Inverse kinematics method of gantry crane, storage medium and electronic equipment
By setting an encoder at the end of the motor of the portal crane and using neural network learning, the kinematic inverse solution of the crane is achieved, the inverse solution problems caused by complex structure in the existing technology are solved, and efficient adaptation and real-time control of cranes of different structures are achieved.
Patent Information
- Application Number
- CN202411596353.2
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-11-11
- Publication Date
- 2025-02-28
AI Technical Summary
The structure of the gantry crane is complex, and it is difficult for the prior art to use a unified and convenient kinematic inverse solution method to adapt to models of different structures.
By setting an encoder at the end of the motor of the crane, recording the motion angle and amplitude distance, the kinematic inverse solution results are obtained using neural network learning. This method does not rely on complex structural solutions, and performs inverse kinematic solutions through experimental point acquisition.
It realizes an efficient kinematic inverse solution for the gantry crane, which is suitable for crane models of different structures, without re-derivation and analysis, is widely applicable and easy to control operations in real time.
Smart Images
Figure CN119528022A_ABST
Abstract
Description
Technical field:
[0001] The invention belongs to the technical field of crane automatic control, and in particular relates to a kinematic inverse solution method, a storage medium and an electronic device for a portal crane. Background technology:
[0002] Today, with the rapid development of automation technology, the automation transformation of port machinery is of great significance to the intelligent development of ports. The automation transformation of gantry cranes is a key step in complying with the global trend of port modernization, and is of great significance for improving competitiveness and achieving sustainable development. The automated control of gantry cranes needs to consider the trajectory control of the end load of the crane, and the inverse kinematics solution is of great significance in the trajectory control of the crane. The inverse kinematics solution can calculate the specific motion parameters required for each joint of the crane (such as rotation, amplitude change, lifting, etc.) according to the target position and posture (i.e., position, orientation, etc.) of the desired end effector (such as a hook or grab) in space. This is crucial for achieving precise loading and unloading operations, especially in automated control, to ensure that the goods are efficiently and safely transported to the designated location. In the prior art, gantry cranes have complex structures, and the solution methods for different structures are different, making it difficult to use a unified and convenient solution method. Therefore, the present invention provides a kinematic inverse solution method, storage medium and electronic device for a gantry crane to solve the above problems. Summary of the invention:
[0003] The purpose of the present invention is to address the deficiencies in the prior art and to provide a kinematic inverse solution method, storage medium and electronic equipment for a gantry crane, which can efficiently adapt to models of various structures and obtain kinematic inverse solution results.
[0004] The present invention adopts the following technical solutions:
[0005] (I) The present invention provides a kinematic inverse solution method for a gantry crane, characterized in that it comprises the following steps:
[0006] S1. An encoder is set at the end of the luffing motor of the crane to record the output value of the luffing motor encoder at the luffing position, and the movement angle of the luffing motor is obtained through the encoder output value; the luffing distance / luffing lifting height of the crane hook corresponding to the current luffing position is measured and recorded;
[0007] S2, substitute the output value of the luffing motor encoder and the luffing distance / luffing lifting height data of the crane hook into the neural network for learning, and obtain the relationship between the movement angle of the luffing motor and the luffing distance / luffing lifting height of the crane hook under pure luffing motion;
[0008] S3. Encoders are respectively arranged at the ends of the hoisting motor and the rotating motor of the crane to record the encoder output values of the hoisting motor and the rotating motor during the hoisting and rotating motions, and the motion angles of the hoisting motor and the rotating motor are obtained through the encoder output values;
[0009] S4. Obtain the corresponding relationship between the rotation angle of the crane hook and the movement angle of the rotating motor according to the reduction ratio of the rotating structure, and obtain the corresponding relationship between the movement angle of the lifting motor and the lifting height of the crane hook according to the reduction ratio of the pulley block and the lifting motor.
