Metal wire bending prediction method, device and medium based on clamp holding posture detection

Abstract

The embodiments of this specification disclose a wire bending prediction method, device and medium based on the detection of the holding forceps posture, which relates to the field of digital intelligence monitoring technology and is used to solve the problem that the existing method is difficult to predict the next wire bending operation in time, resulting in poor training effect of holding forceps. The method includes: obtaining the initial posture data of the mold to be trained in the current training scene, and performing feature extraction; planning the deformation path of the mold to be trained based on the extracted feature data and the preset standard structure model; determining the operation position data of the bending forceps based on the path, and predicting the predicted force data of each operation position in combination with the adjacent operation position data and the historical training data. Multi-source data of the wire bending process are collected by using multiple types of sensors on the bending forceps to determine the current force data of the bending forceps; the current force data is compared with the predicted force data, and the next force data of the bending operation is determined according to the comparison result, so as to guide the completion of the wire bending operation according to the next force data.

Description

Technical Field

[0001] The present invention relates to the field of data processing technology, and in particular to a method, device and medium for predicting wire bending based on pliers holding posture detection. Background Art

[0002] Wire bending is currently widely used in many fields. For example, in the field of medical equipment, flexible metal wires such as those made of stainless steel, pure titanium, titanium alloys, etc. are widely used to fix or assist in fixation of fractures, as well as in scenarios such as orthodontics due to their high flexibility, high strength, excellent corrosion resistance, and biocompatibility. In the electrical field, when manufacturing electronic equipment, such as circuit boards and integrated circuits, it is often necessary to bend metal wires or metal sheets into specific shapes to connect different electronic components, thereby improving the manufacturing accuracy and performance of electronic equipment. The current wire bending process mainly relies on manual operation by operators, which requires workers to have rich experience and skills to determine the pliers holding posture to bend the wire. Therefore, the operator's bending skills directly affect the quality of wire bending.

[0003] Currently, in order to improve the quality and accuracy of wire bending by operators, a training platform is provided to enable operators to train wire bending operations using training molds in required scenarios. However, traditional monitoring of wire bending operations involves testing the finished product after processing. Since the test feedback is only provided after the wire bending operation is completed, it is difficult to correct the operator's mistakes during the training process in a timely manner, which may lead to the formation of bad habits. In addition, if the finished product is unqualified, it means a waste of time and materials. Especially during the training process, the number of unqualified finished products may be considerable, thereby increasing the training cost. The method based on real-time guidance of the operator cannot adapt to the training scenario of multiple operators, and the labor cost is high. Therefore, how to predict the next wire bending operation in a timely manner to assist the operator in mastering the clamp holding posture and force data is a technology that needs to be solved. Summary of the Invention

[0004] In order to solve the above technical problems, one or more embodiments of this specification provide a method, device and medium for predicting wire bending based on clamp holding posture detection.

[0005] One or more embodiments of this specification adopt the following technical solutions:

[0006] One or more embodiments of this specification provide a wire bending prediction method based on pliers holding posture detection, the method comprising:

[0007] Acquire initial posture data of the mold to be trained in the current training scene, perform feature extraction on the initial posture data, and determine feature data of the mold to be trained;

[0008] Performing deformation path planning on the mold to be trained based on the feature data and a preset standard structure model to determine the deformation path corresponding to the mold to be trained;

[0009] Determining operating position data of the bending pliers based on the deformation path, and determining predicted force data corresponding to each operating position data of the bending pliers according to adjacent operating position data and historical training data;

[0010] Acquiring multi-source data of a wire bending process based on multiple types of sensors pre-installed on the bending pliers, and determining current force data of the bending pliers based on the multi-source data; wherein the current force data includes: force direction, force magnitude, and gripping posture;

[0011] The current force data is compared with the predicted force data to determine the next force data of the bending operation according to the comparison result, and the next force data is transmitted to the current terminal to guide the completion of the wire bending operation.

[0012] Optionally, in one or more embodiments of the present specification, performing deformation path planning on the mold to be trained based on the feature data and a preset standard structure model to determine the deformation path corresponding to the mold to be trained specifically includes:

[0013] Based on pre-deformation feature data corresponding to historical deformation data stored in a preset database, determining historical deformation data that matches the feature data, and determining a standard structural model corresponding to the feature data based on the historical deformation data;

[0014] Determining a deformation target state corresponding to the mold to be trained based on difference data between the standard structural model and the mold to be trained;

[0015] The arrangement state of the structure to be deformed corresponding to the mold to be trained is used as the initial solution of the path, and the annealing parameters are initialized; wherein the annealing parameters include: an initial temperature corresponding to the exploration range of the initial state of the mold to be trained, and a temperature attenuation factor corresponding to the path accuracy;

[0016] Establishing a simulation model of the structure to be deformed of the mold to be trained based on the three-dimensional data of the mold to be trained, so as to adjust the simulation model of the structure to be deformed based on the initial temperature;

[0017] Calculating difference data between the adjusted simulation model of the structure to be deformed and the deformation target state, adjusting the annealing parameters according to the difference data, and using the arrangement state of the structure to be deformed corresponding to the adjusted simulation model of the structure to be deformed as the current solution of the path;

[0018] The current solution of the path is iteratively adjusted according to the adjusted annealing parameters to obtain a final solution of the path, and the final solution of the path is used as the deformation path corresponding to the mold to be trained.

