Photovoltaic aluminum frame polishing control method and system for adjusting polishing head through mechanical arm

Abstract

The invention discloses a photovoltaic aluminum frame polishing control method and system with a polishing head adjusted through a mechanical arm, and relates to the technical field of photovoltaic aluminum frame polishing control. A three-degree-of-freedom mechanical arm is used for adjusting and controlling a pressing mechanism so that an abrasive belt polishing head can be attached to a photovoltaic aluminum frame; normal contact force and tangential friction force are monitored in real time through a six-dimensional force and torque sensor, and three-dimensional grinding characteristics are constructed in combination with an abrasive belt grinding motor load; based on the characteristics, a polishing state index is calculated through a polishing state evaluation model, and the deviation degree between the current polishing state and an ideal process reference is reflected; according to the grinding state index change curve and the trend thereof, whether grinding is completed or failed or not is automatically judged, and stable grinding quality is ensured; through fusion of force control fitting, multi-source sensing information processing and real-time state evaluation, efficient and accurate polishing of the photovoltaic aluminum frame is achieved, and the production efficiency and the product quality are improved.

Description

Technical Field

[0001] This invention belongs to the field of photovoltaic aluminum frame grinding control technology, specifically, it relates to a method and system for controlling the grinding of photovoltaic aluminum frames by adjusting the grinding head with a robotic arm. Background Technology

[0002] With rapid economic development and increasing global emphasis on clean energy, the photovoltaic industry has ushered in unprecedented development opportunities. As an important component of photovoltaic modules, the production and processing quality of photovoltaic aluminum frames directly affects the strength and service life of the modules.

[0003] In existing technologies, the grinding of photovoltaic aluminum frames largely relies on manual experience or semi-automated equipment for control. Common problems include a lack of real-time force control and status perception capabilities, over-reliance on preset parameters leading to an inability to dynamically adapt to changes in processing conditions, and poor grinding quality consistency. Specifically, traditional methods typically rely on motor current or a single force sensor for rough judgment, failing to simultaneously acquire multi-source information such as normal contact force, tangential friction force, and motor load, resulting in a lack of comprehensive judgment of the grinding status. Robotic arm control often uses position control rather than force control, making it difficult to ensure stable contact between the belt grinding head and the curved surface of the aluminum frame, easily causing uneven grinding pressure, leading to over-grinding, under-grinding, or surface damage. Furthermore, existing technologies generally lack real-time evaluation and dynamic decision-making mechanisms for the grinding process, failing to adjust or stop operations promptly based on the actual grinding status, affecting processing efficiency and posing quality risks.

[0004] To address the aforementioned problems, this invention proposes a method and system for controlling the grinding of photovoltaic aluminum frames using a robotic arm that adjusts the grinding head. Summary of the Invention

[0005] To address the shortcomings of existing technologies, this invention provides a method and system for controlling the grinding of photovoltaic aluminum frames using a robotic arm that adjusts the grinding head, thus solving the problems of difficulty in quantitative perception and complexity in intelligent control during the grinding process of photovoltaic aluminum frames.

[0006] The objective of this invention can be achieved through the following technical solutions: A method for controlling the grinding of photovoltaic aluminum frames using a robotic arm that adjusts the grinding head, the method comprising: Step 1: Obtain the photovoltaic aluminum frame to be polished, and use the pre-built three-degree-of-freedom robotic arm to control the clamping mechanism so that the sanding head fits the photovoltaic aluminum frame. Step 2: Real-time monitoring of the normal contact force and tangential friction force between the clamping mechanism and the sanding head using a six-dimensional force and torque sensor between the clamping mechanism and the sanding head during the sanding process; Real-time monitoring of the motor load associated with the belt sander motor during the sanding process, and construction of three-dimensional sanding features with normal contact force and tangential friction force; Step 3: Construct a polishing status assessment model, using the real-time determined three-dimensional polishing features as input, and output the polishing status index associated with the photovoltaic aluminum frame. Step 4: Obtain the polishing status index corresponding to the time associated with the photovoltaic aluminum frame according to the timeline, plot the polishing status index change curve based on the determined polishing status index, determine the change trend of the polishing status index change curve, and evaluate the polishing status of the photovoltaic aluminum frame.

[0007] As a further embodiment of the present invention, in step one, the three-degree-of-freedom robotic arm includes three degrees of freedom: forward and backward, up and down, and pitch. Each degree of freedom is driven by an independent AC servo motor; Each degree of freedom employs a transmission mechanism with a large transmission ratio and self-locking function.