[0010] Furthermore, in the steps S1 and S2, the spatial coordinates of the crane hook at the minimum amplitude variation position and the maximum amplitude variation position, the output value of the amplitude variation motor encoder, and the spatial coordinates of the crane hook at the amplitude variation s times between the minimum amplitude variation position and the maximum amplitude variation position are recorded respectively, and the spatial coordinates of the crane hook at the amplitude variation s times between the minimum amplitude variation position and the maximum amplitude variation position are substituted into the neural network for learning to obtain the relationship between the movement angle of the amplitude variation motor and the amplitude variation distance / amplitude variation lifting height of the crane hook under pure amplitude variation motion.
[0011] Furthermore, in step S1, a QR code label for positioning is pasted on the crane hook, and a scale plate is installed directly below the crane hook, and a visual module capable of capturing the QR code is set up at the far end of the crane; the visual module recognizes the movement of the center position of the QR code label and the projected scale reading of the crane hook on the scale plate, and obtains the spatial coordinates of the crane hook in the current luffing position, thereby obtaining the luffing distance / luffing lifting height of the crane hook.
[0012] Furthermore, the method for determining the variable distance / variable lifting height of the crane hook in the variable position is as follows: 1) calibrating the distortion of the camera of the visual module, and correcting the two-dimensional coordinates (xp, yp) of the two-dimensional code in the camera image through affine transformation. If the camera resolution is m×n, then 0<xp<m, 0<yp<n, and the two-dimensional pixel coordinates of the center coordinates of the two-dimensional code label are (x i ,y i )(i=1,2,3…,n), n=s+2, the coordinate corresponding to the minimum amplitude variation distance is (x 0 ,y 0 ), the coordinate corresponding to the maximum amplitude variation distance is (x n ,y n );2) mark the distance d corresponding to the scale plate i (i=1,2,3…,n), n=s+2, the value corresponding to the minimum amplitude variation distance is d 0 , the value corresponding to the maximum amplitude variation distance is d n ; 3) x i With d i Correspondingly, get the distance corresponding to each pixel Then the luffing lifting height of the crane hook at each luffing position during the luffing process is obtained. l f = Δ ×(x i -x 0 ).
[0013] Furthermore, in the steps S1 and S3, the encoder value e of the motor corresponding to each movement is recorded. e The range is determined by the resolution of the installed encoder, that is, the number of pulses corresponding to one rotation of the motor. If the resolution of the encoder is em, then the angle of rotation of the motor is The angle of each motor movement is recorded as a i (i=1,2,3…,n).
[0014] Furthermore, the visual module is capable of observing the entire crane, and the length of the scale plate is greater than the maximum amplitude variation distance of the crane.
[0015] (ii) The present invention provides a computer-readable storage medium storing a computer program, wherein the computer program enables a computer to execute the kinematic inverse solution method for the gantry crane described above.
[0016] (ii) The present invention provides an electronic device, comprising: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the computer program, the kinematic inverse solution method for the gantry crane described above is implemented.
[0017] Beneficial effects of the present invention:
[0018] (1) The present invention discloses a kinematic inverse solution method for a gantry crane. The end of the crane hook is tracked and located according to a visual device, and the two-dimensional coordinates of the end of the hook are accurately output. Within the operable range of the crane boom, the motion stroke is subdivided into multiple segments, and the encoder data at the tail end of each motor is matched with the two-dimensional coordinates of the end of the boom. Multiple groups of data are collected, and then the corresponding point sets are trained through deep learning to finally obtain a curve of the inverse kinematic solution result.
[0019] (2) The present invention does not rely on complex structural solutions, but only needs to perform inverse kinematics solutions by experimental point selection, which provides a feasible and more efficient way to solve complex inverse kinematic problems and has wide applicability.
[0020] (3) The inverse kinematics solution of the present invention can be very fast and suitable for real-time control operations. At the same time, for different crane models, there is no need to re-derive the analysis, only retrain or fine-tune the neural network. This means that the neural network method is more universal and easy to adapt to the inverse kinematics problems of various mechanical structures. In some cases, empirical data from actual operations (such as non-ideal factors such as friction and joint clearance) can be directly integrated into the neural network training, making the model closer to the motion behavior of the real world and improving the practicality of the model. Description of the drawings:
[0021] Figure 1 is a flow chart of the method of the present invention;
[0022] Figure 2 It is a working block diagram of the present invention. Specific implementation method:
[0023] In order to make the purpose, technical solution and advantages of the embodiments of the present invention clearer, the technical solution in the embodiments of the present invention will be clearly and completely described in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present invention.
