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

Abstract

The embodiment of the invention discloses a metal wire bending prediction method and device based on clamp holding posture detection and a medium, relates to the technical field of digital intelligent monitoring, and is used for solving the problem that the clamp holding training effect is poor due to the fact that the next metal wire bending operation is difficult to predict in time in an existing mode. The method comprises the steps of obtaining initial attitude data of a to-be-trained mold in a current training scene, and performing feature extraction; according to the extracted feature data and a preset standard structure model, planning a deformation path of the to-be-trained mold; and determining operation position data of the bending clamp based on the path, and predicting predicted force application data of each operation position in combination with adjacent operation position data and historical training data. The method comprises the following steps: collecting multi-source data in a metal wire bending process by utilizing multiple types of sensors on a bending clamp so as to determine current force application data of the bending clamp; and comparing the current force application data with the predicted force application data, and determining next force application data of the bending operation according to a comparison result so as to guide the completion of the metal wire bending operation according to the next force application data.

Description

Technical Field

[0001] The present specification relates to the field of data processing technology, and in particular to a method, device and medium for predicting metal wire bending based on clamp 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 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, integrated circuits, etc., 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 clamp holding posture for wire bending. 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, the traditional monitoring of wire bending operations is to detect the finished product after the processing is completed. Since the detection feedback is provided only after the wire bending operation is completed, it is difficult to correct the operator's mistakes in the operation training process in time, 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 instructor 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 time to assist the operator in mastering the clamp holding posture and force application 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 the present 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 the present specification provide a method for predicting wire bending based on clamp 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] According to the feature data and the preset standard structure model, a deformation path planning is performed on the mold to be trained to determine the deformation path corresponding to the mold to be trained;

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

[0010] According to the multi-type sensors pre-installed on the bending pliers, multi-source data of the wire bending process are obtained to determine the current force data of the bending pliers according to the multi-source data; wherein the current force data includes: force direction, force magnitude, and holding 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 according to the feature data and the preset standard structure model to determine the deformation path corresponding to the mold to be trained specifically includes:

[0013] Based on the pre-deformation feature data corresponding to the historical deformation data stored in the preset database, determine the historical deformation data matching the feature data, and determine the standard structural model corresponding to the feature data according to the historical deformation data;

[0014] Determining a deformation target state corresponding to the mold to be trained according to difference data between the standard structure 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] Based on the three-dimensional data of the mold to be trained, a simulation model of the structure to be deformed of the mold to be trained is established, 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 taking 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 the present specification, based on the deformation path, determining the clamp holding operation position data specifically includes:

[0020] According to the deformation path and the arrangement state of the structure to be deformed corresponding to the mold to be trained, the adjustment data of the structure to be deformed corresponding to each structure to be deformed is determined; wherein the adjustment data of the structure to be deformed 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 bending points of the wires and the clamping points are summarized according to the preset traversal direction to determine the clamp holding 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 operation position data of the bending pliers according to the adjacent operation position data and the historical deformation training data specifically includes:

[0025] Determining a moving track of the bending pliers at the adjacent operating position through the adjacent operating position data and the preset initial direction, so as to determine a force application direction of the bending pliers based on the moving track;

[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 forceps;

[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, multi-source data of the wire bending process is obtained according to multi-type sensors pre-installed on the bending pliers, so as to determine the current force data of the bending pliers according to the multi-source data, specifically including:

[0030] Acquire initial multi-source data collected by multiple types of sensors pre-placed on the bending pliers; 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] Determine a pressure concentration point according to the pressure distribution corresponding to the pressure data, so as to obtain a holding posture of the bending forceps 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 of the bending operation according to the comparison result specifically includes:

[0035] Comparing the current force application data with the predicted force application data, and determining a deviation value between the current force application data and the predicted force application 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, the adjustment data of the current force data is determined according to the deviation value, so as to use the adjustment data as the next force data of the bending operation.

