Novel stand column guide sleeve automatic locking device

Abstract

The novel stand column guide sleeve automatic locking device comprises a cylinder body and a guide sleeve, the guide sleeve is connected to the upper portion of the cylinder body in a threaded mode, an annular cavity is formed in the upper portion in the cylinder body, two grooves are symmetrically formed in the left side and the right side of the upper end of the cylinder body, the grooves communicate with the annular cavity, pawls are rotationally connected into the grooves, and a ratchet wheel is fixed to the outer portion of the guide sleeve. A pull rope is fixed to the outer wall of the pawl, the tail end of the pull rope movably penetrates through the cylinder body, an annular guide rail is fixed to the outside of the cylinder body, and a rotating ring is slidably connected into the annular guide rail. According to the novel automatic locking device for the stand column guide sleeve, when the guide sleeve and a cylinder body are installed, self-locking can be automatically completed by rotating the guide sleeve until the pawl is clamped with the ratchet wheel, the guide sleeve is prevented from loosening, and when the guide sleeve needs to be detached, the rotating ring is rotated to pull the pawl to be separated from the ratchet wheel through the two pull ropes, so that the guide sleeve is detached. And then the guide sleeve can be disassembled by reversely rotating the guide sleeve, so that the whole process of disassembling the guide sleeve is convenient and simple.

Description

Technical Field

[0001] The invention relates to the technical field of locking devices, in particular to a novel automatic locking device for a column guide sleeve. Background Art

[0002] Due to frequent use and pressure, the guide sleeves of existing columns often become loose. This is mainly because the columns are often in frequent use and have to withstand pressure from different directions. These pressures are easily destroyed by repeated action. The original stable installation state of the guide sleeve can cause it to loosen or even fall off. In order to deal with this problem, the industry has taken some corresponding solutions at this stage, but these measures have their own limitations. For example, a common practice is to weld an L-shaped iron block on the outer cylinder to buckle the guide sleeve, hoping to prevent it from loosening. In theory, this method can limit the displacement of the guide sleeve to a certain extent and play a role in preventing loosening. However, many disadvantages are exposed in actual application. Since the welding parts are subject to vibration, stress changes and various complex external forces during the operation of the equipment for a long time, the welding points often fall off. Moreover, when the column is subjected to a large pressure and rotates and rises against the guide sleeve, the L-shaped iron block itself is easily deformed due to the large external force. Once the deformation exceeds a certain limit, it can no longer effectively restrict the guide sleeve, which eventually causes the iron block to fall off, causing the guide sleeve to loosen again, losing the originally expected anti-loosening effect. In addition, there are also some designs that use a bayonet to support the slot to prevent the guide sleeve from loosening. Although this structure can inhibit the loosening of the guide sleeve to a certain extent, it itself has the problem of complex structure. When the guide sleeve needs to be disassembled for maintenance or replacement, the operation is extremely cumbersome. The closing screw must be removed first, and then the guide sleeve must be slightly tightened, and then the fixing rod must be pulled outward to disengage the bayonet from the slot. The whole process has many steps, which not only consumes a lot of time and manpower, but also requires high proficiency of the operator. If you are not careful, it may damage related components, which brings many inconveniences to the maintenance of the equipment.

[0003] In view of the above-mentioned problems existing in the existing structures for preventing the loosening of the column guide sleeve and disassembly, there is an urgent need for a new device and method that can not only reliably prevent the guide sleeve from loosening and ensure its stability during the use of the column, but also conveniently and quickly complete the operation when disassembly is required, so as to meet the actual needs for efficient and stable use of the column guide sleeve in industrial production. Summary of the invention

[0004] The purpose of the present invention is to overcome the defects of the prior art and provide a novel automatic locking device for a column guide sleeve to solve the problems raised in the above-mentioned background technology.

[0005] A novel automatic locking device for a column guide sleeve comprises a cylinder body and a guide sleeve, wherein the guide sleeve is threadedly connected to the upper part of the cylinder body, an annular cavity is provided in the upper part of the cylinder body, two grooves are symmetrically opened on the left and right sides of the upper end of the cylinder body, the grooves are communicated with the annular cavity, a pawl is rotatably connected inside the groove, a ratchet is fixed to the outside of the guide sleeve, the ratchet is connected to the pawl, a pull rope is fixed to the outer wall of the pawl, the end of the pull rope movably penetrates the cylinder body, an annular guide rail is fixed to the outside of the cylinder body, a swivel is slidably connected to the inside of the annular guide rail, the end of the pull rope movably penetrates the inside of the annular guide rail, and the end of the pull rope is fixed to the inner wall of the swivel.

[0006] Preferably, an outer wall of the guide sleeve is provided with an external thread below the ratchet, and an inner wall of the cylinder body is provided with an internal thread below the annular cavity, and the external thread cooperates with the internal thread.

