are respectively hinged to a rear wheel through the rear steering knuckle; the front steering cylinder and the rear steering cylinder are connected through a hydraulic system.

[0037] The hydraulic system consists of a fuel tank, a filter, a hydraulic pump, an electric motor, a fully hydraulic steering gear, a two-position, four-way solenoid valve, a relief valve, and a cooler. The fully hydraulic steering gear and the two-position, four-way solenoid valve are both controlled by an industrial computer. The oil outlet of the fuel tank is connected to the oil inlet of the filter, which is in turn connected to the oil inlet of the hydraulic pump. The input shaft of the hydraulic pump is connected to the output shaft of the electric motor. The oil outlet of the hydraulic pump is connected to the oil inlet of the fully hydraulic steering gear. The two oil outlets of the fully hydraulic steering gear are connected to the rodless chamber of the front steering cylinder and the oil return port of the two-position, four-way solenoid valve. The rod chamber of the front steering cylinder is connected to the oil inlet of the two-position, four-way solenoid valve. The two oil outlets of the two-position, four-way solenoid valve are connected to the rodless chamber and the rod chamber of the rear steering cylinder. The oil return port of the fully hydraulic steering gear is connected to the oil inlet of the relief valve, and the oil outlet of the relief valve is connected to the oil return port of the fuel tank via a cooler.

[0038] The beneficial effects of the present invention are as follows:

[0039] 1. The present invention proposes a method for optimizing the seedling avoidance steering mode. By comparing the minimum steering radius of the paddy field weeder with half of the steering spacing, some inappropriate steering modes can be eliminated. The weights of the remaining steering modes are calculated using the hierarchical analysis method, and the scheme with the largest weight is selected as the optimal ridge head steering mode.

[0040] 2. This invention improves the traditional pure tracking algorithm and simulates the preview distance L when manually driving a vehicle. d It will change with the vehicle speed v. The preview distance is expressed as: L d =0.55v+L0, making the steering process smoother and the tracking effect more precise.

[0041] 3. The present invention proposes a method for obtaining the actual position of seedlings based on image processing.

[0042] 4. The present invention proposes a method for establishing an artificial potential field based on the position of the seedlings in the visual recognition image, and fine-tuning the steering trajectory according to the resultant force to achieve seedling-avoiding steering.

[0043] 5. The present invention uses satellite navigation to globally plan the working path of the paddy field weeder when the ridge head is turned, and uses visual navigation to fine-tune the working path of the paddy field weeder when the ridge head is turned, so that the weeder can turn along the satellite navigation path and at the same time fine-tune the turning trajectory according to visual information to achieve seedling-avoiding turning.

[0044] 6. The present invention proposes two steering methods: front-wheel steering only and four-wheel counter-steering through a hydraulic system. Different steering methods can be selected according to needs. BRIEF DESCRIPTION OF THE DRAWINGS

[0045] Figure 1 This is a three-dimensional diagram of the overall structure of the paddy field weeder used in the present invention;

[0046] Figure 2 It is a path diagram of various steering modes of the present invention;

[0047] Figure 3 It is a hierarchical structure diagram of selecting the optimal turning path by using the hierarchical analysis method in the present invention;

[0048] Figure 4 It is an analysis flow chart of the analytic hierarchy process in the present invention;

[0049] Figure 5 This is a schematic diagram of the steering of the bicycle model adopted by the present invention;

[0050] Figure 6 It is a geometric relationship model diagram of pure tracking algorithm;

[0051] Figure 7 This is the result of the preprocessing (grayscale, Ostu binarization, and morphological operation) of the seedling image by the present invention;

[0052] Figure 8 This is the result diagram of the present invention using the minimum circumscribed circle to wrap the seedlings;

[0053] Figure 9 This is a schematic diagram of a model for generating a new path for avoiding seedlings and turning by using an artificial potential field method in the present invention;

[0054] Figure 10 It is a geometric diagram of solving the seedling influence radius of the present invention;

[0055] Figure 11 is a flow chart of the method of the present invention;

[0056] Figure 12 It is a structural diagram of the four-wheel steering mechanism of the present invention;

[0057] Figure 13 It is a schematic diagram of the hydraulic system of the present invention. DETAILED DESCRIPTION

[0058] The present invention will be further described below with reference to the accompanying drawings.

[0059] like Figure 1 As shown, the paddy field weeder used in the present invention includes a weeder body, an industrial camera 1, dual satellite antennas 2, an angle sensor, a display 4, and an industrial computer 5. The weeder body adopts the structure described in the prior art, including a weeder body, wheels, a weeding mechanism, and a four-wheel steering mechanism 3, which drives the wheels. The industrial camera 1, dual satellite antennas 2, and display 4 are all mounted on the weeder body via brackets, with the industrial camera 1 located directly in front of the weeder body. An angle sensor is mounted on the front wheel. The four-wheel steering mechanism 3 is driven by a motor, and the wheels are driven by another motor, allowing rear-wheel drive. The input of the display 4 is connected to the industrial computer 5, and the signal outputs of the dual satellite antennas 2, the industrial camera 1, and the angle sensor are all connected to the industrial computer 5. The motor is controlled by the industrial computer 5. The industrial computer 5 receives satellite navigation information from the dual satellite antennas 2, visual navigation information from the industrial camera 1, and the current front wheel angle information from the angle sensor, and controls the motor to drive the four-wheel steering mechanism, thereby steering the wheels.