[0024] Example 1
[0025] An embodiment of the present invention provides a kinematic inverse solution method for a portal crane, comprising:
[0026] Step 1: Measure and record the luffing distance / luffing lifting height of the crane hook in the luffing position, and the movement angle of the luffing motor in the luffing position.
[0027] (1) A QR code label (feature point) for positioning is attached to the crane hook, and a scale plate is installed directly below the crane hook. The length of the scale plate is greater than the maximum amplitude variation distance of the crane.
[0028] (2) Set up a vision module, a computer, and a wireless receiving terminal at the remote end of the crane to collect feature points for visual calculation and receive encoder data for solution. It is necessary to ensure that the vision module can observe the entire crane and can capture the QR code on the hook.
[0029] (3) An encoder is set at the end of the luffing motor of the crane to record the output value of the luffing motor encoder at the luffing position. The data is remotely transmitted to the wireless receiving terminal through the wireless transmission module and uploaded to the computing computer. The movement angle of the luffing motor can be obtained through the encoder output value.
[0030] Record the encoder value e of the motor corresponding to each movement. The range of e is determined by the resolution of the installed encoder, that is, the number of pulses corresponding to one rotation of the motor. The resolution of the encoder is e m , then the angle of motor rotation The angle of each motor movement is recorded as ai (i=1,2,3…,n).
[0031] (4) When the rotating motor and the lifting motor are stationary, the crane is manually controlled to reach the minimum amplitude change position. The visual module recognizes the QR code label, marks the scale position of the vertical projection of the crane hook load at the minimum amplitude change position to the ground, records the amplitude change distance / amplitude change lifting height of the crane hook at the minimum amplitude change position, and records the position value output by the current amplitude change motor encoder.
[0032] (5) When the rotating motor and the lifting motor are stationary, the crane is manually controlled to reach the maximum amplitude change position. The visual module recognizes the QR code label, marks the scale position of the maximum amplitude change position where the crane hook load is vertically projected to the ground, records the amplitude change distance / amplitude change lifting height of the crane hook at the maximum amplitude change position, and records the position value output by the current amplitude change motor encoder.
[0033] (6) When the rotating motor and the lifting motor are stationary, the crane is manually controlled to change its amplitude s times in the variable amplitude working space (from the minimum variable amplitude position to the maximum variable amplitude position). Each time it moves, the visual module recognizes the QR code label, marks the current variable amplitude position and the corresponding scale plate position, records the variable amplitude distance / variable lifting height of the crane hook at the current variable amplitude position, and records the position value output by the current variable amplitude motor encoder.
[0034] Step 2: Substitute the recorded output value of the luffing motor encoder and the actual measured luffing distance / luffing lifting height data of the crane hook into the neural network for training, and obtain the relationship between the movement angle of the luffing motor and the luffing distance / luffing lifting height of the crane hook under pure luffing motion.
[0035] Through neural network training, the two-dimensional relationship y=f(x) is obtained, where y is the encoder value of the motor, and x can be the amplitude variation distance of pure amplitude variation motion or the amplitude variation lifting height under the amplitude variation motion. Through this neural network, the kinematic inverse model of pure amplitude variation motion can be trained.
[0036] Step 3: Set encoders at the ends of the hoisting motor and rotating motor of the crane, record the encoder output values of the hoisting motor and rotating motor during hoisting and rotating motion, and transmit the data remotely to the wireless receiving terminal through the wireless transmission module and upload it to the computer. The motion angle of the hoisting motor and rotating motor can be obtained through the encoder output value.