[0007] Preferably, a convex ring is fixed to the outside of the guide sleeve, and the convex ring is located above the ratchet.

[0008] Preferably, a movable plate is fixed to one end of the pawl, a fixed plate is provided on one side of the movable plate, the lower end of the fixed plate is fixed to the bottom of the groove, and a spring is connected between the movable plate and the fixed plate.

[0009] Preferably, an annular groove is formed at both upper and lower ends of the inner wall of the annular guide rail, and an annular boss is fixed at both upper and lower ends of the rotating ring. The annular boss is located inside the annular groove and is slidably connected to the annular groove.

[0010] Preferably, the outer wall of the rotating ring is annular and has multiple anti-slip convex strips evenly fixed thereon.

[0011] Preferably, the swing angle range of the pawl is configured as follows:

[0012] The minimum swing angle of the pawl in the groove is α min , the maximum swing angle is α max ;

[0013] The minimum swing angle is α min The mathematical model is:

[0014] F 推 *h 推 ≥k*x 0 +Δx*L 1 *cosα min +μ*F N *L 1 ;

[0015] The maximum swing angle is α max The mathematical model is:

[0016] F 拉 *h拉 ≥k*x 0 +Δx*L 1 *cosα max +μ*F N *L 1 ;

[0017] Where, L 1 is the length of the pawl, that is, the distance from the rotation fulcrum to the end of the pawl, k is the elastic coefficient of the spring, x 0 is the initial compression of the spring, Δx is the extension of the spring when the pawl swings to the limit position, F 拉 is the pulling force of the rope on the pawl, h 拉 is the tension F 拉 The vertical distance between the action point and the fulcrum of the pawl, F 推 is the thrust of the ratchet wheel on the pawl, h 推 is the thrust F 推 The vertical distance between the action point and the fulcrum of the pawl, μ is the friction coefficient between the pawl and the groove, F N is the positive pressure between the pawl and the groove, and the minimum swing angle of the pawl in the groove is α min , the maximum swing angle is α max .

[0018] Preferably, the minimum swing angle α of the pawl is solved by a machine learning algorithm. min and the maximum swing angle α max , including the steps of:

[0019] S1. Collect multiple sets of data examples of parameters related to the automatic locking device of the column guide sleeve under different working conditions and specifications. For each example, record the following parameters: the length of the pawl L 1 , the elastic coefficient k of the spring, the initial compression of the spring x 0 , the elongation Δx of the spring when the pawl swings to the limit position, and the pulling force F of the pull rope on the pawl 拉 , tension F 拉 The vertical distance h between the action point and the fulcrum of the pawl 拉 , the thrust F of the ratchet on the pawl 推 Thrust F 推 The vertical distance h between the action point and the fulcrum of the pawl 推 The friction coefficient μ between the pawl and the groove is recorded, and the minimum swing angle α of the corresponding pawl actually measured in the groove is recorded. min and the maximum swing angle α max , organize all the collected data into a structured form, clean the data, and normalize it;

[0020] S2. Use the radial basis function kernel as the kernel function of the support vector regression model, and its formula is:

[0021] K(x i ,x j ) = exp(-γ||x i -x j || 2 )

[0022] Where γ is the width parameter, and γ>0, and the initial setting is γ=0.1; x i and x j Represent two sample vectors in the input space respectively;

[0023] S3. Use part of the cleaned and normalized preprocessed data set as the training set and the other part as the test set. Input the training set data into the constructed support vector regression model. Start the model training according to the set model parameters. Use the test set data to evaluate the trained SVR model. Input the input feature data of the test set into the model to obtain the predicted α. min and α max The value is then compared with the actual angle value in the test set. The model performance is optimized by continuously adjusting the parameters and retraining and evaluating until the minimum swing angle α of the pawl is predicted based on the input device parameters. min and the maximum swing angle α max , after obtaining a model that meets the accuracy requirements, for the new device, obtain its actual L 1 , k, x 0 , Δx, F 拉 、h 拉 、F 推 、h 推 , μ parameter values, after cleaning and normalization, the processed parameter data is input into the trained model, and the model will output the predicted α min and α max Value, based on the predicted α min and α max The value is used to configure the swing angle range of the pawl in the groove.