[0008] As a further aspect of the present invention, in step one, the specific method of using a pre-constructed three-degree-of-freedom robotic arm to control the clamping mechanism to make the sanding head fit against the photovoltaic aluminum frame is as follows: Obtain any photovoltaic aluminum frame to be polished, and denote it as A; Using a 3D mapping tool with the three-degree-of-freedom robotic arm base coordinate system as a reference, the 3D morphology of the photovoltaic aluminum frame A entering the grinding area was scanned. Obtain the point cloud dataset associated with the photovoltaic aluminum frame A; All surfaces to be polished in the photovoltaic aluminum frame A are determined from the point cloud dataset, and the normal vector of each surface to be polished is obtained. Take the normal vector N of any surface to be polished, and with the current position of the belt polishing head as the starting point, generate a straight line L from the starting point to the direction of the normal vector N, and extract the normal contact force Fset preset by the operator; The three-degree-of-freedom robotic arm is advanced along a straight line L, and the normal contact force, denoted as F_n, is simultaneously acquired from the real-time feedback of the six-dimensional force and torque sensor. When the normal contact force Fn is first reached and maintained within the range of [Fset−ΔF, Fset+ΔF] for a continuous time T, the three-degree-of-freedom robotic arm is instructed to activate its self-locking function to complete the bonding of the sanding head to the photovoltaic aluminum frame. T is the time period preset by the operator, and ΔF is the normal contact force tolerance preset by the operator. Once the surface to be polished corresponding to the normal vector N is polished, the normal vectors associated with other surfaces to be polished are determined in the same way to complete the bonding process.

[0009] As a further aspect of the present invention, the specific method for constructing the three-dimensional polishing features in step two is as follows: A six-dimensional force and torque sensor is used to monitor the normal contact force and tangential friction force between the clamping mechanism and the belt grinding head during the grinding process, and the force is serialized in time sequence into normal contact force sequence Fn_1, Fn_2, ..., Fn_j and tangential friction force sequence Ft_1, Ft_2, ..., Ft_j, where j represents the total number of moments in the grinding process; The motor load of the belt grinding motor associated with the belt grinding head is monitored in real time during the grinding process and time-aligned with the normal contact force sequence to obtain the motor load sequence P1, P2, ..., Pj; Combine the normal contact force Fn_i, tangential friction force Ft_i, and motor load Pi corresponding to any time i, and denote it as the three-dimensional polishing feature RUB_i associated with the photovoltaic aluminum frame A at time i, where i is the counting index, 1≤i≤j.

[0010] As a further aspect of the present invention, the specific method for outputting the polishing status index associated with the photovoltaic aluminum frame in step three is as follows: Based on the determined 3D polishing features RUB_i, the polishing state evaluation model is used: Obtain the dimensionless grinding state index SI_i associated with the surface to be ground at time i; Wherein, Fn_ref, Ft_ref, and P_ref are the process reference values ​​preset by the operator for normal contact force, tangential friction force, and motor load, respectively; α and β are the weighting coefficients preset by the operator, and α+β= 1; k,m,h are exponential adjustment factors greater than zero. Similarly, determine the grinding state index corresponding to the three-dimensional grinding feature associated with each of the j time points, and record it as a grinding state index sequence in chronological order, denoted as: SI_1, SI_2, ..., SI_j.

[0011] As a further aspect of the present invention, the specific method for plotting the grinding state index change curve based on the determined grinding state index in step four is as follows: Obtain the polishing state index sequence SI_1, SI_2, ..., SI_j; Construct a two-dimensional coordinate system with the timeline as the horizontal axis and the values ​​of the grinding state index as the vertical axis. Mark all the grinding state indices in the grinding state index sequence SI_1, SI_2, ..., SI_j as data points in the constructed two-dimensional coordinate system in a time-aligned manner, obtaining j data points. Fit the j data points to a curve to obtain the grinding state index change curve S.

[0012] As a further aspect of the present invention, the specific method for evaluating the polishing condition of the photovoltaic aluminum frame in step four is as follows: S71. Obtain the grinding state index change curve S and its two-dimensional coordinate system; S72. Determine the scale mark with a value of 1 on the vertical axis, and construct a straight line L1 that passes through this scale mark, is parallel to the horizontal axis, and is perpendicular to the vertical axis. S73. Starting from time 0, monitor the slope between any two data points on the grinding state index change curve S. If the slope between any two data points is ≥0, no action is taken, and monitoring continues until any data point on the grinding state index change curve S coincides with the straight line L1. Grinding is stopped at the time of this data point, which is considered as grinding completion. S74. If the slope between any two data points is less than 0, obtain the grinding state index of the latter data point on the vertical axis of the two-dimensional coordinate system, mark it as the grinding state index to be detected, and compare it with the minimum threshold for qualified grinding preset by the operator. S75. If the polishing status index to be tested is greater than or equal to the minimum threshold for qualified polishing, then polishing shall be stopped and the polishing shall be considered complete. S76. If the polishing state index to be tested is less than the minimum threshold for qualified polishing, then continuously monitor the polishing state index associated with the subsequent X data points. If the polishing state index associated with any data point is greater than or equal to the minimum threshold for qualified polishing, then stop polishing and consider polishing to be completed. If the grinding status index associated with X data points is less than the minimum threshold for qualified grinding, grinding will be stopped, considered a grinding failure, and the operator will be alerted by an audible and visual alarm. Here, X is a value preset by the operator.