[0037] Record the encoder value e of the motor corresponding to each movement. The range of e is determined by the resolution of the installed encoder, that is, the number of pulses corresponding to one rotation of the motor. The resolution of the encoder is e m , then the angle of motor rotation The angle of each motor movement is recorded as ai (i=1,2,3…,n)
[0038] Step 4: Obtain the corresponding relationship between the rotation angle of the crane hook and the movement angle of the rotating motor according to the reduction ratio of the rotating structure, and obtain the corresponding relationship between the movement angle of the lifting motor and the lifting height of the crane hook according to the reduction ratio of the pulley block and the lifting motor. This is a conventional method in the art and will not be described in detail here.
[0039] The obtained correspondence between the movement angle of the luffing motor and the luffing distance / luffing lifting height of the crane hook, the correspondence between the rotation angle of the crane hook and the movement angle of the rotating motor, and the correspondence between the movement angle of the lifting motor and the lifting height of the crane hook are the inverse kinematic solutions of the crane.
[0040] Example 2
[0041] Based on Example 1, the embodiment of the present invention describes a method for measuring the luffing distance / luffing lifting height of a crane hook in a luffing position, specifically:
[0042] (1) Calibrate the camera distortion and correct the two-dimensional coordinates (xp, yp) of the QR code in the camera image through affine transformation. If the camera resolution is m×n, then 0<xp<m, 0<yp<n, and the two-dimensional pixel coordinates of the center coordinates of the QR code label are (x i ,y i )(i=1,2,3…,n), n=s+2, the coordinate corresponding to the minimum amplitude variation distance is (x 0 ,y 0 ), the coordinate corresponding to the maximum amplitude variation distance is (x n ,y n ).
[0043] (2) The distance d corresponding to the marking scale plate i (i=1,2,3…,n), n=s+2, the value corresponding to the minimum amplitude variation distance is d 0 , the value corresponding to the maximum amplitude variation distance is d n .
[0044] (3) Match xi with di to obtain the distance corresponding to each pixel Then the luffing lifting height of the crane hook at each luffing position during the luffing process is obtained. l f = Δ×(xi-x0).
[0045] Example 3
[0046] This embodiment provides a computer-readable storage medium storing a computer program, wherein the computer program enables a computer to execute the kinematic inverse solution method for the gantry crane described in Embodiment 2.
[0047] Example 4
[0048] This embodiment provides an electronic device, including: a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, the kinematic inverse solution method for the gantry crane described in Example 3 is implemented.
[0049] The present invention discloses a kinematic inverse solution method for a gantry crane. In the automated control of a gantry crane, kinematic solution is very important. The forward kinematic equation of a gantry crane can be easily obtained by a geometric method. However, if the trajectory planning of the end of the cantilever is required, the inverse kinematic solution of the gantry crane is required. Due to the complex structure of the gantry crane cantilever, the inverse kinematic solution has the characteristics of high dimension and nonlinearity. Therefore, we track and locate the end of the hook of the crane according to the visual device, and accurately output the two-dimensional coordinates of the end of the hook; within the operable range of the crane cantilever, the motion stroke is subdivided into multiple sections, and the encoder data at the tail end of each motor is matched with the two-dimensional coordinates of the end of the cantilever, multiple groups of data are collected, and then the corresponding point set is trained by deep learning, and finally a curve of the inverse kinematic solution result is obtained. The present invention does not rely on complex structural solution, and only needs to perform inverse kinematic solution by experimental point selection. It has wide applicability and is easy to use in engineering practice.
[0050] The above are only preferred implementations of the present invention. The protection scope of the present invention is not limited to the above embodiments. All technical solutions under the concept of the present invention belong to the protection scope of the present invention. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications without departing from the principle of the present invention should be regarded as the protection scope of the present invention.