[0060] The present invention discloses a seedling avoidance and steering method for a paddy field weeder based on satellite and visual navigation. The method first selects the optimal steering mode through an industrial computer, obtains the position information of the seedlings in the visual image by using the visual navigation information of the industrial camera, and simultaneously obtains the mapping relationship between the camera coordinate system and the world coordinate system according to the installation position and posture information of the industrial camera, thereby obtaining the position information of the seedlings in the world coordinate system; judges whether the seedlings will be crushed according to the distance between the coordinate point on the path of the optimal steering mode and the seedling position point within the seedling influence radius; if the distance is less than a threshold value L1, it is considered that the seedlings will be crushed, and an artificial potential field method is used to establish a coordinate point based on the coordinate point that is considered to be crushed. Based on the gravitational potential field, a repulsive potential field is established according to the position of the seedlings, and the resultant force is calculated. A new path that can avoid the seedlings is planned based on the resultant force. This new path is also composed of a series of discrete coordinate points. These new path coordinate points replace the original path coordinate points that are judged to crush the seedlings, completing the local fine-tuning of the original path. The satellite navigation information of the dual satellite antennas is then used to obtain the position and posture information of the paddy field weeder, calculate the preview distance of the paddy field weeder in the current posture, and based on the geometric relationship between the new path and the position and posture of the paddy field weeder, use the improved pure tracking algorithm to calculate the target turning angle of the front wheel of the paddy field weeder, so that the paddy field weeder can track the new path when turning at the ridge head. The details are as follows:

[0061] Step 1: Select the optimal steering method through the industrial computer, as follows:

[0062] like Figure 2 、 Figure 3 、 Figure 4 and Figure 11As shown, the steering modes include U-shaped steering, smooth steering, bulb-shaped steering, hook-shaped steering and fishtail-shaped steering. The path of the U-shaped steering includes two straight line segments with a spacing of the steering width w and a semicircle segment tangent to the two straight line segments and with a radius of the minimum steering radius r1 of the paddy field weeder; the path of the smooth steering includes two straight line segments with a spacing of the steering width w, two quarter-circle segments with a radius of the minimum steering radius r1 of the paddy field weeder connected between the two straight line segments, and a straight line segment connecting the two quarter-circle segments; the path of the bulb-shaped steering includes two straight line segments with a spacing of the steering width w, two straight line segments The path of the hook-shaped turn includes two straight line segments with a spacing of the turning width w, a quarter-circle segment with a radius of the minimum turning radius r1 of the paddy field weeder and a circle segment with a central angle of θ1, which is tangent to the two quarter-circle segments; the path of the fishtail turn includes two straight line segments with a spacing of the turning width w, a quarter-circle segment with a radius of the minimum turning radius r1 of the paddy field weeder and a circle segment tangent to the quarter-circle segment and the other straight line segment; the path of the fishtail turn includes two straight line segments with a spacing of the turning width w, a radius tangent to the two straight line segments, and a circle segment tangent to the quarter-circle segment and the other straight line segment. The minimum turning radius r1 of the paddy field weeder and the two intersecting quarter circle segments and the straight line segment connecting the two quarter circle segments are used. First, based on the comparison between the minimum turning radius r1 of the paddy field weeder and half of the turning width w, inappropriate turning modes are eliminated. For example, U-turn requires the minimum turning radius r1 of the paddy field weeder to be equal to half of the turning spacing w, and smooth turn requires the minimum turning radius r1 of the paddy field weeder to be less than half of the turning spacing w. Then, the hierarchical structure model is established using the hierarchical analysis method to select the optimal turning mode. Specifically, the decision goal, consideration factors (decision criteria) and The decision-making scheme is designed as the highest layer T, the middle layer A and the lowest layer B according to the mutual relationship. The highest layer is the target layer, which is the selection of the optimal steering method. The middle layer A is the factors that need to be considered in selecting the optimal steering method, including the steering seedling pressing amount A1, the steering time A2, the lateral offset distance A3 of the steering final position and the heading angle A4 of the steering final position. The lowest layer B is the scheme layer, which includes various steering methods remaining after excluding unsuitable steering methods (for example, the bulb-shaped steering B1, the hook-shaped steering B2 and the fishtail-shaped steering B3 are left in this embodiment); the relative importance judgment matrix of each factor to the target is constructed A=(a ij ) 4×4 , a ij for a i Factor and a j The relative importance of factors to the goal is divided into five levels: equally important, slightly important, obviously important, strongly important, and extremely important (corresponding scale values ​​are 1, 3, 5, 7, and 9). ij >0、a ij =1 / a ji 、aii =1; In order to prevent the problem from being analyzed by the hierarchical analysis method, a consistency test is required. If the consistency test fails, the judgment matrix is ​​modified. After the consistency test, the eigenvector corresponding to the maximum eigenvalue of the judgment matrix A is calculated and normalized to obtain the weight of each factor on the target W = [w1 w2 w3 w4], w i , i=1,2,3,4 is the weight of the i-th factor to the target; then, construct the judgment matrix of each scheme relative to each factor n is the number of solutions, and consistency testing is performed on them. After consistency testing, each judgment matrix is ​​calculated The eigenvector corresponding to the maximum eigenvalue is normalized to obtain the weight of each scheme on the factor c ki , i=1,2,3,4 is the weight of the kth solution for the ith factor, k=1,2,…,n; finally, the hierarchical total sorting consistency test is performed. After the test passes, the total weight matrix is ​​calculated and the weight d with the largest weight in the total weight matrix is ​​selected. k The corresponding solution is the optimal steering method.