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

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

[0040] According to the number of the deviation parameters and the size of the deviation value, the wire bending operation of the operation position data is evaluated, and the evaluation result is stored in a storage unit corresponding to the bending pliers; wherein the bending pliers has a corresponding label, and the label is associated with the id of the current operator;

[0041] Calling 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 the training plan of 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 the 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] Based on the edge strength map, extract the edge of the structure to be deformed in the mold to be trained, and connect the edge of the structure to be deformed according to the edge connection algorithm to obtain the edge contour 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, and 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] According to the initial posture data and the attribute characteristics of each key point, the key feature points of the mold to be trained are identified; wherein the key feature points include: the vertices of the structure to be deformed, the center of the groove, and the boundary line 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 are determined; wherein the feature data include: pressing position coordinate data, inclination angle data of the structure to be deformed, and rotation angle data of the structure to be deformed.

[0050] One or more embodiments of the present specification provide a wire bending prediction device based on clamp 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: execute any of the above-mentioned methods.

[0054] One or more embodiments of the present specification provide a non-volatile computer storage medium storing computer executable instructions, wherein the computer executable instructions are configured to be able 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 obtaining the initial posture data of the mold to be trained, the key information of the initial posture data can be extracted as feature data, and then the deformation path planning is performed based on the feature data and the standard structural model, which can ensure that the deformation process of the mold to be trained can be carried out within the expected requirements. Determining the operating position of the bending pliers based on the deformation path and then obtaining the predicted force data corresponding to each operating position will help plan the force strategy in advance. Using the multi-type sensors on the bending pliers to obtain multi-source data in real time, the force application during the bending process can be accurately grasped. Then, by comparing the current force data with the predicted force data, the deviation in the operation can be discovered in time, and the next force data can be adjusted according to the comparison results. This helps to achieve dynamic adjustment and optimization of the operator's bending operation, ensure that the bending process always follows the correct path, and realize the training of the operator while ensuring the product operation quality. Transmitting the next force data to the current terminal can guide the operator to make adjustments and targeted training in real time. This real-time feedback mechanism helps the operator to correct errors in time and improve operating skills and bending quality. BRIEF DESCRIPTION OF THE DRAWINGS

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

[0058] Figure 1A schematic flow chart of a method for predicting wire bending based on detection of holding forceps posture 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 metal wire bending prediction device based on clamp 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] In order to enable those skilled in the art to better understand the technical solutions in this specification, the technical solutions in the embodiments of this specification will be clearly and completely described below in conjunction with the drawings in the embodiments of this specification. Obviously, the described embodiments are only part of the embodiments of this specification, not all of the embodiments. Based on the embodiments of this specification, all other embodiments obtained by ordinary technicians in this field 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 comprises Figure 1 It can be seen that in one or more embodiments of the present specification, a method for predicting wire bending 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] In order to facilitate the training of metal bending for operators in different demand scenarios, the initial posture data of the mold to be trained in the current training scene will be obtained in the embodiments of this specification. For example, if the current training scene is an electrician training scene, electronic devices such as circuit boards or integrated circuits are obtained as the mold to be trained, and if the current training scene is dental correction in a medical scene, a tooth model is obtained as the mold to be trained. After obtaining the initial posture data of the mold to be trained, in order to facilitate the determination of the standardized deformation path of metal wire bending, the embodiments of this specification first need to extract features from the initial posture data and determine the feature data of the mold to be trained, so that path planning can be performed based on the feature data later.