Claims
1. A kinematic inverse solution method for a portal crane, characterized in that: The following steps are involved: S1. An encoder is set at the end of the luffing motor of the crane to record the output value of the luffing motor encoder at the luffing position, and the movement angle of the luffing motor is obtained through the encoder output value; the luffing distance / luffing lifting height of the crane hook corresponding to the current luffing position is measured and recorded; S2, substitute the output value of the luffing motor encoder and the luffing distance / luffing lifting height data of the crane hook into the neural network for learning, and obtain the relationship between the movement angle of the luffing motor and the luffing distance / luffing lifting height of the crane hook under pure luffing motion; S3. Encoders are respectively arranged at the ends of the hoisting motor and the rotating motor of the crane to record the encoder output values of the hoisting motor and the rotating motor during the hoisting and rotating motions, and the motion angles of the hoisting motor and the rotating motor are obtained through the encoder output values; S4. Obtain the corresponding relationship between the rotation angle of the crane hook and the movement angle of the rotating motor according to the reduction ratio of the rotating structure, and obtain the corresponding relationship between the movement angle of the lifting motor and the lifting height of the crane hook according to the reduction ratio of the pulley block and the lifting motor.
2. The kinematic inverse solution method of a portal crane according to claim 1, characterized in that: In the steps S1 and S2, the spatial coordinates of the crane hook at the minimum amplitude variation position and the maximum amplitude variation position, the output value of the amplitude variation motor encoder, and the spatial coordinates of the crane hook at the amplitude variation s times between the minimum amplitude variation position and the maximum amplitude variation position are recorded respectively, and the spatial coordinates of the crane hook at the amplitude variation s times between the minimum amplitude variation position and the maximum amplitude variation position, and the output value of the amplitude variation motor encoder are substituted into the neural network for learning to obtain the relationship between the movement angle of the amplitude variation motor and the amplitude variation distance / amplitude variation lifting height of the crane hook under pure amplitude variation motion.
3. The kinematic inverse solution method of a gantry crane according to claim 2, characterized in that: In the step S1, A QR code label for positioning is attached to the crane hook, a scale plate is installed directly below the crane hook, and a visual module capable of capturing the QR code is set up at the far end of the crane; The visual module recognizes the movement of the center position of the QR code label and the projected scale reading of the crane hook on the scale plate to obtain the spatial coordinates of the crane hook at the current luffing position, that is, the luffing distance / luffing lifting height of the crane hook.
4. The kinematic inverse solution method of a gantry crane according to claim 3, characterized in that: The method for determining the luffing distance / luffing lifting height of the crane hook in the luffing position is: 1) Calibrate the distortion of the camera of the visual module and correct the two-dimensional coordinates (x p ,y p ), if the camera resolution is m×n, then 0<x p <m,0<y p <n, the 2D pixel coordinates of the center coordinates of the 2D code label are (x i ,y i )(i=1,2,3 … ,n), n=s+2, the coordinate corresponding to the minimum amplitude change distance is (x0,y0), and the coordinate corresponding to the maximum amplitude change distance is (x n ,y n ); 2) Mark the distance d corresponding to the scale plate i (i=1,2,3 … ,n), n=s+2, the value corresponding to the minimum amplitude change distance is d0, and the value corresponding to the maximum amplitude change distance is d n ; 3) x i With d i Correspondingly, get the distance corresponding to each pixel Then the luffing height l of the crane hook at each luffing position during the luffing process is obtained. f =Δ×(x i -x0).
5. The kinematic inverse solution method of a portal crane according to claim 1, characterized in that: In the steps S1 and S3, Record the encoder value e of the motor corresponding to each movement. The range of e is determined by the resolution of the installed encoder, that is, the number of pulses corresponding to one rotation of the motor. The resolution of the encoder is e m , then the angle of motor rotation The angle of each motor movement is recorded as a i (i=1,2,3 … ,n).
6. The kinematic inverse solution method of a portal crane according to claim 3, characterized in that: The visual module can observe the entire crane, and the length of the scale plate is greater than the maximum amplitude variation distance of the crane.
7. A computer-readable storage medium storing a computer program, characterized in that: The computer program enables a computer to execute the kinematic inverse solution method for a portal crane as described in any one of claims 1 to 6.
8. An electronic device, characterized in that: include: A memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the computer program, the kinematic inverse solution method for the gantry crane as described in any one of claims 1 to 6 is implemented.