[0024] Beneficial effects of the present invention: The device realizes the automatic locking function by the cooperation of the ratchet and the pawl. When the guide sleeve and the cylinder body are threaded to the appropriate position, the ratchet and the pawl interact with each other. Under the action of the spring elastic force, the pawl engages the ratchet, which can effectively prevent the guide sleeve from loosening due to the reverse force (such as vibration, external force interference, etc. during the operation of the column), ensuring the stability of the guide sleeve under normal working conditions, and ensuring the stability and accuracy of the entire column system. This is especially important for some application scenarios that require high positioning accuracy and structural stability (such as the column structure in high-precision machining equipment). By setting components such as pull ropes, annular guide rails and swivels, when the guide sleeve needs to be disassembled, it is only necessary to rotate the swivel and use the pull rope to pull the pawl and ratchet apart, and then the guide sleeve can be easily rotated in the opposite direction for disassembly. The operation is simple and convenient. Compared with some traditional complex locking and disassembly structures, it greatly improves the efficiency of maintenance operations such as repairing and replacing the guide sleeve, and reduces maintenance time and labor costs. The convex ring arranged on the outside of the guide sleeve can cover the groove when it rotates to cover the top of the cylinder body, effectively preventing external dust from entering the groove and the relevant matching parts inside the device. The entry of dust often increases the friction and wear between components, and may even affect the normal operation of components such as the pawl. This dust-proof design of the convex ring helps to keep the inside of the device clean, further ensure the long-term stable and reliable operation of the device, and reduce the failure rate caused by dust factors. The annular guide rail and the swivel cooperate with the annular boss and the annular groove to achieve the limiting effect of the swivel, ensuring that the swivel can stably drive the pull rope to operate the pawl during the rotation process, avoiding axial movement and deviation of the swivel and affecting the normal operation of the device. At the same time, the anti-skid convex strips arranged on the outer wall of the swivel can play a good anti-skid role when rotating the swivel, so that the operator can apply the rotation force more stably and accurately when performing disassembly and other operations, improving the convenience and controllability of the operation.

[0025] By using mathematical models and combining machine learning algorithms to accurately set the minimum and maximum swing angles of the pawl in the groove, the pawl can move accurately in different operation stages. When installing the guide sleeve, it can ensure that the pawl can be smoothly pushed into the groove by the ratchet; in the normal working and loosening prevention stage, the pawl can reliably pop out and engage with the ratchet, effectively preventing the guide sleeve from rotating in the opposite direction and loosening; when removing the guide sleeve, it can ensure that the pull rope pulls the pawl and the ratchet to completely separate, so that the entire device can accurately and reliably complete the corresponding actions in various operation links such as installation, use, and disassembly, reducing unexpected situations caused by unreasonable angles, such as jamming, inability to lock normally, or difficulty in disassembly, and improving the accuracy and reliability of the overall operation of the device. Due to the above-mentioned precise angle configuration, abnormal wear and uneven force between components that may be caused by the unsmooth movement of the pawl and unstable engagement are avoided, so that the various components of the device can work under reasonable mechanical conditions, reducing the risk of premature damage to the components, thereby helping to extend the service life of the entire new column guide sleeve automatic locking device, reduce the cost of use and the frequency of repair and replacement of components.

[0026] Collecting multiple sets of device-related parameter data instances of different working conditions and different specifications for training machine learning algorithms enables the final trained model to learn various possible actual parameter combinations and the corresponding appropriate angle ranges. In this way, the configuration method is not limited to a specific working condition or device specification, but has strong versatility and can be widely used in new column guide sleeve automatic locking devices with different working environments and different manufacturing parameters, which enhances the applicability and flexibility of the device in actual engineering applications. With the help of machine learning algorithms, a large amount of actual data is analyzed and learned. Compared with the traditional empirical or theoretical calculation single configuration method, the complex nonlinear relationship between input parameters (such as pawl length, spring-related parameters, rope tension, etc.) and the target angle can be more comprehensively and meticulously excavated. By continuously optimizing the model parameters to minimize the prediction error, the appropriate pawl swing angle range can be accurately predicted according to the actual device parameters. This data-driven optimization configuration method can better meet the actual working conditions, obtain more accurate and effective configuration results, and further improve the performance of the device.

[0027] To sum up, through multi-faceted design and optimized configuration, this application has demonstrated significant beneficial effects in improving the operating accuracy, reliability, automatic locking and disassembly convenience, dust prevention and many other aspects of the device, and can better meet the performance requirements of the automatic locking device of the column guide sleeve in actual engineering applications. BRIEF DESCRIPTION OF THE DRAWINGS

[0028] Figure 1 It is an overall schematic diagram of the present invention;

[0029] Figure 2 It is a top view cross-sectional schematic diagram of the cylinder body of the present invention;

[0030] Figure 3 For the present invention Figure 2 A partial enlarged view of;

[0031] Figure 4 It is a bottom view schematic diagram of the guide sleeve of the present invention.