[0067] Specifically, in one or more embodiments of the present specification, initial posture data of the mold to be trained is obtained to extract features from the initial posture data to determine feature data of the mold to be trained, which specifically includes the following process:

[0068] For a complex mold to be trained, it may include a structure that needs to be deformed by a metal wire and other fixed structures. Therefore, in order to analyze the structure that needs to be deformed, the three-dimensional data of the mold to be trained will be obtained in the embodiment of this specification to convert the three-dimensional data into two-dimensional slice data of each plane in the three-dimensional coordinate system, and then the gradient of the two-dimensional slice data in each direction is calculated according to the preset gradient operator. By obtaining the three-dimensional data of the mold to be trained and converting it into two-dimensional slice data of each plane, the process can capture the geometric shape and details of the mold with high precision. Then, in order to more easily identify the edge of the structure to be deformed in the mold, the gradients corresponding to the two-dimensional slice data of each plane are combined to obtain an edge strength map. For example, in a certain scene, for the two-dimensional slice data of each plane, a preset gradient operator such as a Sobel operator, a Prewitt operator, etc. is used to calculate the gradient of the two-dimensional slice data in the x direction and the y direction. The gradient operator returns a matrix of the same size as the original slice data, in which each element represents the gradient value of the corresponding position in the x or y direction. Then for each two-dimensional slice, the gradient values ​​in the x direction and the y direction are combined using the Pythagorean theorem, that is, the gradient amplitude is calculated. 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 strength map, the edge of the structure to be deformed in the mold to be trained is extracted according to the edge strength map, and the edge of the structure to be deformed is connected according to the edge connection algorithm to obtain the edge contour of the structure to be deformed corresponding to the two-dimensional slice data of each plane. Then, the edge contour corresponding to the two-dimensional slice data of each plane is reconstructed in three dimensions to obtain the three-dimensional edge contour information of the structure to be deformed, and the region of the structure to be deformed of the mold to be trained is segmented according to 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 result. It should be noted that the edge contour data on each slice will be extracted during the three-dimensional reconstruction process, and then the corresponding contour lines on adjacent slices are connected by small triangles to form a curved surface to obtain the three-dimensional edge contour information. Then, the key feature points of the mold to be trained are identified according to the initial posture data and the attribute characteristics of each key point; it should be noted that the key feature points include: the vertex of the structure to be deformed, the center of the groove, and the boundary between the structure to be deformed and the fixed structure. According to 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 according to the three-dimensional edge contour information, and the key deformation area in the mold can be accurately identified. 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 for the mold to be trained according to the feature data and a preset standard structure model, and determining a 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 according to the acquired characteristic data and its standard structural model, so as to determine 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 the subsequent bending of the metal wire so that the bending of the metal wire meets the quality requirements.

[0072] Specifically, in one or more embodiments of the present specification, a deformation path planning is performed on the mold to be trained according to the feature data and the preset standard structure model, and a deformation path corresponding to the mold to be trained is determined, 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, and there is a preset database that stores the historical data of the deformation of the mold to be trained in the past, including the feature data before deformation, such as mold size, material type, etc. and the 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, it is possible to determine the target state that the mold to be trained should reach after deformation, that is, the required part shape.

[0074] Then, 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: the initial temperature corresponding to the exploration range of the initial state of the mold to be trained, and the temperature attenuation factor corresponding to the path accuracy. Then, according to the three-dimensional data of the mold to be trained, a simulation model of the structure to be deformed of the mold to be trained is established, so that the simulation model of the structure to be deformed is adjusted according to the initial temperature. The difference data between the adjusted simulation model of the structure to be deformed and the deformation target state is calculated to adjust the annealing parameters according to the difference data, and the arrangement state of the structure to be deformed corresponding to the adjusted simulation model of the structure to be deformed is used as the current solution of the path. The current solution of the path is iteratively adjusted according to the adjusted annealing parameters to obtain the 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.

[0075] In the above process, by matching with the historical deformation data, the standard structure model corresponding to the current feature data can be accurately found. Then, the difference data between the standard structure model and the mold to be trained can be used to accurately determine the deformation target state of the mold to be trained, thereby improving the accuracy of deformation prediction. In addition, 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 accuracy and speed of the search. Therefore, by establishing a simulation model of the structure to be deformed and adjusting it according to the initial temperature, the search space can be quickly narrowed and the solution efficiency can be improved. Then, the iterative adjustment process can continuously approach the optimal solution, and finally obtain the deformation path of the mold to be trained, ensuring the controllability and stability of the deformation process. By continuously adjusting the difference data between the simulation model and the deformation target state, it can be ensured that the deformation path finally obtained is highly consistent with the expected target. This not only improves production efficiency, but also reduces dependence 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 the adjacent operation position data and the historical training data.