[0013] A photovoltaic aluminum frame grinding control system for adjusting the grinding head of a robotic arm, the system comprising: The force-controlled bonding module acquires the photovoltaic aluminum frame to be polished and uses a pre-constructed three-degree-of-freedom robotic arm to control the clamping mechanism so that the sanding head is bonded to the photovoltaic aluminum frame. The multi-source sensor information fusion module uses a six-dimensional force and torque sensor between the clamping mechanism and the sanding head to monitor the normal contact force and tangential friction force between the clamping mechanism and the sanding head in real time during the sanding process. Real-time monitoring of the motor load associated with the belt sander motor during the sanding process, and construction of three-dimensional sanding features with normal contact force and tangential friction force; The real-time evaluation module for the polishing status index constructs a polishing status evaluation model, taking the real-time determined three-dimensional polishing features as input and outputting the polishing status index associated with the photovoltaic aluminum frame. The polishing status determination module obtains the polishing status index corresponding to the time associated with the photovoltaic aluminum frame according to the timeline, plots the polishing status index change curve based on the determined polishing status index, determines the change trend of the polishing status index change curve, and evaluates the polishing status of the photovoltaic aluminum frame.

[0014] The beneficial effects of this invention are: This invention achieves precise control of the belt abrasive head and workpiece contact through a three-degree-of-freedom robotic arm, combined with real-time monitoring of normal contact force and tangential friction force by a six-dimensional force and torque sensor, and integrates motor load data to construct three-dimensional grinding features. This solves the problem of the photovoltaic aluminum frame to be ground and the belt abrasive head contact during the grinding process, and simultaneously enables adaptive control of the belt abrasive head to cope with photovoltaic aluminum frames of different shapes and postures, which greatly improves grinding quality and production efficiency. This invention constructs a three-dimensional feature vector that comprehensively reflects the physical state of the grinding process by precisely aligning and integrating the normal contact force, tangential friction force, and motor load in time. Its core advantage lies in the use of a state evaluation model that introduces a nonlinear adjustment factor to transform multidimensional physical quantities into intuitive and quantitative grinding state indices. This enables real-time, accurate, and objective digital evaluation of grinding quality, far exceeding traditional judgment methods that rely on a single parameter or human experience. This invention plots the real-time calculated grinding state index into a grinding state index change curve, and intelligently determines the grinding endpoint based on its dynamic relationship with the reference line and the preset qualified threshold. This achieves closed-loop quality decision-making and adaptive termination of the grinding process, accurately capturing the optimal stopping time when the grinding state index change curve reaches the ideal value to ensure optimal quality. At the same time, it intelligently identifies and handles abnormal working conditions where the grinding state index change curve declines. Through multiple judgment logics, it effectively distinguishes between acceptable normal fluctuations and real grinding failures, thereby greatly avoiding over-grinding or under-grinding, significantly improving production yield and efficiency, and finally achieving fully unmanned intelligent operation through audible and visual alarms. Attached Figure Description

[0015] The invention will now be further described with reference to the accompanying drawings.

[0016] Figure 1 This is a schematic diagram of the system described in this invention; Figure 2 This is a flowchart illustrating the method described in Embodiment 2 of the present invention; Figure 3 This is a flowchart illustrating the method described in Embodiment 4 of the present invention; Figure 4 This is a flowchart illustrating the method described in Embodiment 5 of the present invention. Detailed Implementation

[0017] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0018] Example 1 A photovoltaic aluminum frame grinding control system for adjusting the grinding head using a robotic arm, such as Figure 1 As shown, this system includes the following: This system is a photovoltaic aluminum frame grinding control system for adjusting the grinding head of a robotic arm. It mainly includes the following modules: force control bonding module, multi-source sensor information fusion module, real-time evaluation module for grinding status index, and grinding status determination module.

[0019] Among them, the force-controlled bonding module is a precursor module for subsequent modules, integrating three-dimensional machine vision perception and high-precision force-position hybrid control. Specifically, it first obtains real-time point cloud data of the aluminum frame to be polished through a 3D line laser scanner, then quickly reconstructs its three-dimensional shape through point cloud processing algorithms, and accurately calculates the spatial pose and normal vector of each surface to be polished. Subsequently, the surface to be polished, which fits the photovoltaic aluminum frame, is automatically generated based on the normal vector. The three-degree-of-freedom robotic arm then drives the belt polishing head to approach the workpiece in an adaptive force control mode according to this path. By reading the normal contact force fed back by the six-dimensional force sensor in real time, the belt polishing head is able to fit with the complex curved workpiece.

[0020] The three-degree-of-freedom robotic arm includes three degrees of freedom: forward and backward, up and down, and pitch, thereby automatically adjusting the spatial position and shape of the belt sander head to meet the sanding operation requirements in different positions. At the same time, each degree of freedom is driven by an independent AC servo motor to achieve high-precision linear or angular displacement control. Furthermore, each degree of freedom adopts a transmission mechanism with a large transmission ratio and self-locking function, which effectively reduces vibration and improves rigidity and stability.