[0019] Optionally, in one or more embodiments of this specification, determining the forceps operation position data based on the deformation path specifically includes:

[0020] Determining the deformation path and the arrangement state of the deformation structure corresponding to the mold to be trained, the deformation structure adjustment data corresponding to each deformation structure; wherein the deformation structure adjustment data includes: rotation adjustment, tilt adjustment, and movement adjustment;

[0021] By comparing the adjustment data of adjacent structures to be deformed, the relative position and posture transformation data of each structure to be deformed are determined;

[0022] determining the coordinate position of the wire bending point and the coordinate position of the clamping point based on the relative position and the posture transformation data;

[0023] The coordinate positions of the wire bending points and the clamping points are summarized according to a preset traversal direction to determine the clamp operation position data; wherein the preset traversal direction includes: clockwise direction and counterclockwise direction.

[0024] Optionally, in one or more embodiments of the present specification, determining the standard force data corresponding to each operating position data of the bending pliers according to the adjacent operating position data and the historical deformation training data specifically includes:

[0025] Determining a movement trajectory of the bending pliers at the adjacent operation positions by using the adjacent operation position data and a preset initial direction, so as to determine a force application direction of the bending pliers based on the movement trajectory;

[0026] Inputting the historical deformation training data into a preset deep learning network for training to obtain a force prediction model;

[0027] Inputting the force direction and the adjacent operation position data into the force prediction model to obtain the force magnitude and holding posture of the bending pliers;

[0028] The force direction, force magnitude and holding posture of the bending pliers are summarized to determine standard force data corresponding to each operation position data of the bending pliers.

[0029] Optionally, in one or more embodiments of the present specification, obtaining multi-source data of the wire bending process according to multiple types of sensors pre-installed on the bending pliers, and determining the current force data of the bending pliers according to the multi-source data, specifically includes:

[0030] Acquire initial multi-source data collected by multiple types of sensors pre-placed on the bending clamp; wherein the multi-source data includes: pressure data, displacement data and angle data;

[0031] According to the pressure data, obtaining the force applied by the bending pliers;

[0032] determining a pressure concentration point according to the pressure distribution corresponding to the pressure data, and obtaining a holding posture of the bending pliers according to the pressure concentration point and the angle data;

[0033] The force direction of the bending pliers is acquired according to the displacement data, and the current force data of the bending pliers is determined based on the force magnitude, the holding posture and the force direction.

[0034] Optionally, in one or more embodiments of the present specification, comparing the current force data with the predicted force data to determine the next force data for the bending operation according to the comparison result specifically includes:

[0035] comparing the current force data with the predicted force data to determine a deviation between the current force data and the predicted force data;

[0036] If the deviation value is within the preset deviation range, acquiring the next operation position data adjacent to the operation position data corresponding to the predicted force data, and using the predicted force data corresponding to the next operation position data as the next force data for the bending operation;

[0037] If the deviation value exceeds the preset deviation range, adjustment data of the current force data is determined according to the deviation value, so as to use the adjustment data as next force data for the bending operation.

[0038] Optionally, in one or more embodiments of this specification, after comparing the current force data with the predicted force data, the method further includes:

[0039] Determining the number of deviation parameters and the magnitude of the deviation values ​​between the current force data and the predicted force data according to a comparison result between the current force data and the predicted force data;

[0040] Evaluate the wire bending operation of the operation position data based on the number of deviation parameters and the magnitude of the deviation value, and store the evaluation result in a storage unit corresponding to the bending pliers; wherein the bending pliers have a corresponding label, and the label is associated with the ID of the current operator;

[0041] Recalling the evaluation result in the storage unit according to the ID of the current operator to determine the operation position data of the current operator that exceeds the preset deviation range;

[0042] The operation position data exceeding the preset deviation range is transmitted to the preset evaluation terminal, so that the preset evaluation terminal determines a training plan for the current operator in combination with the operation position data exceeding the preset deviation range.

[0043] Optionally, in one or more embodiments of the present specification, obtaining initial posture data of the mold to be trained, performing feature extraction on the initial posture data, and determining feature data of the mold to be trained specifically includes:

[0044] Acquire three-dimensional data of the mold to be trained, convert the three-dimensional data into two-dimensional slice data of each plane in a three-dimensional coordinate system, and calculate the gradient of the two-dimensional slice data in each direction according to a preset gradient operator;

[0045] The gradients corresponding to the two-dimensional slice data of each plane are combined to obtain an edge intensity map;

[0046] Extracting edges of the structure to be deformed in the mold to be trained based on the edge strength map, and connecting the edges of the structure to be deformed according to an edge connection algorithm to obtain edge contours of the structure to be deformed corresponding to the two-dimensional slice data of each plane;

[0047] Performing three-dimensional reconstruction on the edge contours corresponding to the two-dimensional slice data of each plane to obtain three-dimensional edge contour information of the structure to be deformed, segmenting the region of the structure to be deformed of the mold to be trained based on the three-dimensional edge contour information, and determining initial posture data of the region of the structure to be deformed based on the segmentation result;

[0048] Identify key feature points of the mold to be trained based on the initial posture data and the attribute characteristics of each key point; wherein the key feature points include: vertices of the structure to be deformed, centers of grooves, and boundaries between the structure to be deformed and the fixed structure;

[0049] Based on the key feature points, the preset coordinate system and the reference point, the feature data of the mold to be trained is determined; wherein the feature data includes: pressing position coordinate data, tilt angle data of the structure to be deformed, and rotation angle data of the structure to be deformed.