[0063] Among them, such as Figure 5 、 Figure 6 As shown in the figure, the paddy field weeder is simplified into a bicycle model, and the calculation formula for the turning radius of the paddy field weeder is R = L / tanδ, where R is the turning radius of the paddy field weeder, L is the wheelbase between the front and rear wheels at the bottom of the paddy field weeder, and δ is the target turning angle of the front wheel. When the target turning angle of the front wheel is the largest, the minimum turning radius r1 of the paddy field weeder is calculated.

[0064] Among them, the consistency index CI is defined as (λ-m) / (m-1), λ is the maximum eigenvalue of the judgment matrix A, m is the order of the judgment matrix, when CI=0, it indicates complete consistency, the closer CI is to 0, the better the consistency, and the larger the CI, the worse the consistency; in order to better measure the size of CI, the random consistency index RI is introduced, and the fourth-order judgment matrix RI=0.90 is determined by looking up the table (average random consistency index table), and the consistency ratio CR=CI / RI is defined. Generally, when the consistency ratio CR<0.1, it is considered that the inconsistency degree of A is within the allowable range, there is satisfactory consistency, and the consistency test passes; otherwise, the judgment matrix is ​​modified to meet the consistency requirements.

[0065] Among them, in the hierarchical total sorting consistency test, the hierarchical total sorting consistency ratio is set Where CI i is the consistency index of the solution layer relative to the i-th factor in the factor layer, RI i is the corresponding random consistency index, if the hierarchical total ranking consistency ratio CR T<0.1, indicating that the consistency of the total hierarchical ranking is acceptable. The total weight (the weight of the solution layer relative to the target layer) is obtained by combining the weights of the intermediate layer relative to the target layer and the solution layer relative to the intermediate layer. The total weight matrix D = [d1, d2, ..., d n ], d k represents the weight of the k-th solution to the target, Select the largest weight d in the total weight matrix k The corresponding solution is the optimal steering method. In this embodiment, after eliminating the inappropriate steering methods, three steering methods are left. The weight calculation of the three steering methods to the target is shown in Table 1.

[0066] Table 1 Total weight of steering mode to target

[0067]

[0068] Step 2: Use the visual navigation information of the industrial camera to obtain the position information of the seedlings in the visual image. At the same time, according to the installation position and posture information of the industrial camera, the mapping relationship between the camera coordinate system and the world coordinate system is obtained, thereby obtaining the position information of the seedlings in the world coordinate system, as follows:

[0069] The visual images collected by the industrial camera are converted to grayscale using 2G-RB color images and then binarized using Ostu. Then, morphological operations are performed on the binary images. The results are as follows: Figure 7 As shown in the figure, set the radius range and use the minEnclosingCircle function to find the minimum circumscribed circle that encloses the seedlings in the image. The center of the circle is the position coordinate P of the seedling in the image. uv ,like Figure 8 As shown; at the same time, the intrinsic parameter matrix K of the industrial camera is obtained according to the industrial camera calibration m , the external parameter matrix R of the industrial camera is obtained from the installation position and posture information of the industrial camera m , then the mapping relationship between the camera coordinate system and the world coordinate system is P uv =K m R m P W According to this relationship, the position coordinates P of the seedling in the world coordinate system can be obtained W .