[0077] In order to ensure that the bending pliers can move along a predetermined path during the bending process of the operator, the cost of training materials can be reduced while improving the operator's mastery of skills, reducing unnecessary waste of force and possible damage. In the embodiment of this specification, the operating position of the bending pliers will be determined according to the deformation path, and the predicted force data corresponding to each operating position data of the bending pliers will be determined according to 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. Among them, it can be understood that during the bending process of the wire, 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 the present specification, based on the deformation path, determining the clamp holding operation position data specifically includes:

[0079] First, according to the deformation path and the arrangement state of the structure to be deformed corresponding to the mold to be trained, the adjustment data of the structure to be deformed corresponding to each structure to be deformed is determined. Among them, the adjustment data of the structure to be deformed includes: rotation adjustment, tilt adjustment, and movement adjustment. Then, by comparing the adjustment data of the adjacent structures to be deformed, the relative position and posture transformation data of each structure to be deformed are determined. Based on the relative position and posture transformation data, the coordinate position of the wire bending point and the coordinate position of the clamping point are determined. Then, according to the preset traversal direction, the coordinate position of each wire bending point and the coordinate position of each clamping point are summarized to determine the clamp holding operation position data. Among them, the preset traversal direction includes: clockwise direction and counterclockwise direction. For example, taking the dental correction scene as an example, the mold to be trained can be compared to the tooth arrangement state, and the deformation path represents the movement trajectory of the tooth from the current arrangement state to the ideal arrangement state. Then, if the front tooth area needs to be adjusted at this time, one of the front teeth is called the structure to be deformed A, which needs to be rotated inward, i.e., rotation adjustment, slightly tilted downward, i.e., tilt adjustment, and moved forward, i.e., movement adjustment, to achieve the ideal arrangement position. According to the deformation path of the tooth, i.e. the direction and distance of the tooth movement and the current tooth arrangement state, the adjustment data required for the structure A to be deformed, i.e. the specific data of the rotation adjustment, tilt adjustment and movement adjustment, can be calculated at this time. Then, the adjacent teeth, such as the teeth next to the structure A to be deformed, will be compared, which is called the adjustment data of the structure B to be deformed. By comparing the rotation, tilt and movement adjustment data of A and B, the relative position and posture transformation data between them can be determined. For example, if tooth A rotates inward by 5 degrees and tooth B remains in place, then A has a 5-degree rotation difference relative to B. Based on the relative position and posture transformation data, the operator can determine the points on the wire that need to be bent and the clamping points for fixing the wire. These points will ensure that the wire can apply the correct force to the teeth according to the predetermined deformation path, thereby achieving the movement of the teeth. For example, if tooth A needs to move forward, the bending point of the wire at the position of tooth A may need to be bent forward to provide a forward pulling force. At the same time, the clamping point will be installed on tooth A to ensure that the wire can be firmly fixed on the tooth. Finally, according to the preset traversal direction, such as clockwise or counterclockwise, the coordinate positions of all wire bending points and clamping points are summarized to determine the clamp operation position data. These data will correct the operator's bending operation in real time.

[0080] In this process, not only the rotation adjustment and tilt adjustment of the structure to be deformed are taken into account, but also the movement adjustment, 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, and these data provide an accurate basis for the subsequent determination of the coordinate positions of the wire bending points and clamping points. And this process can flexibly determine the clamp operation position data according to different deformation paths and arrangement states 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 product quality and consistency.