[0032] In the figure: 1-cylinder body, 2-guide sleeve, 3-convex ring, 4-annular cavity, 5-groove, 6-pawl, 7-movable plate, 8-fixed plate, 9-spring, 10-pull rope, 11-annular guide rail, 12-swivel, 13-anti-slip convex strip, 14-annular boss, 15-ratchet. DETAILED DESCRIPTION

[0033] See also Figure 1-Figure 4 A novel automatic locking device for a guide sleeve of a column comprises a cylinder body 1 and a guide sleeve 2, wherein the guide sleeve 2 is threadedly connected to the upper part of the cylinder body 1, an annular cavity 4 is provided in the upper inner part of the cylinder body 1, an outer thread is provided on the outer wall of the guide sleeve 2 and below the ratchet 15, an inner thread is provided on the inner wall of the cylinder body 1 and below the annular cavity 4, the outer thread is threadedly matched with the inner thread, and when the guide sleeve 2 is assembled with the cylinder body 1 of the column, the outer thread of the guide sleeve 2 can be threadedly connected with the inner thread of the cylinder body 1;

[0034] Two grooves 5 are symmetrically formed on the left and right sides of the upper end of the cylinder body 1, and the grooves 5 are communicated with the annular cavity 4. A ratchet 6 is rotatably connected inside the groove 5, and a movable plate 7 is fixed to one end of the ratchet 6. A fixed plate 8 is provided on one side of the movable plate 7. The lower end of the fixed plate 8 is fixed to the bottom of the groove 5. A spring 9 is connected between the movable plate 7 and the fixed plate 8. Because the spring 9 gives the ratchet 6 an elastic force, the claw of the ratchet 6 protrudes out of the groove 5.

[0035] A ratchet 15 is fixed to the outside of the guide sleeve 2, and the ratchet 15 is connected to the pawl 6. A pull rope 10 is fixed to the outer wall of the pawl 6, and the end of the pull rope 10 is movably penetrated by the cylinder body 1. An annular guide rail 11 is fixed to the outside of the cylinder body 1, and a swivel 12 is slidably connected to the inside of the annular guide rail 11. The end of the pull rope 10 is movably penetrated to the inside of the annular guide rail 11, and the end of the pull rope 10 is fixed to the inner wall of the swivel 12. When the guide sleeve 2 and the cylinder body 1 are threadedly connected to the ratchet 15 and the pawl 6, When in contact, the ratchet 15 will first push the pawl 6 away, so that the pawl 6 is retracted into the groove 5. Then, when the guide sleeve 2 stops rotating, the pawl 6 is connected to the ratchet 15 to prevent the guide sleeve 2 from rotating in the opposite direction and loosening, thereby playing a role of automatic locking. When the guide sleeve 2 needs to be removed, the swivel 12 is rotated to use the two pull ropes 10 to pull the pawl 6 to rotate and retract into the groove 5 and separate from the ratchet 15. Then, the guide sleeve 2 can be rotated in the opposite direction to disassemble the guide sleeve 2, thereby making the entire process of disassembling the guide sleeve 2 convenient and simple.

[0036] A convex ring 3 is fixed to the outside of the guide sleeve 2, and the convex ring 3 is located above the ratchet 15. When the guide sleeve 2 and the cylinder body 1 rotate until the convex ring 3 covers the cylinder body 1, the groove 5 can be covered to prevent dust from entering.

[0037] An annular groove is formed on both upper and lower ends of the inner wall of the annular guide rail 11, and an annular boss 14 is fixed on both upper and lower ends of the swivel 12. The annular boss 14 is located inside the annular groove and is slidably connected to the annular groove. When the swivel 12 rotates inside the annular guide rail 11, the annular boss 14 can be driven to slide inside the annular groove. Because there is a space between the inner wall of the swivel 12 and the inside of the annular guide rail 11, it is convenient for the pull rope 10 to enter the space. Therefore, the annular boss 14 and the annular groove can play a limiting role in the swivel 12.

[0038] The outer wall of the rotating ring 12 is evenly fixed with a plurality of anti-skid convex strips 13 in a ring shape. By arranging the anti-skid convex strips 13 on the outer wall of the rotating ring 12, an anti-skid effect can be achieved when the rotating ring 12 is rotated.