[0021] In the multi-source sensor information fusion module, a six-dimensional force and torque sensor between the clamping mechanism and the sanding belt grinding head is used to monitor the normal contact force and tangential friction force between the clamping mechanism and the sanding belt grinding head during the grinding process. The motor load associated with the belt sander motor during the sanding process is monitored in real time using an electrical parameter detection instrument. The normal contact force, tangential friction force, and motor load are aligned on the time axis to construct a three-dimensional sanding feature.

[0022] The three-dimensional polishing feature represents the overall state of interaction between the belt polishing head and the photovoltaic aluminum frame at a specific moment.

[0023] The real-time evaluation module for the grinding state index includes the most important real-time evaluation module for the grinding state index. This module is a nonlinear state evaluation model that has been verified by a large number of process experiments (the process experiment verification is part of the existing technology, which will not be elaborated in detail in this solution, but can be understood as training and verifying it with actual data until the verification is passed).

[0024] The grinding status evaluation model takes the real-time determined three-dimensional grinding features as input, compares them with the preset process benchmark values, and analyzes the complex mapping relationship between each physical quantity and the grinding result through configurable weight coefficients and nonlinear exponential adjustment factors, and outputs a dimensionless grinding status index.

[0025] The closer the polishing state index is to 1, the closer it is to the process reference value; conversely, the further it is from the process reference value, the closer it is to the process reference value.

[0026] In the polishing status determination module, the polishing status index corresponding to the time associated with the photovoltaic aluminum frame is obtained according to the timeline. Based on the determined polishing status index, the polishing status index change curve is plotted, the change trend of the polishing status index change curve is determined, and the polishing status of the photovoltaic aluminum frame is evaluated. Specifically, the slope change trend of the polishing status index change curve is analyzed in real time. Based on this trend, combined with the comparison with the preset qualified threshold, multi-level decision-making is performed: when the ideal value is reached, the process is perfectly stopped; if the value drops briefly but is still above the threshold, the process is terminated early to improve efficiency; if the value continues to deteriorate and falls below the threshold, an alarm is triggered and the process is stopped, which is judged as a failure. This ensures the minimum quality threshold while maximizing production efficiency, and achieves the final closed loop of unmanned operation through audible and visual alarms.

[0027] This embodiment uses a force-controlled bonding module to accurately acquire the three-dimensional shape of the aluminum frame and drive the belt sander head to adaptively bond. A multi-source sensor information fusion module monitors the normal contact force, tangential friction force, and motor load during the sanding process in real time to construct three-dimensional sanding features. A real-time sanding state index evaluation module uses a nonlinear state evaluation model to compare the three-dimensional sanding features with the process reference value and output a sanding state index. A sanding state determination module compares the trend of the index change curve with a preset threshold to achieve multi-level decision-making, so as to achieve high-precision bonding sanding between the belt sander head and the photovoltaic aluminum frame, monitor the sanding quality in real time, automatically determine the sanding state and execute corresponding operations, and ultimately maximize production efficiency while ensuring sanding quality, realizing closed-loop control of unmanned operation.

[0028] Example 2 This embodiment further discloses the steps involved in the force-controlled bonding module based on Embodiment 1, such as... Figure 2 As shown, it specifically includes the following: First, obtain any photovoltaic aluminum frame to be polished and denote it as A. In general, the photovoltaic aluminum frame A to be polished is transported from the production line of photovoltaic aluminum frames.

[0029] When the photovoltaic aluminum frame A enters the polishing area (the area where the photovoltaic aluminum frame is polished), the three-dimensional mapping tool first performs a three-dimensional shape scan of the photovoltaic aluminum frame A entering the polishing area with the three-degree-of-freedom robotic arm base coordinate system as a reference, and obtains the point cloud dataset associated with the photovoltaic aluminum frame A. Among them, scanning based on a fixed base coordinate system: the three-degree-of-freedom robotic arm base coordinate system can effectively eliminate the error caused by the positional deviation of the photovoltaic aluminum frame A placed in the grinding area, ensuring the absolute alignment of the scanning data with the robotic arm's motion space.

[0030] Secondly, all the surfaces to be polished of the photovoltaic aluminum frame A are determined from the determined point cloud dataset (by using a point cloud processing algorithm, the point cloud sets belonging to each plane can be automatically identified), and the normal vector of each surface to be polished is obtained. The normal vector is a vector perpendicular to the surface to be polished.

[0031] Next, determine the normal vector of any surface and denote it as N. Generate a straight path along the direction of the calculated normal vector N and denote it as line L. Line L is the ideal trajectory for the belt sander head to approach the surface of the photovoltaic aluminum frame. Adjust the three-degree-of-freedom robotic arm to ensure that the belt sander head approaches the target plane vertically along line L in the correct posture. Approaching along the normal direction (straight line L direction) can effectively avoid unnecessary collisions or scratches between the sanding head and the photovoltaic aluminum frame, thereby avoiding unnecessary losses.