[0070] Step 3. The path of the optimal steering mode is composed of a series of discrete coordinate points; the following planning is performed on each coordinate point on the path of the optimal steering mode, starting from the first coordinate point to a preset number of coordinate points thereafter: the seedlings are selected within a radius range with the coordinate point as the center and ρ0 as the radius, and when the distance between the coordinate point and the selected i-th seedling position point is calculated to be less than the threshold L1, it is considered that the rear wheel of the paddy field weeder will crush the i-th seedling when continuing to travel along the original path to the coordinate point; starting from the first coordinate point that is judged to be likely to crush the seedling to the last coordinate point that is likely to crush the seedling, a new path is replanned based on the resultant force calculated by the artificial potential field method, and each coordinate point on the original path that is likely to crush the seedling is replaced by a coordinate point on the new path that is not likely to crush the seedling. In this way, when the rear wheel of the paddy field weeder drives along the newly planned path, it can achieve seedling avoidance and steering from the first coordinate point to the preset number of coordinate points. In addition, during the driving process along the new path, the above planning is continuously performed on the unplanned coordinate points on the new path, thereby continuously updating the subsequent path, so that the paddy field weeder can achieve the purpose of seedling avoidance and steering along the real-time updated path. W s is the row spacing of seedlings, L s is the spacing between seedlings, such as Figure 10 shown.

[0071] like Figure 9 As shown, the artificial potential field method is used and the calculated resultant force is as follows:

[0072]

[0073] Among them, k0 is the gravitational gain coefficient, k i is the repulsion gain coefficient of the selected i-th seedling, n1 is the number of repulsions, F att (q) is the gravitational force obtained by establishing the gravitational potential field based on the coordinate points on the path, is the repulsive force obtained by establishing the repulsive potential field based on the selected position of the i-th seedling;

[0074] Assume that the rear wheel position of the paddy field weeder is q(x,y), and the coordinate point on the path that will crush the seedlings is q g (g x ,g y ), gravitational potential field U att (q) and the gravitational force F att The expressions of (q) are:

[0075]

[0076]

[0077] Among them, η is the proportional gain coefficient, ρ(q,q g) is the Euclidean distance between the rear wheel position of the paddy field weeder and the coordinate point on the path that will crush the seedlings. The direction is defined as the direction from the rear wheel position of the paddy field weeder to the coordinate point on the path. The gravitational potential field is mainly related to ρ(q,q g ), as ρ(q,q g ) gradually shrinks, and the gravitational potential field gradually decreases.

[0078] The repulsive potential field U established based on the selected seedling position rep The expression is:

[0079]

[0080] γ is the proportional gain coefficient, q0 is the seedling position point, and the corresponding repulsive force is the negative gradient of the repulsive potential field, expressed as:

[0081]

[0082] Among them, when the rear wheels of the paddy field weeder travel along the re-planned new path, the satellite navigation information of the dual satellite antennas is used to obtain the position information and posture information of the paddy field weeder, and the angle sensor at the front wheel is used to obtain the current turning angle information of the front wheel. According to the geometric relationship between the re-planned new path and the position and posture of the paddy field weeder, the improved pure tracking algorithm is used to calculate the target turning angle of the front wheel of the paddy field weeder, and the current turning angle of the front wheel is subtracted from the front wheel target turning angle δ to obtain the required turning angle of the front wheel, wherein the required turning angle of the front wheel is positive when it indicates a right turn, and negative when it indicates a left turn. The four-wheel steering mechanism 3 is driven by the motor, thereby driving the front wheel (of course, both the rear wheel and the front wheel can also rotate) to rotate the required turning angle of the front wheel, so that the paddy field weeder can track the re-planned new path when turning at the ridge head;

[0083] Traditional pure tracking algorithm preview distance L d is a fixed value, but the preview distance L d The tracking effect of the pure tracking algorithm has a relatively large impact, including tracking accuracy, stability, tracking speed and reaction time. The present invention simulates the preview distance L when manual driving d The preview distance expression is given as: L d =0.55v+L0, where L0 is the initial value of the preview distance;

[0084] Calculate the target turning angle of the front wheel of the paddy field weeder as follows:

[0085] like Figure 6 As shown in the figure, the pure tracking algorithm takes the rear wheel driving position as the tangent point and the longitudinal direction of the paddy field weeder as the tangent line. By controlling the front wheel target turning angle δ, the paddy field weeder can move along a path passing through the preview point (g x ,g y) arc travel, and the preview point (g x ,g y ) on the new path after re-planning; therefore, the present invention firstly calculates the path according to the preview distance L d Get the preview point on the new path after replanning (g x ,g y ), the following formula is derived using the law of sine:

[0086]

[0087] In formula (1), α is the heading angle, which is the current posture of the paddy field weeder and the preview point (g x ,g y ) and the angle between the line connecting the rear wheel driving position points.