[0050] One or more embodiments of this specification provide a wire bending prediction device based on pliers holding posture detection, including:

[0051] at least one processor; and,

[0052] a memory communicatively connected to the at least one processor; wherein,

[0053] The memory stores instructions that can be executed by the at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to: perform any of the above methods.

[0054] One or more embodiments of this specification provide a non-volatile computer storage medium storing computer-executable instructions, wherein the computer-executable instructions are configured to execute any of the above-described methods.

[0055] At least one of the above technical solutions adopted in the embodiments of this specification can achieve the following beneficial effects:

[0056] By acquiring the initial posture data of the mold to be trained, key information can be extracted as feature data. Deformation path planning is then performed based on this feature data and the standard structural model, ensuring that the deformation process of the mold to be trained is carried out within expected requirements. Determining the operating positions of the bending clamp based on the deformation path and obtaining predicted force data corresponding to each operating position facilitates pre-planning of force application strategies. By utilizing multiple sensors on the bending clamp to acquire multi-source data in real time, precise force application during the bending process can be obtained. Comparing current force data with predicted force data allows for timely identification of operational deviations and allows for adjustment of the next force data based on the comparison results. This facilitates dynamic adjustment and optimization of the operator's bending operation, ensuring that the bending process always follows the correct path, thus ensuring operator training while maintaining product quality. Transmitting the next force data to the current terminal provides real-time guidance for operator adjustments and targeted training. This real-time feedback mechanism helps operators correct errors promptly, improving their skills and bending quality. BRIEF DESCRIPTION OF THE DRAWINGS

[0057] In order to more clearly illustrate the embodiments of this specification or the technical solutions in the prior art, the following briefly introduces the drawings required for the embodiments or the description of the prior art. Obviously, the drawings described below are only some of the embodiments described in this specification. For those skilled in the art, other drawings can be obtained based on these drawings without inventive work. In the drawings:

[0058] Figure 1A schematic flow chart of a method for predicting wire bending based on pliers holding posture detection provided in an embodiment of this specification;

[0059] Figure 2 A schematic diagram of a process for determining the next force application data in an application scenario provided in an embodiment of this specification;

[0060] Figure 3 A schematic diagram of the structure of a wire bending prediction device based on pliers holding posture detection provided in an embodiment of this specification;

[0061] Figure 4 A schematic diagram of the structure of a non-volatile storage medium provided in an embodiment of this specification. DETAILED DESCRIPTION

[0062] The embodiments of this specification provide a method, device, and medium for predicting wire bending based on clamp holding posture detection.

[0063] To help those skilled in the art better understand the technical solutions in this specification, the following will provide a clear and complete description of the technical solutions in the embodiments of this specification, in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of this specification, not all of them. All other embodiments obtained by those skilled in the art based on the embodiments of this specification without creative work should fall within the scope of protection of this specification.

[0064] like Figure 1 As shown, the embodiment of this specification provides a schematic flow chart of a method for predicting wire bending based on clamp holding posture detection, which is composed of Figure 1 It can be seen that in one or more embodiments of this specification, a wire bending prediction method based on clamp holding posture detection includes the following steps:

[0065] S101: Acquire initial posture data of a mold to be trained in a current training scene, perform feature extraction on the initial posture data, and determine feature data of the mold to be trained.

[0066] To facilitate metal bending training for operators in various scenarios, the embodiments of this specification acquire the initial posture data of the mold to be trained within the current training scenario. For example, if the current training scenario is an electrician's training practice, electronic devices such as circuit boards or integrated circuits are used as the training mold. If the current training scenario is a dental brace in a medical setting, a dental model is used as the training mold. After obtaining the initial posture data of the mold to be trained, in order to facilitate the determination of a standardized deformation path for wire bending, the embodiments of this specification first extract features from the initial posture data to determine the feature data of the mold to be trained, so that path planning can be performed based on this feature data.

[0067] Specifically, in one or more embodiments of the present specification, obtaining initial posture data of a mold to be trained, performing feature extraction on the initial posture data, and determining feature data of the mold to be trained specifically include the following process:

[0068] For complex molds to be trained, which may include structures that need to be deformed by wires and other fixed structures, in order to analyze the structures that need to be deformed, the embodiments of this specification obtain 3D data of the mold to be trained, convert the 3D data into 2D slice data for each plane in a 3D coordinate system, and then calculate the gradient of the 2D slice data in each direction using a preset gradient operator. By obtaining the 3D data of the mold to be trained and converting it into 2D slice data for each plane, this process can capture the mold's geometry and details with high precision. Then, to more easily identify the edges of the structure to be deformed in the mold, the gradients corresponding to the 2D slice data for each plane are combined to obtain an edge strength map. For example, in a certain scenario, for each plane's 2D slice data, a preset gradient operator such as the Sobel operator or the Prewitt operator is used to calculate the gradient of the 2D slice data in the x and y directions. The gradient operator returns a matrix of the same size as the original slice data, where each element represents the gradient value in the x or y direction at the corresponding position. Then, for each 2D slice, the gradient values ​​in the x and y directions are combined using the Pythagorean theorem to calculate the gradient amplitude. Then the gradient amplitudes of all two-dimensional slices are combined to form a new matrix image, which is the edge intensity map. Each pixel value in the edge intensity map represents the gradient amplitude at the corresponding position, reflecting the strength of the edge at that position.