[0032] When the three-degree-of-freedom robotic arm advances along a straight line L, causing the sanding head to fit against the photovoltaic aluminum frame, the normal contact force fed back by the six-dimensional force and torque sensor is acquired in real time and recorded as F_n.

[0033] When the normal contact force Fn is first reached and remains within the range of [Fset−ΔF, Fset+ΔF] (ΔF is the operator's preset normal contact force tolerance) for a continuous period of T (T is the time period preset by the operator), the three-degree-of-freedom robotic arm is instructed to activate its self-locking function. At this point, it is considered that the judgment of the sanding head adhering to the photovoltaic aluminum frame for a continuous period of T is completed. This can avoid misjudgments caused by sensor noise, slight vibration, or brief contact.

[0034] The above describes the operation process for one surface of the photovoltaic aluminum frame A to be polished. If the polishing of this surface is completed, the normal vectors associated with other surfaces of the photovoltaic aluminum frame A to be polished will be determined, and the subsequent bonding and polishing operations will be completed.

[0035] In this embodiment, point cloud data of the aluminum frame is obtained through three-dimensional topographic scanning to ensure that the scanned data is absolutely aligned with the movement space of the robotic arm and to eliminate errors caused by positional deviations. Then, all surfaces to be polished and their normal vectors are identified from the point cloud data, and a straight path is generated along the direction of the normal vector to guide the polishing head to approach the target plane in the correct posture, avoiding collisions or scratches. When the polishing head is in contact with the aluminum frame, the normal contact force fed back by the six-dimensional force sensor is monitored in real time. After the force value remains stable within the preset range, the robotic arm self-locks to complete the contact. This process operates sequentially on each surface to be polished, aiming to achieve high-precision and stable polishing operations, improve polishing quality and efficiency, and ensure the accuracy and safety of operation through self-locking and force control mechanisms, reduce misjudgments and losses, and ultimately achieve efficient production of unmanned polishing.

[0036] Example 3 This embodiment further discloses a method for constructing three-dimensional polishing features based on a multi-source sensor information fusion module, based on embodiment 2, specifically including the following: First, as described in Example 1, the sanding head and the surface of the photovoltaic aluminum frame to be sanded were fitted together, and then the sanding operation was carried out. A six-dimensional force and torque sensor between the clamping mechanism and the abrasive belt grinding head is used to monitor the normal contact force and tangential friction force between the clamping mechanism and the abrasive belt grinding head in real time during the grinding process; The determined normal contact force and tangential friction force are then serialized according to time sequence, that is, the normal contact force and tangential friction force are sorted according to the order of time. After sorting, the normal contact force sequence and tangential friction force sequence are obtained, and are represented as: Fn_1, Fn_2, ..., Fn_j and Ft_1, Ft_2, ..., Ft_j, respectively. By constructing serialized data, extremely brief events during the grinding process can be effectively captured, such as encountering a particularly hard point, the sanding belt slipping momentarily, or a slight collision with a foreign object. These events are ignored in low-frequency sampling. In the above, j represents the total number of moments in the grinding process, and the total number of moments increases with the duration of the grinding process. That is, the normal contact force sequence Fn_1, Fn_2, ..., Fn_j and the tangential friction force sequence Ft_1, Ft_2, ..., Ft_j both increase with the duration of the grinding process. Fn_1 to Fn_j represent the normal contact force from the first moment to the jth moment, and Ft_1 to Ft_j represent the tangential friction force from the first moment to the jth moment.

[0037] Then, the motor load of the belt grinding motor associated with the belt grinding head is determined in real time during the grinding process, and recorded and sorted in the same time order. The motor load sequence is then obtained, which is aligned with the normal contact force sequence Fn_1, Fn_2, ..., Fn_j and the tangential friction force sequence Ft_1, Ft_2, ..., Ft_j, and is represented as: P1, P2, ..., Pj.

[0038] From the determined normal contact force sequence Fn_1, Fn_2, ..., Fn_j, tangential friction force sequence Ft_1, Ft_2, ..., Ft_j, and motor load sequence P1, P2, ..., Pj, determine the normal contact force Fn_i, tangential friction force Ft_i, and motor load Pi corresponding to any time i. Combine the extracted normal contact force Fn_i, tangential friction force Ft_i, and motor load Pi as the three-dimensional polishing feature RUB_i associated with the photovoltaic aluminum frame A, where i is the counting index, with a value range from 1 to j.