[0088] Simplified:

[0089]

[0090] Substituting formula (2) into R = L / tanδ, we can obtain the target turning angle of the front wheel of the paddy field weeder:

[0091]

[0092] L d =0.55v+L0 Substituting into formula (3), we get:

[0093]

[0094] like Figure 12As shown, as a preferred embodiment, the four-wheel steering mechanism 3 is composed of a front steering mechanism and a rear steering mechanism with the same structure and symmetrical arrangement; the front steering mechanism is composed of a front wheel 612, a front axle 615, a front steering knuckle 617, a front steering knuckle arm 623, a front steering ball head 624 and a front connecting rod 616; the cylinder body of the front steering cylinder 67 is fixed to the bottom of the lawn mower body through the front axle 615, and the piston rod of the front steering cylinder 67 is hinged to one end of the front connecting rod 616 and one end of a front steering knuckle arm 623 through the front steering ball head 624; the other end of the front connecting rod 616 is hinged to one end of the other front steering knuckle arm 623; the other ends of the two front steering knuckle arms 623 are respectively connected to the front steering knuckle 617 A front wheel 612 is hinged; the rear steering mechanism consists of a rear wheel 618, a rear axle 620, a rear steering knuckle 619, a rear steering knuckle arm 625, a rear steering ball head 626 and a rear connecting rod 621; the cylinder body of the rear steering cylinder 68 is fixed to the bottom of the lawn mower body through the rear axle 620, and the piston rod of the rear steering cylinder 68 is hinged through the rear steering ball head 626 and one end of the rear connecting rod 621 and one end of a rear steering knuckle arm 625; the other end of the rear connecting rod 621 is hinged to one end of the other rear steering knuckle arm 625; the other ends of the two rear steering knuckle arms 625 are respectively hinged to a rear wheel 618 through the rear steering knuckle 619; the front steering cylinder 67 and the rear steering cylinder 68 are connected through a hydraulic system 622.

[0095] like Figure 13 As shown, the hydraulic system 622 consists of an oil tank 61, a filter 62, a hydraulic pump 63, an electric motor 64, a fully hydraulic steering gear 65, a two-position four-way solenoid valve 69, a relief valve 610, and a cooler 611. The fully hydraulic steering gear and the two-position four-way solenoid valve 69 are both controlled by an industrial computer. The oil outlet of the oil tank 61 is connected to the oil inlet of the filter 62 to filter impurities in the oil. The oil outlet of the filter 62 is connected to the oil inlet of the hydraulic pump 63. The input shaft of the hydraulic pump 63 is connected to the output shaft of the electric motor 64, which provides power. The oil outlet of the hydraulic pump 63 is connected to the oil inlet (P port) of the fully hydraulic steering gear 65. The two oil outlets (A port and B port) of the fully hydraulic steering gear 65 are connected to the rodless cavity of the front steering cylinder 67 and the two-position four-way solenoid valve 69. The return oil port (T port) is connected, the rod chamber of the front steering cylinder 67 is connected to the oil inlet (P port) of the two-position four-way solenoid valve 69; the two oil outlets (A port and B port) of the two-position four-way solenoid valve 69 are connected to the rodless chamber and the rod chamber of the rear steering cylinder 68; the return oil port (T port) of the full hydraulic steering gear 65 is connected to the oil inlet of the overflow valve 610, and the oil outlet of the overflow valve 610 is connected to the oil return port of the oil tank 61 through the cooler 611, and the oil returns to the oil tank 61 after cooling.

[0096] More preferably, the two oil outlets (port A and port B) of the fully hydraulic steering gear 65 are connected to the rodless chamber of the front steering cylinder 67 and the return oil port (port T) of the two-position four-way solenoid valve 69 through the steering valve block 66, and the steering valve block 66 plays a role in stabilizing the pressure.

[0097] More preferably, the valve core in the full hydraulic steering gear 65 is fixed to the steering wheel via the steering column 614, and manual steering can be performed via the steering wheel when needed.

[0098] The motor 64 drives the hydraulic pump 63 to rotate, thereby driving the oil in the oil tank 61 to flow into the P port of the full-hydraulic steering gear 65; when steering is not required, the P port of the full-hydraulic steering gear 65 is connected to the T port, and the oil flowing into the P port of the full-hydraulic steering gear directly flows back to the oil tank 61 from its T port; when only the front wheels are selected for steering and turning left, the P port of the full-hydraulic steering gear 65 is controlled to be connected to the B port, and the two-position four-way solenoid valve 69 is controlled to be de-energized. At this time, the oil flows from the P port of the full-hydraulic steering gear 65 through its B port into the T port of the two-position four-way solenoid valve 69, and then from the T port of the two-position four-way solenoid valve 69 through its P port into the T port of the front steering cylinder 67. The piston rod of the front steering cylinder 67 retracts, and drives the two front wheels 612 to turn left through the front steering knuckle arm 623 and the front connecting rod 616. The oil in the rodless chamber of the front steering cylinder 67 flows back to the oil tank 61 from the A port and T port of the full-hydraulic steering gear 65. When only the front wheels are turned and the steering is right, the P port and the A port of the full hydraulic steering gear 65 are controlled to be connected, and the two-position four-way solenoid valve 69 is controlled to be de-energized. At this time, the oil enters the P port of the front steering cylinder 67 from the P port of the full hydraulic steering gear 65 through its A port, and the piston rod of the front steering cylinder 67 is pushed out, driving the two front wheels 612 to rotate to the right through the front steering knuckle arm 623 and the front connecting rod 616. The oil in the rod chamber of the front steering cylinder 67 enters the B port of the full hydraulic steering gear 65 from the P port of the two-position four-way solenoid valve 69 through its T port, and then flows back to the oil tank 61 from the B port of the full hydraulic steering gear 65 through its T port.

