Preston equation-based weld joint material removal amount prediction and polishing method

Abstract

The invention relates to a weld joint material removal amount predicting and polishing method based on a Preston equation, and relates to the field of mine equipment machining. The method comprises the steps of obtaining machining parameter data in a historical weld joint polishing process; constructing a mathematical relationship model between the material removal amount and the processing parameter data based on a Preston equation; the machining parameter data are trained through a machine learning algorithm, the mathematical relation model is optimized, and a weld material removal amount prediction model is determined; determining a predicted weld material removal amount based on the weld material removal amount prediction model; according to the predicted weld material removal amount, initial machining parameters are determined; and weld joint polishing machining is conducted based on the initial machining parameters, and the machining parameters are regulated and controlled in real time. Under the condition, it can be ensured that the expected material removal effect can be achieved in each grinding operation, so that the welding seam machining precision and consistency are remarkably improved, meanwhile, material waste is reduced, the machining time is shortened, and the production efficiency is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of mining equipment processing, and particularly to a method for predicting the material removal amount of weld materials and grinding based on the Preston equation. Background Art

[0002] In the fields of mining equipment manufacturing and maintenance, weld processing plays a crucial role. It not only relates to the structural integrity and safety of the equipment but also directly determines the service life of the equipment. Welding, as one of the main means of connecting metal components, plays an irreplaceable role in ensuring that mining machinery can withstand extreme working conditions. However, the welds formed after welding often need to be ground to meet the surface quality and dimensional accuracy requirements of the design.

[0003] However, most traditional weld grinding methods rely on manual experience. Although this method has met certain requirements in the past, with the continuous improvement of the industry's requirements for product quality, its limitations have gradually emerged. It is inevitable that the material removal is uneven during the manual grinding process, which not only affects the appearance quality of the final product but more importantly may weaken the mechanical properties of the weld area, thereby shortening the overall life of the equipment. Another significant problem existing in the current weld grinding processing technology is the inability to accurately predict the material removal during the grinding process and adjust the grinding parameters in real time, resulting in a greater impact on the processing quality and efficiency. Due to the lack of an effective feedback mechanism, it is difficult for operators to make timely adjustments according to the actual situation, making variables such as pressure and speed during the grinding process unable to always maintain the optimal state. In addition, workpieces with different materials and thicknesses have different requirements for the grinding process, and it is difficult for traditional methods to take these differences into account, resulting in waste of resources and increased costs. Summary of the Invention

[0004] The purpose of the present invention is to provide a method for predicting the material removal amount of weld materials and grinding based on the Preston equation to solve the problems existing in the above-mentioned prior art.

[0005] To achieve the above purpose, the technical solution adopted by the present invention is as follows:

[0006] In the first aspect, the present invention provides a method for predicting the material removal amount of weld materials and grinding based on the Preston equation. The method uses a grinding robot for processing, and the method includes:

[0007] S1. Obtain the processing parameter data during the historical weld grinding process. The processing parameter data at least includes the grinding force, the inclination angle, and the grinding wheel linear velocity;

[0008] S2. Construct a mathematical relationship model between the material removal amount and the processing parameter data based on the Preston equation;

[0009] S3. Train the machining parameter data through a machine learning algorithm, optimize the mathematical relationship model, and determine a prediction model for the weld material removal amount;

[0010] S4. Based on the prediction model for the weld material removal amount, determine the predicted weld material removal amount;

[0011] S5. Determine the initial machining parameters according to the predicted weld material removal amount;

[0012] S6. Perform weld grinding based on the initial machining parameters and adjust the machining parameters in real time.

[0013] In a possible implementation, in step S2, the Preston equation includes:

[0014] G = K·P·v·cos(θ);

[0015] where G is the weld material removal amount, K is a proportionality constant, P is the grinding force, v is the linear speed of the grinding wheel, and θ is the inclination angle; the proportionality constant K is obtained by performing a regression analysis on historical grinding data and reflects the removal rate characteristics between a specific grinding wheel material and the material to be ground; the grinding force P is dynamically adjusted and adjusted to an optimal value according to the real-time feedback of the grinding effect; the linear speed v of the grinding wheel includes a preset multi-level speed setting to adapt to weld materials of different hardnesses and thicknesses; the measurement of the inclination angle θ is obtained in real time through an inclination angle sensor to ensure precise control and adjustment of the inclination angle θ.