[0069] After obtaining the edge intensity map, the edges of the structure to be deformed in the mold to be trained are extracted based on the edge intensity map, and the edges of the structure to be deformed are connected according to the edge connection algorithm to obtain the edge contours of the structure to be deformed corresponding to the two-dimensional slice data of each plane. The edge contours corresponding to the two-dimensional slice data of each plane are then reconstructed in three dimensions to obtain the three-dimensional edge contour information of the structure to be deformed. The region of the structure to be deformed in the mold to be trained is segmented based on the three-dimensional edge contour information, and the initial posture data of the region of the structure to be deformed is determined based on the segmentation results. It should be noted that during the three-dimensional reconstruction process, the edge contour data on each slice is extracted, and then the corresponding contour lines on adjacent slices are connected using small triangles to form a curved surface to obtain the three-dimensional edge contour information. The key feature points of the mold to be trained are then identified based on the initial posture data and the attribute characteristics of each key point. It should be noted that the key feature points include: the vertices of the structure to be deformed, the center of the groove, and the boundary between the structure to be deformed and the fixed structure. Based on the key feature points obtained above and the preset coordinate system and reference points, the feature data of the mold to be trained is determined; wherein, the feature data includes: pressing position coordinate data, inclination angle data of the structure to be deformed, and rotation angle data of the structure to be deformed. It can be understood that the pressing position coordinate data is used to determine the relative position and arrangement of the structure to be deformed, the inclination angle data can be determined based on the angle between the vertex of the structure to be deformed and the reference line where the reference point is located after being projected onto a specific plane, and the rotation angle data can be determined based on the projection of the key feature points in different directions of the coordinate system. Then, the area of ​​the structure to be deformed of the mold to be trained is segmented based on the three-dimensional edge contour information, which can accurately identify the key deformation area in the mold. At the same time, by identifying the key feature points, the feature data of the mold, such as the pressing position coordinates, inclination angle and rotation angle, etc., can be further determined, providing a data basis for subsequent metal bending path planning and deformation control.

[0070] S102: performing deformation path planning on the mold to be trained based on the feature data and a preset standard structure model to determine the deformation path corresponding to the mold to be trained.

[0071] After obtaining the characteristic data of the mold to be trained based on the above step S101, the embodiment of this specification will plan the deformation path of the mold to be trained based on the acquired characteristic data and its standard structural model, thereby determining the deformation path corresponding to the mold to be trained. Through precise deformation path planning, the deformation process that meets the quality requirements can be determined, which helps to assist and guide the operator in bending the metal wire subsequently, so that the bending of the metal wire meets the quality requirements.

[0072] Specifically, in one or more embodiments of this specification, deformation path planning is performed on the mold to be trained based on the feature data and the preset standard structure model to determine the deformation path corresponding to the mold to be trained, which specifically includes the following process:

[0073] First, based on the pre-deformation feature data corresponding to the historical deformation data stored in the preset database, the historical deformation data that matches the feature data is determined, and at the same time, the standard structural model corresponding to the feature data is determined based on the historical deformation data. Then, based on the difference data between the standard structural model and the mold to be trained, the deformation target state corresponding to the mold to be trained is determined. For example, a metal part of a specific shape needs to be processed. There is a preset database that stores historical deformation data of the mold to be trained in the past, including pre-deformation feature data such as mold size, material type, etc. and corresponding deformation data. By comparing the difference data between the standard structural model and the mold to be trained, that is, the metal part of the specific shape, the target state that the mold to be trained should reach after deformation, that is, the required part shape, can be determined.

[0074] The arrangement state of the structure to be deformed corresponding to the mold to be trained is then used as the initial path solution, and annealing parameters are initialized. The annealing parameters include an initial temperature corresponding to the exploration range of the initial state of the mold to be trained and a temperature attenuation factor corresponding to the path accuracy. A simulation model of the structure to be deformed of the mold to be trained is then established based on the three-dimensional data of the mold to be trained, and the simulation model is adjusted based on the initial temperature. The difference between the adjusted simulation model of the structure to be deformed and the target state is calculated, and the annealing parameters are adjusted based on this difference. The arrangement state of the structure to be deformed corresponding to the adjusted simulation model of the structure to be deformed is then used as the current path solution. The current path solution is iteratively adjusted based on the adjusted annealing parameters to obtain a final path solution, which is then used as the deformation path corresponding to the mold to be trained.

[0075] In this process, by matching historical deformation data, a standard structural model corresponding to the current feature data can be accurately found. The difference data between the standard structural model and the mold to be trained can then be used to precisely determine the target deformation state of the mold to be trained, thereby improving the accuracy of deformation prediction. Furthermore, the introduction of annealing parameters allows greater flexibility in searching for the optimal solution. The initial temperature determines the breadth of the search space, while the temperature attenuation factor controls the search precision and speed. Therefore, by establishing a simulation model of the structure to be deformed and adjusting it based on the initial temperature, the search space can be rapidly narrowed, improving solution efficiency. The iterative adjustment process then continuously approaches the optimal solution, ultimately obtaining the deformation path of the mold to be trained, ensuring controllability and stability of the deformation process. By continuously adjusting the difference data between the simulation model and the target deformation state, the resulting deformation path is ensured to be highly consistent with the desired target. This not only improves production efficiency but also reduces reliance on manual experience, making deformation prediction more scientific and reliable.

[0076] S103: determining the operation position data of the bending pliers based on the deformation path, and determining the predicted force data corresponding to each operation position data of the bending pliers according to adjacent operation position data and historical training data.