[0039] Example 4 This embodiment, based on embodiment 3, further discloses the steps involved in the real-time evaluation module of the grinding status index, such as... Figure 3 As shown, it specifically includes the following: The implementation process of this embodiment mainly relies on the grinding status evaluation model pre-constructed by the operator. The grinding status evaluation model is as follows: The construction principle of the grinding status evaluation model is as follows: Grinding is essentially a process of energy transfer and material removal. The grinding status assessment model normalizes multiple directly measurable physical quantities and constructs a dimensionless index that can indirectly and comprehensively reflect the grinding quality through nonlinear mapping and weighted fusion. Next, the three-dimensional polishing feature RUB_i determined in Example 3 is substituted into this polishing state evaluation model to obtain the polishing state index SI_i associated with the three-dimensional polishing feature RUB_i. Among them, Fn_ref, Ft_ref, and P_ref are the process reference values ​​for normal contact force, tangential friction force, and motor load, respectively, which are preset by the operator according to the material of photovoltaic aluminum frame A and the specifications of the sanding belt; α and β are both weighting coefficients preset by the operator, and α+β= 1, which are used to allocate the contribution of normal contact force and tangential friction force to the grinding state. The k, m, and h are all exponential adjustment factors greater than zero, used to characterize the nonlinear relationship between each physical quantity and the grinding state, and are determined by the operator in combination with the actual situation. The polishing state index SI_i is a dimensionless number. The closer it is to 1, the closer the polishing state is to the preset process reference benchmark, and the higher the polishing quality yield. The further it deviates from 1, the greater the risk of over-polishing or under-polishing in the current state, and the lower the yield. The principle of the polishing state evaluation model is to quantify the deviation between the current polishing state of the photovoltaic aluminum frame A and the ideal process reference state. The (Fn_i / Fn_ref), (Ft_i / Ft_ref), and (P_i / P_ref) operations in the grinding state evaluation model dedimensionalize all physical quantities and scale them to around 1, because a ratio of 1 indicates that the current value is in the ideal process baseline state.

[0040] For example, if the motor load increases sharply due to the sanding belt jamming (P_i / P_ref)^n, the sanding condition index will immediately spike when Fn and Ft are normal. The system can quickly detect this anomaly by monitoring the value of the sanding condition index in real time.

[0041] Next, repeat the above steps to process the three-dimensional polishing features associated with each of the j moments in this polishing process, and obtain j polishing state indices. Then, sort the j polishing state indices in chronological order to obtain the polishing state index sequence, denoted as: SI_1, SI_2, ..., SI_j.

[0042] Example 5 This embodiment, based on embodiment 4, further discloses the steps involved in the real-time evaluation module of the grinding state index, such as... Figure 4 As shown, it specifically includes the following: Based on the content described in Example 4, obtain the grinding state index sequence SI_1, SI_2, ..., SI_j associated with any one of the surfaces to be ground on the photovoltaic aluminum frame A throughout the entire grinding process.

[0043] Next, a two-dimensional coordinate system is constructed with the timeline as the horizontal axis and the values ​​of the polishing state index as the vertical axis. Then, the j polishing state indices in the polishing state index sequence SI_1, SI_2, ..., SI_j are plotted in the constructed two-dimensional coordinate system in the order of the timeline, and each polishing state index represents a data point in the two-dimensional coordinate system. In this way, the j data points associated with the j polishing state indices can be obtained.

[0044] Based on the determined j data points, a curve is fitted to the j data points to obtain the curve fitted from the j data points, which is denoted as the grinding state index change curve S.

[0045] By transforming the grinding state index sequence SI_1, SI_2, ..., SI_j into a trend curve, continuous trends can be effectively extracted from discrete data.

[0046] Next, obtain the grinding state index change curve S and its two-dimensional coordinate system, determine the scale with a value of 1 on the vertical axis (the grinding operation has reached the ideal state), and construct a straight line passing through the scale with a value of 1 on the vertical axis, parallel to the horizontal axis and perpendicular to the vertical axis, denoted as L1. The straight line L1 represents the "perfect target" of grinding quality.

[0047] Then, starting from time 0 on the horizontal axis and moving in the positive direction of the horizontal axis, continuously monitor the slope between any two data points on the grinding state index change curve S. If the slope between any two data points is greater than or equal to 0, no processing is done, and monitoring continues until any data point on the grinding state index change curve S coincides with the straight line L1, and grinding stops at the time associated with this data point, which is considered as grinding completion. A slope greater than or equal to 0 between any two data points indicates that the polishing status index is increasing, meaning that the polishing effect is continuously improving. A slope of 0 indicates that the effect has reached a plateau and stabilized at a good level. In this case, polishing should continue because the effect is still improving or being maintained. When any data point on the polishing state index change curve S coincides with the straight line L1, it means that the polishing quality at the current moment has reached the preset "perfect" standard. Stopping polishing immediately is the optimal choice, which ensures quality and avoids over-polishing.

[0048] If the slope between any two data points is less than 0, the grinding state index of the latter data point on the corresponding vertical axis of the two-dimensional coordinate system is obtained, marked as the grinding state index to be detected, and compared with the minimum threshold for qualified grinding preset by the operator. If the polishing condition index to be tested is greater than or equal to the minimum threshold for passing polishing, polishing is stopped and considered to be completed. This achieves an economical and efficient decision-making process, preventing further polishing that could lead to further deterioration of quality and failure due to the already poor condition, or wasting resources.