[0016] In a possible implementation, in steps S1 to S2, the machining parameter data further includes a grinding wheel wear compensation factor and an environmental compensation factor, and the Preston equation includes:

[0017] G = K·P·v·cos(θ)·C wear ·C env ;

[0018] where C wear is the grinding wheel wear compensation factor, indicating the influence of grinding wheel wear on the weld material removal amount; C env is the environmental compensation factor, indicating the influence of environmental factors on the weld material removal amount;

[0019] C wear = 1 - α·W(t);

[0020] where α is the wear sensitivity coefficient; W(t) is the grinding wheel wear degree, indicating the wear amount of the grinding wheel within the usage time t;

[0021] C env = 1 + βT ·(T - T 0 ) + β H ·(H - H 0 );

[0022] Wherein, β T is the temperature sensitivity coefficient; β H is the humidity sensitivity coefficient; T is the ambient temperature monitored in real time; H is the ambient humidity monitored in real time; T 0 is the temperature under reference environmental conditions; H 0 is the humidity under reference environmental conditions.

[0023] In a possible implementation, in step S3:

[0024] The machine learning algorithm is a neural network algorithm. A neural network model is constructed using a deep learning framework. The neural network model includes an input layer, multiple hidden layers, and an output layer. The hidden layer uses an activation function, and the neural network model is trained using historical grinding data, and the network weights are optimized through the backpropagation algorithm. After the neural network model is trained, a weld material removal amount prediction model is obtained;

[0025] The relationship between the input layer, hidden layer, and output layer of the neural network model is:

[0026] G = f(P, v, θ, C wear , C env );

[0027] Wherein, the input layer is the grinding force P, the grinding wheel linear velocity v, the inclination angle θ, the grinding wheel wear compensation factor C wear and the environmental compensation factor C env ; the hidden layer is to extract complex relationships using a non-linear activation function; the output layer is the weld material removal amount G.

[0028] In a possible implementation, in step S6:

[0029] In response to the weld grinding process, the weld material removal amount is monitored in real time, and an error analysis is performed with the weld material removal amount predicted by the weld material removal amount prediction model. According to the error analysis result, the processing parameters are adjusted in real time;

[0030] The error analysis includes:

[0031] ΔG = G predicted - G actual ;

[0032] Wherein, G predicted and G actual are respectively the weld material removal amount predicted by the model and the weld material removal amount monitored in real time;

[0033] The regulation processing parameters include:

[0034] P new = P current + k p ·ΔG;

[0035] V new = V current + k v ·ΔG;

[0036] Wherein, k p and k v are adjustment coefficients, which are respectively used to adjust the grinding force and the linear speed of the grinding disc; P current is the current grinding force; V current is the current linear speed of the grinding disc.

[0037] In a possible implementation manner, in the step S6:

[0038] In response to the welding seam grinding process, the position of the welding seam is captured in real time by a visual welding seam tracker, and the grinding path is automatically adjusted to ensure full coverage of the welding seam;

[0039] The automatic adjustment of the grinding path includes:

[0040] ΔX = X predicted - X actual ;

[0041] Wherein, X predicted is the path generated by the model; X actual is the path monitored in real time;

[0042] X new = X current + ΔX;

[0043] Wherein, X current is the current path position; X new is the newly generated path position; the path optimization is realized by an interpolation or curve fitting algorithm, so that the end of the grinding robot fits the shape of the welding seam.

[0044] In a second aspect, the present invention provides a welding seam material removal amount prediction and grinding device based on the Preston equation. The device is applicable to the welding seam material removal amount prediction and grinding method based on the Preston equation as described above. The device includes:

[0045] An acquisition module, configured to acquire processing parameter data in the historical welding seam grinding process. The processing parameter data at least includes the grinding force, the inclination angle, and the linear speed of the grinding disc;

[0046] A building module for constructing a mathematical relationship model between the material removal amount and the processing parameter data based on the Preston equation;

[0047] A determination module for training the processing parameter data through a machine learning algorithm, optimizing the mathematical relationship model, and determining a weld material removal amount prediction model;

[0048] The determination module is further configured to determine the predicted weld material removal amount based on the weld material removal amount prediction model;

[0049] The determination module is further configured to determine the initial processing parameters according to the predicted weld material removal amount;

[0050] A regulation module for performing weld grinding processing based on the initial processing parameters and regulating the processing parameters in real time.