[0039] In order to improve the accuracy and reliability of the locking and disassembly operations of the pawl 6, it is necessary to accurately set the swing angle range of the pawl 6 in the groove 5. This angle range must ensure that the pawl 6 can be smoothly retracted into the groove 5 under the push of the ratchet 15 (when installing the guide sleeve 2), and can be reliably ejected and engaged with the ratchet 15 under the action of the spring 9 (when working normally and preventing loosening), and can be rotated to a position completely separated from the ratchet 15 under the pull of the pull rope 10 (when disassembling the guide sleeve 2). The specific configuration is as follows:

[0040] When the guide sleeve 2 is installed, the ratchet 15 pushes the pawl 6 into the groove 5. To ensure that the pawl 6 can be smoothly pushed into the groove 5 by the ratchet 15, it is necessary to satisfy that the torque generated by the ratchet 15 on the pawl 6 is greater than or equal to the sum of the elastic torque and the friction torque of the spring 9. The mathematical expression is:

[0041] F 推 *h 推 ≥k*x 0 +Δx*L 1 *cosα min +μ*F N *L 1 ;

[0042] where F N is the positive pressure between the pawl 6 and the groove 5, which can be determined by analyzing the overall force of the device and combining factors such as the gravity of related components. N It can be approximately considered to be related to the component of the spring 9 force in the direction perpendicular to the pawl 6, that is, F N =k*x 0 +Δx*sinα min ;

[0043] During normal operation and to prevent loosening, the pawl 6 pops out and engages with the ratchet 15 under the action of the spring 9. At this stage, the elastic torque of the spring 9 must be sufficient to overcome the friction torque so that the pawl 6 can be stably engaged with the ratchet 15. The relationship is: k*x 0 +Δx*L 1 *cosα max ≥μ*F N *L 1 ;

[0044] When the guide sleeve 2 is removed, the pull rope 10 pulls the pawl 6 and the ratchet 15 to separate. To make the pull rope 10 pull the pawl 6 and the ratchet 15 to separate completely, the torque generated by the pull rope 10 must be greater than the sum of the elastic torque of the spring 9 and the friction torque, that is: F 拉 *h 拉 ≥k*x 0 +Δx*L 1 *cosα max +μ*F N *L 1 ;

[0045] Solve the minimum swing angle α of the pawl 6 through machine learning algorithm min and the maximum swing angle α max , including the steps of:

[0046] S1. Collect multiple sets of parameter data examples of new column guide sleeve automatic locking devices with different working conditions and specifications. These examples can come from test records in the actual production process, experimental data under different laboratory simulation conditions, etc., to ensure that they cover a sufficient number of situations to reflect various possible parameter combinations. For each example, accurately measure and record the following parameters: the length L of the pawl 6 1 , the elastic coefficient k of spring 9, the initial compression amount x of spring 9 0 , the elongation Δx of the spring 9 when the pawl 6 swings to the limit position, and the pulling force F of the pull rope 10 on the pawl 6 拉 , tension F 拉 The vertical distance h between the action point and the rotation fulcrum of the pawl 6 拉 , the thrust F of the ratchet 15 on the pawl 6 推 Thrust F 推 The vertical distance h between the action point and the rotation fulcrum of the pawl 6 推 and the friction coefficient μ between the pawl 6 and the groove 5, and at the same time record the minimum swing angle α actually measured in the groove 5 of the corresponding pawl 6 min and the maximum swing angle α max All collected data are organized into a structured form, for example, a table format, where each row represents a set of parameter values ​​and corresponding angle values ​​corresponding to a different device instance, and each column corresponds to each of the above-mentioned parameters (L 1 , k, x 0 , Δx, F 拉 、h 拉 、F 推 、h 推 , μ, α min , α max ) to form a data set that can be used by machine learning algorithms. Carefully check each data point in the data set and remove those outliers that are obviously inconsistent with physical laws or caused by measurement errors. For example, if the value of the spring elastic coefficient in a set of data exceeds the normal value range of this type of spring by too much, or the recorded angle value appears to be obviously unreasonable to be extremely large or extremely small, etc., the corresponding set of data should be removed from the data set. At the same time, check whether there is any missing data. If there are a small number of missing values, you can use appropriate filling methods to process them according to the distribution characteristics of the data and the correlation with other parameters, such as using the mean of the corresponding parameters of the same type of device to fill the missing values, etc., to ensure the quality and integrity of the data set. Since the numerical ranges of different parameters collected are often quite different, in order to avoid adverse effects on the model due to differences in numerical size during the subsequent training of the machine learning algorithm (for example, features with larger values ​​may dominate optimization processes such as gradient descent during model training, making it difficult for features with smaller values ​​to play a role), it is necessary to input features (i.e., L1 , k, x 0 , Δx, F 拉 、h 拉 、F 推 、h 推 , μ) to perform feature scaling operations. The method of the embodiment of the present application is normalization, mapping the value of each parameter to a specific interval, such as the [0,1] interval. The specific normalization method is, for each parameter, subtract the minimum value of the parameter in all samples from its current value, and then divide it by the difference between the maximum and minimum values ​​of the parameter to obtain the normalized value, so that all features are in a relatively equal position in the subsequent model training, which is more conducive to the model to accurately learn the relationship between each parameter and the target angle.