[0099] When four-wheel counter-steering is selected and the front wheel turns left, the P port of the full hydraulic steering gear 65 is controlled to be connected to the B port, the two-position four-way solenoid valve 69 is energized, the T port of the two-position four-way solenoid valve 69 is connected to the A port, and the B port is connected to the P port; at this time, the oil flows from the P port of the full hydraulic steering gear 65 through its B port into the T port of the two-position four-way solenoid valve 69, and then from the T port of the two-position four-way solenoid valve 69 through its A port into the rodless chamber of the rear steering cylinder 68, and the piston rod of the rear steering cylinder 68 is pushed out. The two rear wheels 618 are driven to rotate to the right through the rear steering knuckle arm 625 and the rear connecting rod 621; the oil in the rod chamber of the rear steering cylinder 68 enters the rod chamber of the front steering cylinder 67 from the B port of the two-position four-way solenoid valve 69 through its P port, the piston rod of the front steering cylinder 67 retracts, and drives the two front wheels 612 to rotate to the left through the front steering knuckle arm 623 and the front connecting rod 616, and the oil in the rodless chamber of the front steering cylinder 67 flows back to the oil tank 61 from the A port of the full hydraulic steering gear 65 through its T port.

[0100] When four-wheel counter-steering is selected and the front wheels turn right, the P port of the full hydraulic steering gear 65 is controlled to be connected to the A port, the two-position four-way solenoid valve 69 is energized, the T port of the two-position four-way solenoid valve 69 is connected to the A port, and the B port is connected to the P port; at this time, the oil flows from the P port of the full hydraulic steering gear 65 through its A port into the rodless cavity of the front steering cylinder 67, the piston rod of the front steering cylinder 67 is pushed out, and drives the two front wheels 612 to rotate to the right through the front steering knuckle arm 623 and the front connecting rod 616; the front steering cylinder The oil in the rod chamber of 67 flows into the B port of the two-position four-way solenoid valve 69 through the P port, and then flows into the rod chamber of the rear hydraulic cylinder from the B port of the two-position four-way solenoid valve 69. The piston rod of the rear hydraulic cylinder retracts and drives the two rear wheels 618 to rotate left through the rear steering knuckle arm 625 and the rear connecting rod 621; the oil in the rodless chamber of the rear hydraulic cylinder flows from the A port of the two-position four-way solenoid valve 69 through its T port into the B port of the full hydraulic steering gear 65, and then flows back to the oil tank 61 from the T port of the full hydraulic steering gear 65.

Claims

1. A method for avoiding seedlings and turning for a paddy field weeder based on satellite and visual navigation, characterized by: Step 1: First, based on the comparison between the minimum steering radius of the paddy field weeder and half of the steering width, inappropriate steering modes are eliminated from several steering modes, including U-shaped steering, smooth steering, bulb-shaped steering, hook-shaped steering, and fishtail-shaped steering. Then, a hierarchical structure model is established using the analytic hierarchy process to select the optimal steering mode. Step 2: Use the visual navigation information of the industrial camera to obtain the position information of the seedlings in the visual image. At the same time, according to the installation position and posture information of the industrial camera, obtain the mapping relationship between the camera coordinate system and the world coordinate system, and obtain the position information of the seedlings in the world coordinate system; Step 3. Plan the coordinate points on the path of the optimal steering method from the first coordinate point to the preset number of coordinate points as follows: select the seedlings within the radius range of ρ0 with the coordinate point as the center, and calculate that when the distance between the coordinate point and the selected i-th seedling position point is less than the threshold L1, it is considered that the rear wheel of the paddy field weeder will crush the i-th seedling when it continues to travel along the original path to the coordinate point; from the first coordinate point that is judged to crush the seedling to the last coordinate point that will crush the seedling, re-plan the new path according to the resultant force calculated by the artificial potential field method, and the coordinate points on the original path that will crush the seedlings are replaced by the coordinate points on the new path that will not crush the seedlings; during the process of the rear wheel of the paddy field weeder traveling along the re-planned new path, the above-mentioned planning is continuously performed on the unplanned coordinate points on the new path, so as to continuously update the subsequent path, so as to realize the paddy field weeder traveling along the real-time updated path to achieve seedling avoidance and steering; the seedling influence radius W s is the row spacing of seedlings, L s The spacing between seedlings.