[0051] In a third aspect, the present invention provides a computer device, which includes a processor and a memory. At least one instruction, at least one program, a code set, or an instruction set is stored in the memory. The processor can load and execute at least one instruction, at least one program, a code set, or an instruction set to implement the weld material removal amount prediction and grinding method based on the Preston equation provided above.

[0052] In a fourth aspect, the present invention provides a computer-readable storage medium, in which at least one instruction, at least one program, a code set, or an instruction set is stored. The processor can load and execute at least one instruction, at least one program, a code set, or an instruction set to implement the weld material removal amount prediction and grinding method based on the Preston equation provided above.

[0053] In a fifth aspect, the present invention provides a computer program product or a computer program, which includes computer program instructions stored in a computer-readable storage medium. The processor reads the computer instructions from the computer-readable storage medium and executes the computer instructions, so that the computer device executes the weld material removal amount prediction and grinding method based on the Preston equation provided above.

[0054] The beneficial effects brought by the technical solution provided by the present invention at least include:

[0055] By obtaining the processing parameter data during the historical weld grinding process, the processing parameter data at least includes the grinding force, the inclination angle, and the grinding wheel linear velocity; constructing a mathematical relationship model between the material removal amount and the processing parameter data based on the Preston equation; training the processing parameter data through a machine learning algorithm, optimizing the mathematical relationship model, and determining a weld material removal amount prediction model; determining the predicted weld material removal amount based on the weld material removal amount prediction model; determining the initial processing parameters according to the predicted weld material removal amount; performing weld grinding processing based on the initial processing parameters, and regulating the processing parameters in real time. In this case, it is possible to ensure that each grinding operation can achieve the expected material removal effect, thereby significantly improving the accuracy and consistency of weld processing, while reducing material waste and processing time, and improving production efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0056] The accompanying drawings are used to provide a further understanding of the present invention and constitute a part of the specification. They are used together with the embodiments of the present invention to explain the present invention, but do not constitute a limitation to the present invention.

[0057] Figure 1 The flowchart of a method for predicting the weld material removal amount and grinding based on the Preston equation provided by an exemplary embodiment of the present invention is shown.

[0058] Figure 2 The structural block diagram of a device for predicting the weld material removal amount and grinding based on the Preston equation provided by an exemplary embodiment of the present invention is shown.

[0059] Figure 3 The structural schematic diagram of a computer device for performing a method for predicting the weld material removal amount and grinding based on the Preston equation provided by an exemplary embodiment of the present invention is shown. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0060] The following will clearly and completely describe the technical solutions in the embodiments of the present invention with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all of the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0061] The present invention will be further described in detail below with reference to the accompanying drawings and embodiments.

[0062] Figure 1The flowchart of a method for predicting the removal amount of weld material and grinding based on the Preston equation provided by an exemplary embodiment of the present invention is shown. The method for predicting the removal amount of weld material and grinding based on the Preston equation uses a grinding robot for processing, and the method includes the following steps:

[0063] Step S1: Obtain the processing parameter data during the historical weld grinding process. The processing parameter data at least includes the grinding force, the inclination angle, and the grinding wheel linear velocity.

[0064] Specifically, the grinding force refers to the pressure exerted by the grinding tool on the weld surface, which directly affects the material removal efficiency and quality; the inclination angle refers to the angle of the grinding tool relative to the weld surface. Different inclination angles will cause changes in the contact area, thereby affecting the grinding effect; the grinding wheel linear velocity represents the speed at which the edge point of the grinding tool moves relative to the workpiece surface, which also has a crucial impact on the material removal rate. In practical applications, in order to ensure that the collected data can truly reflect the grinding process, sensor technology is usually used to monitor various parameters in real time and store the data in the welding database for subsequent analysis. In this way, not only can the accuracy of the prediction model be improved, but also a good data foundation can be laid for realizing intelligent grinding.