[0047] S2. In the model selection stage, we need to consider multiple input parameters (L 1 , k, x 0 , Δx, F 拉 、h 拉 、F 推 、h 推 , μ) to predict two consecutive target angle values ​​(α min and α max ), and there may be a more complex nonlinear relationship between the input and the output. In the embodiment of the present application, the radial basis function kernel is used as the kernel function of the support vector regression model, and its formula is:

[0048] K(x i ,x j ) = exp(-γ||x i -x j || 2 );

[0049] In the formula, x i and x j Represent two sample vectors in the input space, respectively, and are represented by the length L of the ratchet 6 1 , the elastic coefficient k of spring 9, the initial compression amount x of spring 9 0 , the elongation Δx of the spring 9 when the pawl 6 swings to the limit position, and the pulling force F of the pull rope 10 on the pawl 6 拉 , tension F 拉 The vertical distance h between the action point and the rotation fulcrum of the pawl 6 拉 , the thrust F of the ratchet 15 on the pawl 6 推 Thrust F 推 The vertical distance h between the action point and the rotation fulcrum of the pawl 6 推 , a vector composed of the friction coefficient μ parameters between the pawl 6 and the groove 5.

[0050] Among them, γ is the width parameter (γ>0), which controls the width of the kernel function. It is initially set to 0.1 in this application. It controls the distribution of sample data after being mapped to high-dimensional space through the kernel function. Specifically, the value of γ affects the width of the function curve. A smaller γ value will make the Gaussian curve corresponding to the kernel function relatively flat and wide, which means that the influence range between different sample vectors is larger, and the model's discrimination of data in high-dimensional space is relatively less precise. It may be relatively smooth when fitting data, but some detailed information may be lost; while a larger γ value will make the Gaussian curve sharper and narrower, making the influence range between sample vectors smaller. The model is more sensitive to changes in data in high-dimensional space and can capture more delicate data features, but it is also easy to cause overfitting, that is, it fits too closely to the subtle features in the training data and loses the ability to generalize to new data.

[0051] ||x i -x j || 2 Represents the sample vector x i and x j The square of the Euclidean distance between them (that is, the sum of the squares of the differences between the corresponding elements of the two vectors). The characteristic of this kernel function is that it can map the input space to an infinite-dimensional high-dimensional space, and it has a good effect on processing complex nonlinear relationships. It is widely used in many practical nonlinear regression and classification problems. It can help the model to find a suitable regression hyperplane in the high-dimensional space to fit the ratchet swing angle data based on the complex relationship between the device parameters (input vectors corresponding to different samples). These kernel functions change the distribution form and relationship of the data by mapping the data of the original input space to the high-dimensional space, so that the support vector regression model can handle nonlinear problems of different complexities, and then learn the mapping law between input and output (such as the device parameters and the ratchet swing angle in this application) based on the training data, so as to achieve accurate prediction of unknown data.

[0052] x i and x j Represent two sample vectors in the input space respectively. In this application, they are the length L of the ratchet 6. 1 , the elastic coefficient k of spring 9, the initial compression amount x of spring 9 0 , the elongation Δx of the spring 9 when the pawl 6 swings to the limit position, and the pulling force F of the pull rope 10 on the pawl 6 拉 , tension F 拉 The vertical distance h between the action point and the rotation fulcrum of the pawl 6 拉 , the thrust F of the ratchet 15 on the pawl 6 推 Thrust F 推 The vertical distance h between the action point and the rotation fulcrum of the pawl 6 推, the friction coefficient μ between the pawl 6 and the groove 5 (in vector form after normalization and other preprocessing). For example, for the parameter conditions corresponding to two different device samples, x i and x j To represent its corresponding parameter vector.

[0053] ||x i -x j || 2 Represents the sample vector x i and x j The square of the Euclidean distance between the two vectors is calculated by first calculating the difference between the corresponding elements of the two vectors, and then squaring these differences and summing them up, that is, (Here n represents the dimension of the vector, that is, the number of parameters, which in this application is the number of parameters related to the device mentioned above; x ik Represents vector x i The kth element of jk Similarly). Through this kernel function, the SVR model can map the original input space (the space composed of device-related parameters) to a high-dimensional space, and find an optimal regression hyperplane in the high-dimensional space to better fit the swing angle α of the pawl 6. min and α max Output data, process the complex nonlinear relationship between input parameters and output angle, so as to achieve the purpose of accurately predicting the pawl swing angle based on the input device parameters.

[0054] The preprocessed data set is divided into a training set and a test set according to a certain ratio. Usually 80% of the data is used as the training set and 20% of the data is used as the test set. The purpose of this division is to allow the model to learn and train on most of the data, grasp the rules between the input features and the target angle, and then use the remaining small part of the data that has not participated in the training (test set) to test the model's ability to predict new data, that is, the generalization ability of the model, so as to determine whether the model is overfitting or underfitting. For example, if there are 1,000 groups of data samples, 800 groups are selected as training sets for training the SVR model, and the remaining 200 groups are used as test sets for subsequent model evaluation.