[0077] To ensure that the bending pliers can move along a predetermined path during the bending process, the operator's skill mastery is improved while reducing the cost of training materials, unnecessary force waste, and possible damage. In the embodiments of this specification, the operating position of the bending pliers is determined based on the deformation path. At the same time, the predicted force data corresponding to each operating position data of the bending pliers is determined based on the adjacent operating position data and historical training data, so as to facilitate the subsequent real-time correction and adjustment of the operator's operation based on the predicted force data. It can be understood that during the wire bending process, the operating position data of the bending pliers includes the coordinate position of the wire bending point and the coordinate position of the clamping point.

[0078] Specifically, in one or more embodiments of this specification, determining the clamp holding operation position data based on the deformation path specifically includes:

[0079] First, based on the deformation path and the alignment of the deformable structure corresponding to the training mold, the corresponding deformation adjustment data for each deformable structure is determined. This deformation adjustment data includes rotation adjustment, tilt adjustment, and translation adjustment. Then, by comparing the adjustment data of adjacent deformable structures, the relative position and posture transformation data of each deformable structure are determined. Based on the relative position and posture transformation data, the coordinate positions of the wire bending points and the clamping points are determined. The coordinate positions of each wire bending point and each clamping point are then summarized according to a preset traversal direction to determine the forceps operating position data. Preset traversal directions include clockwise and counterclockwise. For example, in an orthodontic scenario, the training mold can be likened to the alignment of teeth, and the deformation path represents the trajectory of the teeth from the current alignment to the ideal alignment. If the anterior area requires adjustment, one of the front teeth, referred to as deformation structure A, needs to be rotated inward (rotational adjustment), tilted slightly downward (tilt adjustment), and moved forward (translation adjustment) to achieve the ideal alignment. Based on the tooth deformation path—the direction and distance of tooth movement—and the current tooth alignment, the required adjustment data for deformed structure A, specifically rotational, tilt, and translation adjustments, can be calculated. The adjustment data for adjacent teeth, such as those next to deformed structure A, are then compared. By comparing the rotation, tilt, and translation adjustment data for tooth A and tooth B, their relative position and posture transformation data can be determined. For example, if tooth A rotates inward 5 degrees while tooth B remains in place, there is a 5-degree rotation difference between tooth A and tooth B. Based on this relative position and posture transformation data, the operator can determine the necessary bends on the wire and the clamping points used to secure the wire. These points ensure that the wire applies the correct force to the tooth along the predetermined deformation path, thereby achieving tooth movement. For example, if tooth A needs to be moved forward, the wire's bend point at tooth A may need to be bent forward to provide a forward pulling force. Furthermore, clamping points are installed on tooth A to ensure the wire is securely fixed to the tooth. Finally, based on the preset traversal direction, such as clockwise or counterclockwise, the coordinate positions of all wire bending and clamping points are summarized to determine the clamp operation position data. This data is used to correct the operator's bending operation in real time.

[0080] In this process, not only the rotation and tilt adjustments of the structure to be deformed are taken into account, but also the movement adjustments, so that the various changes of the structure to be deformed during the deformation process can be fully and meticulously reflected. By comparing the adjustment data of adjacent structures to be deformed, the relative position and posture transformation data of each structure to be deformed can be determined. These data provide an accurate basis for the subsequent determination of the coordinate positions of the wire bending points and clamping points. Moreover, this process can flexibly determine the clamp operation position data according to different deformation paths and the arrangement status of the structure to be deformed, and has strong adaptability. Accurate clamp operation position data can assist operators in improving the accuracy and stability of wire bending during the wire bending process, thereby helping to improve the quality and consistency of the product.

[0081] Furthermore, in one or more embodiments of the present specification, determining the standard force data corresponding to each operating position data of the bending pliers according to the adjacent operating position data and the historical deformation training data specifically includes the following process:

[0082] First, the movement trajectory of the bending forceps at adjacent operating positions is determined using data from adjacent operating positions and a preset initial direction. The force direction of the bending forceps is then determined based on this trajectory. Determining the movement trajectory using data from adjacent operating positions and a preset initial direction accurately depicts the path of the bending forceps during actual operation, ensuring the accuracy of the force direction. Historical deformation training data is then input into a pre-set deep learning network for training to obtain a force prediction model. The force direction and adjacent operating position data are then fed into the force prediction model to determine the force applied and the gripping posture of the bending forceps. The force direction, force applied, and gripping posture of the bending forceps are then summarized to determine the standard force data corresponding to each operating position of the bending forceps. By integrating this historical deformation training data with the deep learning network, the force required for the bending forceps in different operating positions can be predicted. This prediction, based on a large amount of historical data, offers high accuracy. Furthermore, through deep learning network training, the model learns the commonalities and differences between different bending processes, thereby generating a force application scheme that better suits the actual conditions of the mold being trained.

[0083] S104: Acquire multi-source data of the wire bending process based on multiple types of sensors pre-installed on the bending pliers, and determine current force data of the bending pliers based on the multi-source data; wherein the current force data includes: force direction, force magnitude, and holding posture.

[0084] In order to be able to determine the force data of the operator when operating the bending pliers in real time according to the posture of the bending pliers, thereby facilitating the guidance of the operator's operation and improving the bending quality. In the embodiment of this specification, multi-source data generated during the wire bending process will be obtained based on multiple types of sensors pre-set on the bending pliers, so as to determine the current force data of the bending pliers based on the multi-source data. Among them, it should be noted that the current force data of the bending pliers during the wire bending process includes the force direction, force magnitude, and holding posture. By real-time collection and analysis of the force data of the bending pliers during the wire bending process, valuable feedback and guidance information are provided to the operator, which helps to improve the operator's skills and improve the bending quality.