[0065] Step S2: Construct a mathematical relationship model between the material removal amount and the processing parameter data based on the Preston equation.

[0066] Furthermore, the Preston equation includes: G = K·P·v·cos(θ); where G is the weld material removal amount, K is a proportionality constant, P is the grinding force, v is the grinding wheel linear velocity, and θ is the inclination angle; the proportionality constant K is obtained through regression analysis of historical grinding data and reflects the removal rate characteristics between a specific grinding wheel material and the material being ground; the grinding force P is dynamically adjusted to the optimal value according to the real-time feedback of the grinding effect; the grinding wheel linear velocity v includes a preset multi-level velocity setting to adapt to weld materials of different hardnesses and thicknesses; the inclination angle θ is measured in real time by an inclination angle sensor to ensure precise control and adjustment of the inclination angle θ. Specifically, this model aims to establish a quantitative relationship between the weld material removal amount G and processing parameters such as the grinding force P, the grinding wheel linear velocity v, and the inclination angle θ. According to the Preston equation, the material removal rate is proportional to the normal pressure (i.e., the grinding force) and the relative velocity (i.e., the grinding wheel linear velocity), and is affected by a proportionality constant K, which comprehensively reflects the removal rate characteristics between a specific grinding wheel material and the material being ground. The determination of the proportionality constant K is obtained through regression analysis of historical grinding data, which means it not only represents the properties of the material itself but also reflects the result of the interaction between the grinding tool and the workpiece. To ensure optimal performance during grinding, the grinding force P can be dynamically adjusted according to real-time feedback to adapt to changing grinding conditions and ensure an ideal grinding effect. The grinding wheel linear velocity v is designed as a multi-level velocity setting to flexibly meet the requirements of weld materials of different hardnesses and thicknesses. In addition, the inclination angle θ is monitored in real time by an inclination angle sensor installed on the grinding robot, ensuring precise control of the grinding angle, which is crucial for maintaining a stable material removal rate.

[0067] Furthermore, the processing parameter data also includes a grinding wheel wear compensation factor and an environmental compensation factor. The Preston equation includes: G = K·P·v·cos(θ)·C wear ·C env ; where C wear is the grinding wheel wear compensation factor, which represents the impact of grinding wheel wear on the weld material removal amount. It is a dynamic value that can be adjusted in real time according to the usage time, wear degree, etc. of the grinding wheel; C env is the environmental compensation factor, which represents the impact of environmental factors (such as temperature, humidity, etc.) on the weld material removal amount; C wear = 1 - α·W(t); where α is the wear sensitivity coefficient determined by experiments; W(t) is the grinding wheel wear degree, which represents the wear amount of the grinding wheel within the usage time t and can be measured by a real-time sensor; C env = 1 + β T ·(T - T 0 ) + β H ·(H - H 0 ); where β Tis the temperature sensitivity coefficient, which is obtained by experimental calibration; β H is the humidity sensitivity coefficient, which is obtained by experimental calibration; T is the ambient temperature monitored in real time; H is the ambient humidity monitored in real time; T 0 is the temperature under reference environmental conditions; H 0 is the humidity under reference environmental conditions.

[0068] Specifically, considering that the wear of the grinding wheel and the changes in environmental factors such as temperature and humidity during the actual grinding process will affect the material removal amount, a grinding wheel wear compensation factor C wear and an environmental compensation factor C env are introduced when constructing the mathematical model. These two factors quantify the influence of the grinding wheel wear degree and environmental conditions on the weld material removal amount respectively. The grinding wheel wear compensation factor C wear is a dynamic value, which is updated in real time according to the usage time and wear condition of the grinding wheel; while the environmental compensation factor C env reflects the specific influence of the external environment on the grinding process, making the prediction model closer to the actual situation. To sum up, by integrating these complex variables, a more accurate and practical material removal amount prediction model can be constructed to guide the efficient operation of the grinding robot.

[0069] Step S3: Train the processing parameter data through a machine learning algorithm, optimize the mathematical relationship model, and determine the weld material removal amount prediction model.