[0049] If the grinding status index to be tested is less than the minimum threshold for passing grinding, the grinding status index associated with the subsequent X data points is continuously monitored. If the grinding status index associated with any data point is greater than or equal to the minimum threshold for passing grinding, grinding is stopped and considered to be completed. This step handles the worst case and introduces a fault tolerance mechanism. The grinding status index to be tested being less than the minimum threshold for passing grinding indicates that the grinding quality has fallen below the pass line. By continuously monitoring the grinding status index associated with the subsequent X data points, short-term instantaneous interference can be effectively avoided. Furthermore, if any subsequent data point meets the standard, it means that the behavior of falling below the pass line was just an outlier. If the grinding status index associated with X data points is less than the minimum threshold for qualified grinding (confirming that this is a real and continuous process failure), grinding will be stopped and considered a grinding failure. An audible and visual alarm will be used to alert the operator because the automated process can no longer solve this problem. Here, X is a value preset by the operator.

[0050] All data in the formulas described above are numerical calculations performed with dimensions removed. Furthermore, any content not described in detail in this specification is existing technology known to those skilled in the art.

[0051] The above description is merely an example and illustration of the present invention. Those skilled in the art can make various modifications or additions to the specific embodiments described, or use similar methods to replace them, as long as they do not deviate from the invention or exceed the scope defined in the claims, all of which should fall within the protection scope of the present invention.

[0052] It should be stated that all user data collected in this application was collected with the user's consent and authorization. Furthermore, the uses of user data are legal and compliant, and the use and processing of user data comply with the relevant laws, regulations, and standards of the relevant regions.

Claims

1. A method for controlling the grinding of photovoltaic aluminum frames using a robotic arm that adjusts the grinding head, characterized in that, The method includes: Step 1: Obtain the photovoltaic aluminum frame to be polished, and use the pre-built three-degree-of-freedom robotic arm to control the clamping mechanism so that the sanding head fits the photovoltaic aluminum frame. Step 2: Real-time monitoring of the normal contact force and tangential friction force between the clamping mechanism and the sanding head using a six-dimensional force and torque sensor between the clamping mechanism and the sanding head during the sanding process; Real-time monitoring of the motor load associated with the belt sander motor during the sanding process, and construction of three-dimensional sanding features with normal contact force and tangential friction force; Step 3: Construct a polishing status assessment model, using the real-time determined three-dimensional polishing features as input, and output the polishing status index associated with the photovoltaic aluminum frame. Step 4: Obtain the polishing status index corresponding to the time associated with the photovoltaic aluminum frame according to the timeline, plot the polishing status index change curve based on the determined polishing status index, determine the change trend of the polishing status index change curve, and evaluate the polishing status of the photovoltaic aluminum frame.

2. The method according to claim 1, characterized in that, In step one, the three-degree-of-freedom robotic arm includes three degrees of freedom: forward and backward, up and down, and pitch. Each degree of freedom is driven by an independent AC servo motor; Each degree of freedom employs a transmission mechanism with a large transmission ratio and self-locking function.

3. The method according to claim 1, characterized in that, In step one, the specific method of using a pre-constructed three-degree-of-freedom robotic arm to control the clamping mechanism to make the sanding head fit against the photovoltaic aluminum frame is as follows: Obtain any photovoltaic aluminum frame to be polished, and denote it as A; Using a 3D mapping tool with the three-degree-of-freedom robotic arm base coordinate system as a reference, the 3D morphology of the photovoltaic aluminum frame A entering the grinding area was scanned. Obtain the point cloud dataset associated with the photovoltaic aluminum frame A; All surfaces to be polished in the photovoltaic aluminum frame A are determined from the point cloud dataset, and the normal vector of each surface to be polished is obtained. Take the normal vector N of any surface to be polished, and with the current position of the belt polishing head as the starting point, generate a straight line L from the starting point to the direction of the normal vector N, and extract the normal contact force Fset preset by the operator; The three-degree-of-freedom robotic arm is advanced along a straight line L, and the normal contact force, denoted as F_n, is simultaneously acquired from the real-time feedback of the six-dimensional force and torque sensor. When the normal contact force Fn is first reached and maintained within the range of [Fset−ΔF, Fset+ΔF] for a continuous time T, the three-degree-of-freedom robotic arm is instructed to activate its self-locking function to complete the bonding of the sanding head to the photovoltaic aluminum frame. T is the time period preset by the operator, and ΔF is the normal contact force tolerance preset by the operator. Once the surface to be polished corresponding to the normal vector N is polished, the normal vectors associated with other surfaces to be polished are determined in the same way to complete the bonding process.