2. The method for avoiding seedlings and turning a paddy field weeder based on satellite and visual navigation according to claim 1, characterized in that: The hierarchical structure model was established by using the analytic hierarchy process to select the optimal steering method. Specifically, the following steps were taken: ① Design the target layer, the intermediate layer and the scheme layer. The target layer is the selection of the optimal steering method. The intermediate layer is the factors that need to be considered in selecting the optimal steering method, including the steering seedling pressing amount A1, the steering time A2, the lateral offset of the steering final position A3 and the heading angle A4 of the steering final position. The scheme layer includes the various steering methods left after excluding the inappropriate steering methods; ② Construct the judgment matrix A = (a ij ) 4×4 , a ij for a i Factor and a j The relative importance of factors to the target is divided into five levels: equally important, slightly important, obviously important, strongly important, and extremely important. The judgment matrix A is tested for consistency, and the eigenvector corresponding to the maximum eigenvalue of the judgment matrix A is calculated and normalized to obtain the weight of each factor to the target W = [w1 w2 w3 w4], w i , i = 1, 2, 3, 4 is the weight of the i-th factor on the target; ③Construct a judgment matrix for each option relative to each factor n is the number of options, and each judgment matrix Through consistency detection, calculate each judgment matrix The eigenvector corresponding to the maximum eigenvalue is normalized to obtain the weight of each scheme on the factor is the weight of the k-th solution on the i-th factor, k = 1, 2, ..., n; ④ Perform a hierarchical total sorting consistency test. If the hierarchical total sorting consistency test fails, modify the judgment matrix A and the judgment matrix Re-execute steps ② and ③, and re-test the hierarchical total order consistency until the test passes. Calculate the total weight matrix and select the largest weight d in the total weight matrix. k The corresponding solution is the optimal steering method.

3. The method for avoiding seedlings and turning a paddy field weeder based on satellite and visual navigation according to claim 1, characterized in that: If the paddy field weeder is simplified into a bicycle model, the formula for calculating the turning radius of the paddy field weeder is R = L / tanδ, where R is the turning radius of the paddy field weeder, L is the wheelbase between the front and rear wheels at the bottom of the paddy field weeder, and δ is the target turning angle of the front wheel. When the target turning angle of the front wheel is maximum, the minimum turning radius of the paddy field weeder is calculated.

4. The method for avoiding seedlings and turning a paddy field weeder based on satellite and visual navigation according to claim 2, characterized in that: Define the consistency index CI = (λ-m) / (m-1), where λ is the maximum eigenvalue of the judgment matrix A and m is the order of the judgment matrix; introduce the random consistency index RI, check the average random consistency index table, determine the 4th-order judgment matrix RI = 0.90, and define the consistency ratio CR = CI / RI. When the consistency ratio CR is less than 0.1, it is considered that the inconsistency degree of A is within the allowable range.

5. The method for avoiding seedlings and turning a paddy field weeder based on satellite and visual navigation according to claim 2, characterized in that: In the hierarchical total sort consistency test, set the hierarchical total sort consistency ratio Where CI i is the consistency index of the solution layer relative to the i-th factor in the factor layer, RI i is the corresponding random consistency index, if the hierarchical total ranking consistency ratio CR T <0.1, the total hierarchical order consistency is acceptable; the total weight matrix D = [d1, d2, ..., d n ], d k represents the weight of the k-th solution to the target, Select the largest weight d in the total weight matrix k The corresponding solution is the optimal steering method.

6. The method for avoiding seedlings and turning a paddy field weeder based on satellite and visual navigation according to claim 1, characterized in that: Step 2 is as follows: The visual images collected by the industrial camera are converted to grayscale using 2G-RB color images and then binarized using Ostu. Morphological operations are then performed on the binary images. The radius range is set, and the minEnclosingCircle function is used to find the minimum circumscribed circle that encloses the seedlings in the image. The center of the circle is the position coordinate P of the seedling in the image. uv ; At the same time, the intrinsic parameter matrix K of the industrial camera is obtained according to the industrial camera calibration m , the external parameter matrix R of the industrial camera is obtained from the installation position and posture information of the industrial camera m , then the mapping relationship between the camera coordinate system and the world coordinate system is P uv =K m R m P W According to this relationship, the position coordinates P of the seedling in the world coordinate system can be obtained W .

7. The method for avoiding seedlings and turning a paddy field weeder based on satellite and visual navigation according to claim 1, characterized in that: Using the artificial potential field method, the calculated resultant force is as follows: Among them, k0 is the gravitational gain coefficient, k i is the repulsion gain coefficient of the selected i-th seedling, n1 is the number of repulsions, F att (q) is the gravitational force obtained by establishing the gravitational potential field based on the coordinate points on the path, is the repulsive force obtained by establishing the repulsive potential field based on the selected position of the i-th seedling; Assume that the rear wheel position of the paddy field weeder is q(x, y), and the coordinate point on the path that will crush the seedlings is q g (g x , g y ), gravitational potential field U att (q) and the gravitational force F att The expressions of (q) are: Among them, η is the proportional gain coefficient, ρ(q,q g ) is the Euclidean distance between the rear wheel driving position of the paddy field weeder and the coordinate point on the path that will crush the seedlings, and the direction is defined as the direction from the rear wheel driving position of the paddy field weeder to the coordinate point on the path; The repulsive potential field U established based on the selected seedling position rep The expression is: γ is the proportional gain coefficient, q0 is the seedling position point, and the corresponding repulsive force is the negative gradient of the repulsive potential field, expressed as:

8. The method for avoiding seedlings and turning a paddy field weeder based on satellite and visual navigation according to claim 1, characterized in that: When the rear wheels of the paddy field weeder travel along the replanned new path, the satellite navigation information of the dual satellite antennas is used to obtain the position information and posture information of the paddy field weeder, and the angle sensor at the front wheel is used to obtain the current turning angle information of the front wheel. According to the geometric relationship between the replanned new path and the position and posture of the paddy field weeder, an improved pure tracking algorithm is used to calculate the target turning angle of the front wheel of the paddy field weeder. The current turning angle of the front wheel is subtracted from the front wheel target turning angle δ to obtain the required turning angle of the front wheel. A positive required turning angle indicates a right turn, and a negative required turning angle indicates a left turn. The four-wheel steering mechanism is driven by a motor, thereby driving the front wheel to rotate the required turning angle or half of the required turning angle. The rear wheels rotate an angle of the same size and opposite direction as the front wheels, so that the paddy field weeder can follow the replanned new path when turning at the ridge head.

9. The method for avoiding seedlings and turning a paddy field weeder based on satellite and visual navigation according to claim 8, characterized in that: Calculate the target turning angle of the front wheel of the paddy field weeder as follows: Take the rear wheel driving position as the tangent point and the longitudinal direction of the paddy field weeder as the tangent line, according to the preview distance L d Get the preview point on the new path after replanning (g x , g y ), the following formula is derived using the law of sine: In formula (1), α is the heading angle, which is the current posture of the paddy field weeder and the preview point (g x , g y ) and the angle between the line connecting the rear wheel driving position points; preview distance L d =0.55v+L0, where L0 is the initial value of the preview distance; Simplified: Substituting formula (2) into R = L / tanδ, we can obtain the target turning angle of the front wheel of the paddy field weeder: L d =0.55v+L0 Substituting into formula (3), we get:

10. The method for avoiding seedlings and turning a paddy field weeder based on satellite and visual navigation according to claim 8, characterized in that: The four-wheel steering mechanism is composed of a front steering mechanism and a rear steering mechanism with the same structure and symmetrical arrangement; the front steering mechanism is composed of a front wheel, a front axle, a front steering knuckle, a front steering knuckle arm, a front steering ball head and a front connecting rod; the cylinder body of the front steering oil cylinder is fixed to the bottom of the lawn mower body through the front axle, and the piston rod of the front steering oil cylinder is hinged through the front steering ball head and one end of the front connecting rod and one end of a front steering knuckle arm; the other end of the front connecting rod is hinged to one end of the other front steering knuckle arm; the other ends of the two front steering knuckle arms are respectively hinged to one end of the front steering knuckle arm through the front steering knuckle The front wheels are hinged; the rear steering mechanism consists of a rear wheel, a rear axle, a rear steering knuckle, a rear steering knuckle arm, a rear steering ball head and a rear connecting rod; the cylinder body of the rear steering cylinder is fixed to the bottom of the lawn mower body through the rear axle, and the piston rod of the rear steering cylinder is hinged through the rear steering ball head and one end of the rear connecting rod and one end of a rear steering knuckle arm; the other end of the rear connecting rod is hinged to one end of the other rear steering knuckle arm; the other ends of the two rear steering knuckle arms are respectively hinged to a rear wheel through the rear steering knuckle; the front steering cylinder and the rear steering cylinder are connected by a hydraulic system; The hydraulic system consists of an oil tank, a filter, a hydraulic pump, an electric motor, a full hydraulic steering gear, a two-position four-way solenoid valve, a relief valve and a cooler; the full hydraulic steering gear and the two-position four-way solenoid valve are both controlled by an industrial computer; the oil outlet of the oil tank is connected to the oil inlet of the filter, the oil outlet of the filter is connected to the oil inlet of the hydraulic pump, the input shaft of the hydraulic pump is connected to the output shaft of the electric motor, the oil outlet of the hydraulic pump is connected to the oil inlet of the full hydraulic steering gear, the two oil outlets of the full hydraulic steering gear are connected to the rodless cavity of the front steering cylinder and the return oil port of the two-position four-way solenoid valve, the rod cavity of the front steering cylinder is connected to the oil inlet of the two-position four-way solenoid valve; the two oil outlets of the two-position four-way solenoid valve are connected to the rodless cavity and the rod cavity of the rear steering cylinder; the return oil port of the full hydraulic steering gear is connected to the oil inlet of the relief valve, and the oil outlet of the relief valve is connected to the return oil port of the oil tank via the cooler.

Images