[0055] S3. Start the training process and transfer the training set data (including the normalized L 1 , k, x 0 , Δx, F 拉 、h 拉 、F 推 、h 推 , μ and the corresponding α min and α max) is input into the constructed SVR model, and the model training process is started according to the set model parameters (γ). During the training process, the SVR model will continuously adjust its internal parameters (such as the coefficients corresponding to the support vector, etc.) according to the training data to make the prediction error of the model on the training set as small as possible, that is, to make the predicted angle value (α min and α max ) is as close as possible to the actual measured angle value. This optimization adjustment process is based on the SVR optimization algorithm and is achieved by minimizing the loss function. The trained SVR model is evaluated using the test set data. The input feature data of the test set (normalized parameter values) is input into the model to obtain the predicted α min and α max The value is then compared with the actual angle value in the test set. A variety of evaluation indicators are used to measure the prediction accuracy and stability of the model, including mean square error (MSE) and mean absolute error (MAE). The mean square error is calculated by summing the squares of the difference between the predicted value and the true value of each sample and taking the average. It is more sensitive to larger errors and can highlight the performance of the model in terms of prediction accuracy; the mean absolute error is the sum of the absolute values ​​of the difference between the predicted value and the true value of each sample and taking the average. It relatively more intuitively reflects the average deviation between the predicted value and the true value. The performance of the model is judged by calculating the values ​​of these evaluation indicators. If the indicator value is large, it means that the model may still have problems such as poor fitting and poor generalization ability, and further optimization and adjustment are needed. If the model evaluation effect is not ideal, adjust the key parameters of the SVR model, such as changing the value of γ, and use grid search, random search and other methods to systematically find the optimal parameter combination. Grid search is to traverse all possible parameter combinations within the pre-set parameter value range at a certain step length, train the model and evaluate each combination, and finally select the parameter combination with the best performance; random search is to randomly select a certain number of parameter combinations within the parameter value range for trial, which is relatively more time-saving and suitable for situations with a large parameter space. By continuously adjusting the parameters and retraining and evaluating, observing whether the evaluation indicators have improved, and gradually optimizing the model performance, a model with better performance and higher prediction accuracy is obtained, which can accurately predict the minimum swing angle α of the pawl 6 based on the input device parameters. min and the maximum swing angle α max After obtaining a model that meets the accuracy requirements, the actual L 1 , k, x 0 , Δx, F 拉 、h 拉 、F 推 、h 推, μ parameter values, process these parameters according to the previous data preprocessing method (such as normalization), and then input the processed parameter data into the trained model, and the model will output the predicted α min and α max The predicted values ​​are used to configure the swing angle range of the pawl 6 in the groove 5, ensuring the normal operation and reliability of the device in each operation stage (installation, normal operation, disassembly, etc.), and achieving better working performance of the entire automatic locking device. Through the process based on the machine learning algorithm, combined with the given mathematical model, the minimum swing angle α of the pawl 6 in the groove 5 can be solved more accurately. min and the maximum swing angle α max , thereby optimizing the configuration of the device and improving its performance in practical applications.

Claims

1. A novel column guide sleeve automatic locking device, comprising a cylinder body (1) and a guide sleeve (2), wherein the guide sleeve (2) is threadedly connected to the upper part of the cylinder body (1), characterized in that: An annular cavity (4) is provided at the upper inner part of the cylinder body (1), and two grooves (5) are symmetrically provided on the left and right sides of the upper end of the cylinder body (1), and the grooves (5) are communicated with the annular cavity (4). A ratchet (6) is rotatably connected inside the groove (5), and a ratchet (15) is fixed outside the guide sleeve (2), and the ratchet (15) is connected to the ratchet (6). A pull rope (10) is fixed to the outer wall of the ratchet (6), and the end of the pull rope (10) is movably penetrated through the cylinder body (1), and an annular guide rail (11) is fixed outside the cylinder body (1), and a swivel (12) is slidably connected inside the annular guide rail (11), and the end of the pull rope (10) is movably penetrated into the inside of the annular guide rail (11), and the end of the pull rope (10) is fixed to the inner wall of the swivel (12).

2. According to claim 1, a novel automatic locking device for a column guide sleeve is characterized in that: An external thread is provided on the outer wall of the guide sleeve (2) and below the ratchet (15), and an internal thread is provided on the inner wall of the cylinder body (1) and below the annular cavity (4), and the external thread matches the internal thread.