[0085] Specifically, in one or more embodiments of the present specification, multi-source data of a wire bending process is obtained based on multiple types of sensors pre-installed on the bending clamp, so as to determine the current force data of the bending clamp based on the multi-source data, which specifically includes the following process:

[0086] First, initial multi-source data is acquired from multiple sensors pre-installed on the bending pliers. This multi-source data includes pressure data, displacement data, and angle data. The force applied by the bending pliers is then determined based on the pressure data acquired by the sensors. The pressure concentration point is determined based on the pressure distribution corresponding to the pressure data. The gripping posture of the bending pliers is then determined based on the pressure concentration point and the angle data. The force direction of the bending pliers is also determined based on the displacement data acquired by the pre-installed sensors. The current force data of the bending pliers is then determined based on the force magnitude, gripping posture, and direction.

[0087] In addition, in another embodiment of the present specification, multi-source data of the wire bending process is obtained based on multiple types of sensors pre-installed on the bending clamp, so as to determine the current force data of the bending clamp based on the multi-source data. Specifically, this can be achieved through the following process:

[0088] First, initial multi-source data is collected from multiple sensors pre-installed on the bending clamp. This multi-source data includes pressure, displacement, and angle data. This initial multi-source data is then preprocessed to obtain processed multi-source data. Feature extraction is then performed on the processed multi-source data using a Fourier transform to obtain eigenvalues ​​of key features. Historical training data corresponding to the eigenvalues ​​of the key features is then obtained. Probabilities of the values ​​of each adjustable parameter in the force data are determined based on the historical training data. The values ​​of the adjustable parameters of the bending clamp are then determined based on these probabilities. Current force data of the bending clamp is then determined based on the values ​​of the adjustable parameters and the values ​​of the fixed parameters of the bending clamp. The multi-source data collected by the multiple sensors in this process, including pressure, displacement, and angle, comprehensively reflects the various physical conditions during the wire bending process. Preprocessing and feature extraction of this data allows for more accurate determination of the current force data of the bending clamp, thereby improving bending accuracy and consistency.

[0089] S105: Compare the current force data with the predicted force data to determine next force data for the bending operation according to the comparison result, and transmit the next force data to the current terminal to guide the completion of the wire bending operation.

[0090] To ensure accuracy and consistency in wire bending and meet quality requirements, the embodiments of this specification compare the current force data obtained in the above process with the predicted force data. Based on the comparison results, the next force data for the bending operation is determined and transmitted to the current terminal to guide the completion of the wire bending operation. This process allows the system to dynamically adjust based on real-time data feedback, enhancing the adaptability and flexibility of operators during wire bending training through closed-loop adjustments. Adjusting the next force data based on the comparison results can promptly detect and correct deviations in the bending process, achieving real-time optimization of the bending process, helping to reduce scrap rates and improve production efficiency and resource utilization.

[0091] Specifically, in one or more embodiments of this specification, Figure 2 The current force data is compared with the predicted force data to determine the next force data for the bending operation according to the comparison result, which specifically includes the following process:

[0092] Compare the current force data determined above with the predicted force data to determine the deviation between the current force data and the predicted force data. If the deviation is within the preset deviation range, then obtain the next operation position data adjacent to the operation position data corresponding to the predicted force data, and use the predicted force data corresponding to the next operation position data as the next force data for the bending operation. If the deviation exceeds the preset deviation range, determine the adjustment data for the current force data based on the deviation value to use the adjustment data as the next force data for the bending operation. For example, in a certain scenario, assuming that the force direction and the clamp holding posture of the current force data are consistent with the predicted force data, and the force magnitude in the predicted force data is 100 Newtons, the force magnitude of the current force data collected in real time is 110 Newtons. It can be determined that the deviation between the two is 10%, which exceeds the preset range of ±5%. At this point, an adjustment data needs to be calculated to bring the current force data closer to 100 Newtons. If the force is reduced by 10 Newtons, the next force data should be 100 Newtons. Since the deviation is known to be 10 Newtons, the force can be adjusted to 100 Newtons. The adjusted force of 100 Newtons, along with the force direction and gripping posture of the current force data, is used as the next force data. The operating parameters of the bending pliers are then adjusted accordingly to ensure that subsequent bending operations can be performed with the desired accuracy. If the deviation between the current force data and the predicted force data is 3%, which is within the preset range, the current deviation is controllable. The operating position data corresponding to the current predicted force data can be obtained, and the next adjacent operating position data can be found. The predicted force data corresponding to this next operating position data is then used as the next force data for the bending operation. By continuously adjusting the operating parameters of the bending pliers based on the comparison between real-time data and predicted data, precise bending control is achieved.

[0093] By comparing the current force data with the predicted force data in real time, deviations in the operation can be discovered and corrected in a timely manner. When the deviation value is within the preset range, the next operating position recommended by the prediction model and its corresponding predicted force data are directly adopted to ensure high precision and consistency of the bending process. When the deviation value is within an acceptable range, there is no need for complex calculations or adjustments. The direct use of predicted data speeds up the response speed and helps improve the efficiency of training. When the deviation value exceeds the preset range, the system can automatically calculate the adjustment data based on the deviation value and adjust the operating parameters of the bending clamp accordingly. This automated adjustment mechanism achieves continuous optimization of the bending process, and precise force control and timely deviation correction help provide timely auxiliary guidance to operators.