4. The method according to claim 1, characterized in that, In step two, the specific method for constructing the three-dimensional polishing features is as follows: A six-dimensional force and torque sensor is used to monitor the normal contact force and tangential friction force between the clamping mechanism and the belt grinding head during the grinding process, and the force is serialized in time sequence into normal contact force sequence Fn_1, Fn_2, ..., Fn_j and tangential friction force sequence Ft_1, Ft_2, ..., Ft_j, where j represents the total number of moments in the grinding process; The motor load of the belt grinding motor associated with the belt grinding head is monitored in real time during the grinding process and time-aligned with the normal contact force sequence to obtain the motor load sequence P1, P2, ..., Pj; Combine the normal contact force Fn_i, tangential friction force Ft_i, and motor load Pi corresponding to any time i, and denote it as the three-dimensional polishing feature RUB_i associated with the photovoltaic aluminum frame A at time i, where i is the counting index, 1≤i≤j.

5. The method according to claim 4, characterized in that, In step three, the specific method for outputting the polishing status index associated with the photovoltaic aluminum frame is as follows: Based on the determined 3D polishing features RUB_i, the polishing state evaluation model is used: Obtain the dimensionless grinding state index SI_i associated with the surface to be ground at time i; Wherein, Fn_ref, Ft_ref, and P_ref are the process reference values ​​preset by the operator for normal contact force, tangential friction force, and motor load, respectively; α and β are the weighting coefficients preset by the operator, and α+β= 1; k,m,h are exponential adjustment factors greater than zero. Similarly, determine the grinding state index corresponding to the three-dimensional grinding feature associated with each of the j time points, and record it as a grinding state index sequence in chronological order, denoted as: SI_1, SI_2, ..., SI_j.

6. The method according to claim 5, characterized in that, In step four, the specific method for plotting the grinding state index change curve based on the determined grinding state index is as follows: Obtain the polishing state index sequence SI_1, SI_2, ..., SI_j; Construct a two-dimensional coordinate system with the timeline as the horizontal axis and the values ​​of the grinding state index as the vertical axis. Mark all the grinding state indices in the grinding state index sequence SI_1, SI_2, ..., SI_j as data points in the constructed two-dimensional coordinate system in a time-aligned manner, obtaining j data points. Fit the j data points to a curve to obtain the grinding state index change curve S.

7. The method according to claim 6, characterized in that, In step four, the specific method for evaluating the polishing condition of the photovoltaic aluminum frame is as follows: S71. Obtain the grinding state index change curve S and its two-dimensional coordinate system; S72. Determine the scale mark with a value of 1 on the vertical axis, and construct a straight line L1 that passes through this scale mark, is parallel to the horizontal axis, and is perpendicular to the vertical axis. S73. Starting from time 0, monitor the slope between any two data points on the grinding state index change curve S. If the slope between any two data points is ≥0, no action is taken, and monitoring continues until any data point on the grinding state index change curve S coincides with the straight line L1. Grinding is stopped at the time of this data point, which is considered as grinding completion. S74. If the slope between any two data points is less than 0, obtain the grinding state index of the latter data point on the vertical axis of the two-dimensional coordinate system, mark it as the grinding state index to be detected, and compare it with the minimum threshold for qualified grinding preset by the operator. S75. If the polishing status index to be tested is greater than or equal to the minimum threshold for qualified polishing, then polishing shall be stopped and the polishing shall be considered complete. S76. If the polishing state index to be tested is less than the minimum threshold for qualified polishing, then continuously monitor the polishing state index associated with the subsequent X data points. If the polishing state index associated with any data point is greater than or equal to the minimum threshold for qualified polishing, then stop polishing and consider polishing to be completed. If the grinding status index associated with X data points is less than the minimum threshold for qualified grinding, grinding will be stopped, considered a grinding failure, and the operator will be alerted by an audible and visual alarm. Here, X is a value preset by the operator.

8. A photovoltaic aluminum frame grinding control system for adjusting the grinding head using a robotic arm, characterized in that, The system includes: The force-controlled bonding module acquires the photovoltaic aluminum frame to be polished and uses a pre-constructed three-degree-of-freedom robotic arm to control the clamping mechanism so that the sanding head is bonded to the photovoltaic aluminum frame. The multi-source sensor information fusion module uses a six-dimensional force and torque sensor between the clamping mechanism and the sanding head to monitor the normal contact force and tangential friction force between the clamping mechanism and the sanding head in real time during the sanding process. Real-time monitoring of the motor load associated with the belt sander motor during the sanding process, and construction of three-dimensional sanding features with normal contact force and tangential friction force; The real-time evaluation module for the polishing status index constructs a polishing status evaluation model, taking the real-time determined three-dimensional polishing features as input and outputting the polishing status index associated with the photovoltaic aluminum frame. The polishing status determination module obtains the polishing status index corresponding to the time associated with the photovoltaic aluminum frame according to the timeline, plots the polishing status index change curve based on the determined polishing status index, determines the change trend of the polishing status index change curve, and evaluates the polishing status of the photovoltaic aluminum frame.

Images