[0070] Furthermore, the machine learning algorithm is a neural network algorithm. A neural network model is constructed using a deep learning framework. The neural network model includes an input layer, multiple hidden layers, and an output layer. The hidden layer uses an activation function, and the neural network model is trained using historical grinding data. The network weights are optimized through the backpropagation algorithm. After the neural network model is trained, a weld material removal amount prediction model is obtained; the relationship between the input layer, hidden layer, and output layer of the neural network model is: G = f(P, v, θ, C wear , C env ); where the input layer is the grinding force P, the grinding wheel linear velocity v, the inclination angle θ, the grinding wheel wear compensation factor C wear and the environmental compensation factor C env ; the hidden layer is to extract complex relationships using a non-linear activation function; the output layer is the weld material removal amount G.

[0071] Specifically, to improve the prediction accuracy and achieve intelligent grinding, a neural network algorithm is used as the core tool to construct and optimize the prediction model of the weld material removal amount. Using deep learning frameworks such as TensorFlow or PyTorch, a neural network architecture consisting of an input layer, multiple hidden layers, and an output layer can be designed. Among them, the input layer receives processing parameters such as grinding force P, grinding wheel linear velocity v, and inclination angle θ as feature vectors; while the output layer is responsible for giving the expected material removal amount G. Activation functions (such as ReLU, Sigmoid, or Tanh) are used in the hidden layer to introduce non-linear factors, enabling the model to capture more complex patterns. After the model is constructed, the next crucial step is to train the neural network, which depends on a large number of historical grinding data sets that contain records of actual grinding processes and their corresponding material removal amounts under different conditions. During the training process, the data is first divided into a training set and a validation set to evaluate the model performance and prevent overfitting; then, the predicted values are calculated through forward propagation and compared with the true values to calculate the loss function (such as mean squared error MSE); next, the backpropagation algorithm is applied to calculate the gradient according to the chain rule and adjust the weights and bias terms in the network, aiming to minimize the loss function value, so that the model can better fit the data. This process is iterated repeatedly until the convergence condition is met, such as reaching the predetermined maximum number of iterations or the change in loss between consecutive iterations is less than a certain threshold. Once the training is completed and the model performs stably and reliably, the final prediction model of the weld material removal amount is obtained. This model can not only be used to predict the material removal amount under new working conditions, but also provide intelligent decision-making support for the grinding robot, helping it to adjust the grinding parameters in real time to ensure high-quality grinding effects. In addition, with the accumulation of more new data, the model can be retrained periodically to make it continuously evolve to adapt to new process requirements and technological progress. In this way, not only the automation level of the grinding operation is improved, but also the quality control ability of the welded structural parts is significantly enhanced.

[0072] Step S4: Based on the prediction model of the weld material removal amount, determine the predicted weld material removal amount.

[0073] In the embodiments of the present application, once the above neural network model is trained and verified to have good generalization ability, the model can be used to predict the weld material removal amount in a new grinding task. Specifically, in practical applications, an operator or an automated system will input the specific parameters of the weld to be processed currently, such as processing conditions like grinding force P, grinding wheel linear speed v, inclination angle θ, etc. These parameters will be passed as input features to the trained neural network model, and through a series of complex non-linear transformations and calculations, the output layer will finally give the expected material removal amount G. This predicted value not only reflects the possible material removal effect under given conditions but also provides a scientific basis for subsequent adjustment of the grinding process. In addition, to ensure the accuracy of the prediction results, the model can be corrected online by combining real-time monitoring data to adapt to changes in actual working conditions.

[0074] Step S5: Determine the initial processing parameters according to the predicted weld material removal amount.

[0075] In the embodiments of the present application, after obtaining the predicted material removal amount, the next step is to set the most suitable initial processing parameters based on this to ensure that the grinding process is both efficient and does not damage the quality of the weld. This step involves multiple considerations: First, the specific situation of the weld needs to be considered, such as factors like material type, hardness, and surface condition; second, the grinding target should be evaluated, such as the expected roughness and flatness standards; finally, the state of the grinding tool also needs to be comprehensively considered, including the newness of the grinding wheel and its grinding wheel wear compensation factor C wear 's influence. Based on the above factors, the best parameter combination can be searched by looking up a pre-established empirical database or using an optimization algorithm (such as a genetic algorithm). For example, if the prediction indicates that a large material removal amount is required, the grinding force P can be appropriately increased or the grinding wheel linear speed v can be increased; on the contrary, for a more delicate grinding operation, a smaller grinding force and a lower speed should be selected, and the inclination angle θ should be precisely controlled to avoid excessive material removal. In addition, considering that environmental factors (such as temperature, humidity) may also affect the grinding effect, an environmental compensation factor C env should be introduced for fine-tuning. In short, by carefully setting the initial processing parameters, a solid foundation can be laid for achieving high-quality grinding, and it also helps to improve production efficiency and reduce costs.