3. According to claim 1, a novel automatic locking device for a column guide sleeve is characterized in that: A convex ring (3) is fixed to the outside of the guide sleeve (2), and the convex ring (3) is located above the ratchet (15).

4. According to claim 1, the novel automatic locking device for column guide sleeve is characterized in that: A movable plate (7) is fixed to one end of the pawl (6), a fixed plate (8) is provided on one side of the movable plate (7), the lower end of the fixed plate (8) is fixed to the bottom of the groove (5), and a spring (9) is connected between the movable plate (7) and the fixed plate (8).

5. According to claim 1, the novel automatic locking device for the column guide sleeve is characterized in that: The inner wall of the annular guide rail (11) is provided with an annular groove at both upper and lower ends, and an annular boss (14) is fixed at both upper and lower ends of the rotating ring (12). The annular boss (14) is located inside the annular groove and is slidably connected to the annular groove.

6. The novel automatic locking device for column guide sleeve according to claim 1 is characterized in that: The outer wall of the rotating ring (12) is annular and has a plurality of anti-slip convex strips (13) evenly fixed thereon.

7. The novel automatic locking device for column guide sleeve according to claim 1 is characterized in that: The swing angle range of the pawl (6) is configured as follows: The minimum swing angle of the pawl (6) in the groove (5) is α min , the maximum swing angle is α max ; The minimum swing angle is α min The mathematical model is: F 推 *h 推 ≥k*x0+Δx*L1*cosα min +μ*F N *L1; The maximum swing angle is α max The mathematical model is: F 拉 *h 拉 ≥k*x0+Δx*L1*cosα max +μ*F N *L1; Wherein, L1 is the length of the pawl (6), that is, the distance from the rotation fulcrum to the end of the pawl, k is the elastic coefficient of the spring (9), x0 is the initial compression of the spring (9), Δx is the elongation of the spring (9) when the pawl (6) swings to the limit position, and F 拉 is the pulling force of the pull rope (10) on the ratchet (6), h 拉 is the tension F 拉 The vertical distance between the action point and the rotation fulcrum of the pawl (6), F 推 is the thrust of the ratchet (15) on the pawl (6), h 推 is the thrust F 推 The vertical distance between the action point and the rotation fulcrum of the pawl (6), μ is the friction coefficient between the pawl (6) and the groove (5), F N is the positive pressure between the pawl (6) and the groove (5), and the minimum swing angle of the pawl (6) in the groove (5) is α min , the maximum swing angle is α max .

8. The novel automatic locking device for column guide sleeve according to claim 7 is characterized in that: The minimum swing angle α of the pawl (6) is solved by machine learning algorithm min and the maximum swing angle α max , including the steps of: S1. Collect multiple sets of parameter data examples of column guide sleeve automatic locking devices of different working conditions and specifications. For each example, record the following parameters: length L1 of the pawl (6), elastic coefficient k of the spring (9), initial compression x0 of the spring (9), elongation Δx of the spring (9) when the pawl (6) swings to the limit position, and tension F of the pull rope (10) on the pawl (6). 拉 , tension F 拉 The vertical distance h between the action point and the rotation fulcrum of the pawl (6) 拉 , the thrust F of the ratchet wheel (15) on the pawl (6) 推 Thrust F 推 The vertical distance h between the action point and the rotation fulcrum of the pawl (6) 推 and the friction coefficient μ between the pawl (6) and the groove (5), and simultaneously record the minimum swing angle α actually measured by the corresponding pawl (6) in the groove (5) min and the maximum swing angle α max , organize all the collected data into a structured form, clean the data, and normalize it; S2. Use the radial basis function kernel as the kernel function of the support vector regression model, and its formula is: K(x i ,x j )=exp(-γ||x i -x j || 2 ); Where γ is the width parameter, and γ>0, and the initial setting is γ=0.1; x i and x j Represent two sample vectors in the input space respectively; S3. Use part of the cleaned and normalized preprocessed data set as the training set and the other part as the test set. Input the training set data into the constructed support vector regression model. Start the model training according to the set model parameters. Use the test set data to evaluate the trained SVR model. Input the input feature data of the test set into the model to obtain the predicted α. min and α max The value is then compared with the actual angle value in the test set. By continuously adjusting the parameters and retraining and evaluating, the model performance is optimized until the minimum swing angle α of the ratchet (6) is predicted based on the input device parameters. min and the maximum swing angle α max , after obtaining a model that meets the accuracy requirements, for the new device, obtain its actual L1, k, x0, Δx, F 拉 、h 拉 、F 推 、h 推 , μ parameter values, after cleaning and normalization, the processed parameter data is input into the trained model, and the model will output the predicted α min and α max Value, based on the predicted α min and α max The swing angle range of the ratchet pawl (6) in the groove (5) is configured according to the value.

Images