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

[0082] First, the moving trajectory of the bending pliers at the adjacent operating positions is determined by the adjacent operating position data and the preset initial direction, so as to determine the force direction of the bending pliers according to the moving trajectory. By determining the moving trajectory by the adjacent operating position data and the preset initial direction, the path of the bending pliers in the actual operation can be accurately depicted, thereby ensuring the accuracy of the force direction. Then, the historical deformation training data is input into the preset deep learning network for training to obtain a force prediction model. The force direction and the adjacent operating position data are input into the force prediction model to obtain the force magnitude and holding posture of the bending pliers. Then, the force direction, force magnitude and holding posture of the bending pliers are summarized to determine the standard force data corresponding to each operating position data of the bending pliers. By introducing the historical deformation training data and combining the deep learning network, the force magnitude required for the bending pliers at different operating positions can be predicted. This prediction is based on a large amount of historical data and therefore has a high accuracy. And based on the training of the deep learning network, the model can learn the commonalities and differences in different bending processes, thereby generating a force scheme that is more in line with the actual situation of the current mold to be trained.

[0083] S104: Acquire multi-source data of the wire bending process according to the multi-type sensors pre-installed on the bending pliers, so as to determine the current force data of the bending pliers according to 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, so as to facilitate the guidance of the operator's operation and improve the bending quality. In the embodiment of this specification, multi-source data generated in the process of bending the wire will be obtained according to the multi-type 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 bending of the wire includes the direction of force, the magnitude of force, and the 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 skills of the operator and improve the bending quality.

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

[0086] First, obtain the initial multi-source data collected by the multi-type sensors pre-installed on the bending pliers; wherein the multi-source data includes: pressure data, displacement data and angle data. Then, according to the pressure data obtained by the sensor, obtain the force applied by the bending pliers. Determine the pressure concentration point according to the pressure distribution corresponding to the pressure data, and then obtain the holding posture of the bending pliers according to the pressure concentration point and the angle data. At the same time, determine the force direction of the bending pliers according to the displacement data collected by the preset sensor, and then determine the current force data of the bending pliers according to the force, holding posture and force direction.

[0087] In addition, in another embodiment of the present specification, multi-source data of the wire bending process is obtained according to multi-type sensors pre-installed on the bending pliers, so as to determine the current force data of the bending pliers according to the multi-source data, which can be specifically implemented by the following process:

[0088] First, obtain the initial multi-source data collected by the multi-type sensors pre-placed on the bending pliers, wherein the multi-source data includes: pressure data, displacement data and angle data. Then pre-process the initial multi-source data to obtain the processed multi-source data, and extract the features of the processed multi-source data according to Fourier transform to obtain the characteristic values ​​of the key features. Then obtain the historical training data corresponding to the characteristic values ​​of the key features, so as to determine the probability values ​​of the values ​​of the adjustable parameters in the force data according to the historical training data, so as to determine the values ​​of the adjustable parameters of the bending pliers based on the probability values, and determine the current force data of the bending pliers according to the values ​​of the adjustable parameters of the bending pliers and the values ​​of the fixed parameters of the bending pliers. In this process, the multi-source data pressure, displacement, angle, etc. collected by the multi-type sensors can fully reflect the various physical states in the process of bending the metal wire. After pre-processing and feature extraction of these data, the current force data of the bending pliers can be determined more accurately, thereby improving the accuracy and consistency of bending.