[0094] Furthermore, in one or more embodiments of this specification, after comparing the current force data with the predicted force data, the method further includes the following process:

[0095] Based on the comparison results of the current force data and the predicted force data, the number and magnitude of deviation parameters between the current force data and the predicted force data are determined, facilitating a more detailed evaluation of the comparison results. Based on the number and magnitude of the deviation parameters, the wire bending operation of the operation position data is then evaluated, and the evaluation results are stored in a storage unit corresponding to the bending forceps. The bending forceps have a corresponding tag associated with the ID of the current operator. The evaluation results are stored in the storage unit corresponding to the bending forceps and associated with the ID of the current operator, enabling personalized records for each operator. Based on these records, operation position data that exceeds a preset deviation range can be retrieved to develop a targeted training plan for the current operator. This helps improve operator skills and reduce operational errors. Furthermore, the evaluation results in the storage unit can be retrieved based on the current operator ID to determine the operation position data of the current operator that exceeds the preset deviation range. The operation position data that exceeds the preset deviation range is transmitted to a preset evaluation terminal, which then combines the operation position data with the preset deviation range to determine a training plan for the current operator. Combining operational data with personal training plans can motivate operators and enable managers to understand operational conditions in real time and take appropriate measures to intervene and guide operators in improving wire bending quality.

[0096] like Figure 3 As shown in FIG, an embodiment of this specification provides a schematic structural diagram of a wire bending prediction device based on clamp holding posture detection. Figure 3 It can be seen that in one or more embodiments of this specification, a wire bending prediction device based on clamp holding posture detection includes:

[0097] at least one processor; and,

[0098] a memory communicatively connected to the at least one processor; wherein,

[0099] The memory stores instructions that can be executed by the at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to: perform any of the above methods.

[0100] like Figure 4 As shown in FIG, an embodiment of this specification provides a structural diagram of a non-volatile storage medium. Figure 4 It can be seen that in one or more embodiments of this specification, a non-volatile storage medium stores computer-executable instructions 401, and the computer-executable instructions 401 can: execute any of the methods described above.

[0101] The various embodiments in this specification are described in a progressive manner. Similar portions between the various embodiments can be referenced to each other, and each embodiment focuses on the differences from the other embodiments. In particular, the device, apparatus, and non-volatile computer storage medium embodiments are generally similar to the method embodiments, so their descriptions are relatively simplified. For relevant details, refer to the descriptions of the method embodiments.

[0102] The foregoing description of this specification describes specific embodiments. Other embodiments are within the scope of the appended claims. In some cases, the actions or steps recited in the claims can be performed in an order different from that described in the embodiments and still achieve the desired results. Furthermore, the processes depicted in the accompanying drawings do not necessarily require the specific order shown or the sequential order to achieve the desired results. In certain embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0103] The foregoing description is merely one or more embodiments of this specification and is not intended to limit this specification. It will be apparent to those skilled in the art that various modifications and variations may be made to one or more embodiments of this specification. Any modifications, equivalent substitutions, or improvements made within the spirit and principles of one or more embodiments of this specification are intended to be within the scope of the claims of this specification.

Claims

1. A wire bending prediction method based on pliers holding posture detection, characterized in that: The method comprises: Acquire initial posture data of the mold to be trained in the current training scene, perform feature extraction on the initial posture data, and determine feature data of the mold to be trained; Performing deformation path planning on the mold to be trained based on the feature data and a preset standard structure model to determine the deformation path corresponding to the mold to be trained; Determining operating position data of the bending pliers based on the deformation path, and determining predicted force data corresponding to each operating position data of the bending pliers according to adjacent operating position data and historical training data; Acquiring multi-source data of a wire bending process based on multiple types of sensors pre-installed on the bending pliers, and determining current force data of the bending pliers based on the multi-source data; wherein the current force data includes: force direction, force magnitude, and gripping posture; The current force data is compared with the predicted force data to determine the next force data of the bending operation according to the comparison result, and the next force data is transmitted to the current terminal to guide the completion of the wire bending operation.

2. The wire bending prediction method based on clamp holding posture detection according to claim 1, characterized in that: Performing deformation path planning on the mold to be trained based on the feature data and the preset standard structure model to determine the deformation path corresponding to the mold to be trained specifically includes: Based on pre-deformation feature data corresponding to historical deformation data stored in a preset database, determining historical deformation data that matches the feature data, and determining a standard structural model corresponding to the feature data based on the historical deformation data; Determining a deformation target state corresponding to the mold to be trained based on difference data between the standard structural model and the mold to be trained; The arrangement state of the structure to be deformed corresponding to the mold to be trained is used as the initial solution of the path, and the annealing parameters are initialized; wherein the annealing parameters include: an initial temperature corresponding to the exploration range of the initial state of the mold to be trained, and a temperature attenuation factor corresponding to the path accuracy; Establishing a simulation model of the structure to be deformed of the mold to be trained based on the three-dimensional data of the mold to be trained, so as to adjust the simulation model of the structure to be deformed based on the initial temperature; Calculating difference data between the adjusted simulation model of the structure to be deformed and the deformation target state, adjusting the annealing parameters according to the difference data, and using the arrangement state of the structure to be deformed corresponding to the adjusted simulation model of the structure to be deformed as the current solution of the path; The current solution of the path is iteratively adjusted according to the adjusted annealing parameters to obtain a final solution of the path, and the final solution of the path is used as the deformation path corresponding to the mold to be trained.