[0089] S105: Compare the current force data with the predicted force data to determine the next force data of 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 the accuracy and consistency of wire bending and meet the quality requirements for wire bending, the current force data obtained in the above process is compared with the predicted force data in the embodiment of this specification, so as to determine the next force data of 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. This process allows the system to make dynamic adjustments based on real-time data feedback, and enhances the adaptability and flexibility of operators in wire bending training operations through closed-loop adjustment. And by adjusting the next force data according to the comparison results, deviations in the bending process can be discovered and corrected in a timely manner, and the bending process can be optimized in real time, which helps 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 of 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 of 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, assume that the force direction of the current force data is consistent with the clamp holding posture and the predicted force data, and the force magnitude in the predicted force data is 100 Newtons, and the force magnitude of the current force data collected in real time is 110 Newtons. Then it can be determined that the deviation between the two is 10%, which exceeds the preset range of ±5%. At this time, an adjustment data needs to be calculated to adjust the current force data to a level close to 100 Newtons. If it is decided to reduce the force by 10 Newtons, the next force data should be 100 Newtons. Since the known deviation value is 10 Newtons, the force size can be adjusted to 100 Newtons, and the adjusted 100 Newton force size and the force direction and clamp holding posture of the current force data are used as the next force data, and the operating parameters of the bending pliers are adjusted accordingly to ensure that the subsequent bending operation can be performed with the expected accuracy. If the deviation value between the current force data and the predicted force data is 3%, which is within the preset range, it means that the current deviation is controllable, and the operation position data corresponding to the current predicted force data can be obtained, and the adjacent next operation position data can be found. Then, the predicted force data corresponding to the next operation position data is used as the next force data for the bending operation. By continuously adjusting the operating parameters of the bending pliers according to the comparison results of the real-time data and the predicted data, accurate bending control can be 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 accuracy 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 pliers 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] Further, in one or more embodiments of the present specification, after comparing the current force application data with the predicted force application data, the method further includes the following process:

[0095] According to the comparison result of the current force data and the predicted force data, the number of deviation parameters and the size of the deviation value between the current force data and the predicted force data are determined, which is helpful for a more detailed evaluation of the comparison result of the current force data and the predicted force data. Then, according to the number of deviation parameters and the size of the deviation value, the wire bending operation of the operation position data is evaluated, and the evaluation result is stored 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. The evaluation result is stored in the storage unit corresponding to the bending pliers and associated with the ID of the current operator, so that personalized records for each operator can be realized. According to these records, the operation position data exceeding the preset deviation range can be called to formulate a targeted training plan for the current operator. This helps to improve the skill level of the operator and reduce operational errors. In addition, the evaluation result in the storage unit can be called according to the ID of the current operator to determine the operation position data of the current operator exceeding 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 the training plan of the current operator in combination with the operation position data exceeding the preset deviation range. Combining operation data with personal training plans can motivate operators and enable managers to understand the operation status in real time and take appropriate measures to intervene and guide operators to improve the quality of wire bending.

[0096] like Figure 3 As shown, in the embodiment of this specification, a schematic diagram of the structure of a wire bending prediction device based on clamp holding posture detection is provided. Figure 3 It can be seen that in one or more embodiments of the present 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: execute any of the above-mentioned methods.

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

[0101] Each embodiment in this specification is described in a progressive manner, and the same or similar parts between the embodiments can be referred to each other, and each embodiment focuses on the differences from other embodiments. In particular, for the device, equipment, and non-volatile computer storage medium embodiments, since they are basically similar to the method embodiments, the description is relatively simple, and the relevant parts can be referred to the partial description of the method embodiment.

[0102] The above is a description of a specific embodiment of the specification. Other embodiments are within the scope of the appended claims. In some cases, the actions or steps recorded in the claims can be performed in an order different from that in the embodiments and still achieve the desired results. In addition, the processes depicted in the drawings do not necessarily require the specific order or continuous order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0103] The above description is only one or more embodiments of this specification and is not intended to limit this specification. For those skilled in the art, one or more embodiments of this specification may have various changes and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of one or more embodiments of this specification shall be included in the scope of the claims of this specification.

Claims

1. A method for predicting wire bending based on clamp 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; According to the feature data and the preset standard structure model, a deformation path planning is performed on the mold to be trained to determine the deformation path corresponding to the mold to be trained; Based on the deformation path, determining the operation position data of the bending pliers, and determining the predicted force data corresponding to each operation position data of the bending pliers according to the adjacent operation position data and the historical training data; According to the multi-type sensors pre-installed on the bending pliers, multi-source data of the wire bending process are obtained to determine the current force data of the bending pliers according to the multi-source data; wherein the current force data includes: force direction, force magnitude, and holding 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 method for predicting wire bending based on clamp holding posture detection according to claim 1, characterized in that: According to the feature data and the preset standard structure model, a deformation path planning is performed on the mold to be trained to determine the deformation path corresponding to the mold to be trained, specifically including: Based on the pre-deformation feature data corresponding to the historical deformation data stored in the preset database, determine the historical deformation data matching the feature data, and determine the standard structural model corresponding to the feature data according to the historical deformation data; Determining a deformation target state corresponding to the mold to be trained according to difference data between the standard structure 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; Based on the three-dimensional data of the mold to be trained, a simulation model of the structure to be deformed of the mold to be trained is established, 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 taking 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 method for predicting wire bending based on clamp holding posture detection according to claim 1, characterized in that: Based on the deformation path, determining the clamp holding operation position data specifically includes: According to the deformation path and the arrangement state of the structure to be deformed corresponding to the mold to be trained, the adjustment data of the structure to be deformed corresponding to each structure to be deformed is determined; wherein the adjustment data of the structure to be deformed 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 the preset traversal direction to determine the clamp holding operation position data; wherein the preset traversal direction includes: clockwise direction and counterclockwise direction.

4. The method for predicting wire bending based on clamp holding posture detection according to claim 1, characterized in that: Determine the standard force data corresponding to each operation position data of the bending pliers according to the adjacent operation position data and the historical deformation training data, specifically including: Determining a moving track of the bending pliers at the adjacent operating position through the adjacent operating position data and the preset initial direction, so as to determine a force application direction of the bending pliers based on the moving track; 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 forceps; 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 method for predicting wire bending based on clamp holding posture detection according to claim 1, characterized in that: According to the multi-type sensors pre-installed on the bending pliers, multi-source data of the wire bending process are obtained to determine the 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 pliers; 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; Determine a pressure concentration point according to the pressure distribution corresponding to the pressure data, so as to obtain a holding posture of the bending forceps 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 method for predicting wire bending based on clamp 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 of the bending operation according to the comparison result, specifically includes: Comparing the current force application data with the predicted force application data, and determining a deviation value between the current force application data and the predicted force application 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, the adjustment data of the current force data is determined according to the deviation value, so as to use the adjustment data as the next force data of the bending operation.

7. The method for predicting wire bending based on clamp holding posture detection as claimed in claim 6, characterized in that: After comparing the current force application data with the predicted force application data, the method further includes: Determining the number of deviation parameters and the magnitude of the deviation value between the current force application data and the predicted force application data according to a comparison result between the current force application data and the predicted force application data; According to the number of the deviation parameters and the size of the deviation value, the wire bending operation of the operation position data is evaluated, and the evaluation result is stored in a storage unit corresponding to the bending pliers; wherein the bending pliers has a corresponding label, and the label is associated with the id of the current operator; Calling 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 the training plan of the current operator in combination with the operation position data exceeding the preset deviation range.

8. The method for predicting wire bending based on clamp holding posture detection according to claim 1, characterized in that: Acquiring initial posture data of the mold to be trained, extracting features from 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 the 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; Based on the edge strength map, extract the edge of the structure to be deformed in the mold to be trained, and connect the edge of the structure to be deformed according to the edge connection algorithm to obtain the edge contour 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, and implementing segmentation of the structure to be deformed region of the mold to be trained based on the three-dimensional edge contour information, and determining initial posture data of the structure to be deformed region based on the segmentation result; According to the initial posture data and the attribute characteristics of each key point, the key feature points of the mold to be trained are identified; wherein the key feature points include: the vertices of the structure to be deformed, the center of the groove, and the boundary line 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 are determined; wherein the feature data include: pressing position coordinate data, inclination 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 so that the at least one processor can: execute any of the methods described in claims 1-8.

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

Images