3. The wire bending prediction method based on clamp holding posture detection according to claim 1, characterized in that: Determining clamp holding operation position data based on the deformation path specifically includes: Determining the deformation path and the arrangement state of the deformation structure corresponding to the mold to be trained, the deformation structure adjustment data corresponding to each deformation structure; wherein the deformation structure adjustment data includes: rotation adjustment, tilt adjustment, and movement adjustment; By comparing the adjustment data of adjacent structures to be deformed, the relative position and posture transformation data of each structure to be deformed are determined; determining the coordinate position of the wire bending point and the coordinate position of the clamping point based on the relative position and the posture transformation data; The coordinate positions of the wire bending points and the clamping points are summarized according to a preset traversal direction to determine the clamp operation position data; wherein the preset traversal direction includes: clockwise direction and counterclockwise direction.

4. The wire bending prediction method based on clamp holding posture detection according to claim 1, characterized in that: Determining the standard force data corresponding to each operating position data of the bending clamp according to the adjacent operating position data and the historical deformation training data, specifically including: Determining a movement trajectory of the bending pliers at the adjacent operation positions by using the adjacent operation position data and a preset initial direction, so as to determine a force application direction of the bending pliers based on the movement trajectory; Inputting the historical deformation training data into a preset deep learning network for training to obtain a force prediction model; Inputting the force direction and the adjacent operation position data into the force prediction model to obtain the force magnitude and holding posture of the bending pliers; The force direction, force magnitude and holding posture of the bending pliers are summarized to determine standard force data corresponding to each operation position data of the bending pliers.

5. The wire bending prediction method based on clamp holding posture detection according to claim 1, characterized in that: Acquiring multi-source data of a wire bending process according to multiple types of sensors pre-installed on the bending pliers, and determining current force data of the bending pliers according to the multi-source data, specifically including: Acquire initial multi-source data collected by multiple types of sensors pre-placed on the bending clamp; wherein the multi-source data includes: pressure data, displacement data and angle data; According to the pressure data, obtaining the force applied by the bending pliers; determining a pressure concentration point according to the pressure distribution corresponding to the pressure data, and obtaining a holding posture of the bending pliers according to the pressure concentration point and the angle data; The force direction of the bending pliers is acquired according to the displacement data, and the current force data of the bending pliers is determined based on the force magnitude, the holding posture and the force direction.

6. The wire bending prediction method based on pliers holding posture detection according to claim 1, characterized in that: Comparing the current force data with the predicted force data to determine the next force data for the bending operation according to the comparison result, specifically includes: comparing the current force data with the predicted force data to determine a deviation between the current force data and the predicted force data; If the deviation value is within the preset deviation range, acquiring the next operation position data adjacent to the operation position data corresponding to the predicted force data, and using the predicted force data corresponding to the next operation position data as the next force data for the bending operation; If the deviation value exceeds the preset deviation range, adjustment data of the current force data is determined according to the deviation value, so as to use the adjustment data as next force data for the bending operation.

7. The wire bending prediction method based on clamp holding posture detection according to claim 6, characterized in that: After comparing the current force data with the predicted force data, the method further includes: Determining the number of deviation parameters and the magnitude of the deviation values ​​between the current force data and the predicted force data according to a comparison result between the current force data and the predicted force data; Evaluate the wire bending operation of the operation position data based on the number of deviation parameters and the magnitude of the deviation value, and store the evaluation result in a storage unit corresponding to the bending pliers; wherein the bending pliers have a corresponding label, and the label is associated with the ID of the current operator; Recalling the evaluation result in the storage unit according to the ID of the current operator to determine the operation position data of the current operator that exceeds the preset deviation range; The operation position data exceeding the preset deviation range is transmitted to the preset evaluation terminal, so that the preset evaluation terminal determines a training plan for the current operator in combination with the operation position data exceeding the preset deviation range.

8. The wire bending prediction method based on clamp holding posture detection according to claim 1, characterized in that: Acquiring initial posture data of the mold to be trained, performing feature extraction on the initial posture data, and determining feature data of the mold to be trained, specifically includes: Acquire three-dimensional data of the mold to be trained, convert the three-dimensional data into two-dimensional slice data of each plane in a three-dimensional coordinate system, and calculate the gradient of the two-dimensional slice data in each direction according to a preset gradient operator; The gradients corresponding to the two-dimensional slice data of each plane are combined to obtain an edge intensity map; Extracting edges of the structure to be deformed in the mold to be trained based on the edge strength map, and connecting the edges of the structure to be deformed according to an edge connection algorithm to obtain edge contours of the structure to be deformed corresponding to the two-dimensional slice data of each plane; Performing three-dimensional reconstruction on the edge contours corresponding to the two-dimensional slice data of each plane to obtain three-dimensional edge contour information of the structure to be deformed, segmenting the region of the structure to be deformed of the mold to be trained based on the three-dimensional edge contour information, and determining initial posture data of the region of the structure to be deformed based on the segmentation result; Identify key feature points of the mold to be trained based on the initial posture data and the attribute characteristics of each key point; wherein the key feature points include: vertices of the structure to be deformed, centers of grooves, and boundaries between the structure to be deformed and the fixed structure; Based on the key feature points, the preset coordinate system and the reference point, the feature data of the mold to be trained is determined; wherein the feature data includes: pressing position coordinate data, tilt angle data of the structure to be deformed, and rotation angle data of the structure to be deformed.

9. A wire bending prediction device based on clamp holding posture detection, characterized in that: The device comprises: at least one processor; and, a memory communicatively connected to the at least one processor; wherein, The memory stores instructions that can be executed by the at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to: execute the method according to any one of claims 1 to 8.

10. A non-volatile storage medium storing computer-executable instructions, characterized in that: The computer executable instructions can execute the method according to any one of claims 1 to 8.

Images