[0076] Step S6: Perform weld grinding processing based on the initial processing parameters and regulate the processing parameters in real time.

[0077] Furthermore, in response to the real-time monitoring of the weld material removal amount during the weld grinding process, an error analysis is carried out with the weld material removal amount predicted by the weld material removal amount prediction model, and the processing parameters are regulated in real time according to the error analysis results; the error analysis includes: ΔG = G predicted -Gactual ; among them, G predicted and G actual are respectively the weld material removal amounts predicted by the model and monitored in real time; the regulated processing parameters include: P new =P current +k p ·ΔG; V new =V current +k v ·ΔG; where k p and k v are adjustment coefficients, which are respectively used to adjust the grinding force and the grinding wheel linear speed; P current is the current grinding force; V current is the current grinding wheel linear speed.

[0078] Specifically, after determining the initial processing parameters, the grinding robot starts the actual grinding operation on the weld according to these set values. At the same time, in order to ensure that the grinding effect meets the expectations, the system will continuously monitor the key indicators during the grinding process, especially the weld material removal amount. In this process, the data collected by sensors (such as force sensors, displacement sensors, etc.) will be used to calculate the actually occurring material removal amount and compare it with the predicted value given by the weld material removal amount prediction model. Through error analysis, the difference between the two can be identified. If it is found that the actual removal amount deviates from the predicted value, it means that the current processing parameters need to be adjusted to correct this deviation. For example, if the actual removal amount is less than the predicted value, it may be because the grinding force is insufficient or the grinding wheel linear speed is too low. At this time, the grinding force P can be increased or the grinding wheel linear speed v can be increased for compensation; conversely, if the removal amount is too large, these parameters should be appropriately reduced to avoid over-grinding.

[0079] Furthermore, in response to the weld grinding process, the weld position is captured in real time through a vision weld tracker, and the grinding path is automatically adjusted to ensure full coverage of the weld; the automatic adjustment of the grinding path includes: path error calculation, ΔX = X predicted -X actual ; where X predicted is the path generated by the model; X actual is the path monitored in real time; path correction, X new =X current +ΔX; where X current is the current path position; X newis the newly generated path position; path optimization is achieved through interpolation or curve fitting algorithms to make the end of the grinding robot fit the shape of the weld seam. The vision weld tracker is an advanced automated tool that can capture the position information of the weld seam in real time through cameras or other optical sensing devices, including but not limited to the shape, width, and height changes of the weld seam. When detecting an offset in the weld seam position, it can quickly respond and automatically adjust the movement trajectory of the grinding robot to ensure that the grinding head always moves along the correct path, thereby achieving full coverage of the weld seam. In addition, the vision tracking technology can also help identify defects or abnormalities in the weld seam, such as discontinuous weld seams, uneven surfaces, etc., making the grinding process more precise and efficient.

[0080] In summary, by combining real-time monitoring of material removal amount and vision weld tracking technology, the grinding system can flexibly respond to various complex working conditions while maintaining high precision, dynamically optimize processing parameters, and ensure that the final grinding quality reaches the best state. This intelligent grinding method not only improves work efficiency, reduces the uncertainty brought by human factors, but also brings higher production standards and technical levels to the manufacturing industry. During the whole process, all adjustments are completed without affecting the overall production process, reflecting the advantages of modern intelligent manufacturing systems.

[0081] Figure 2 The structural block diagram of a weld seam material removal amount prediction and grinding device provided by an exemplary embodiment of the present invention based on the Preston equation is shown. The weld seam material removal amount prediction and grinding device based on the Preston equation is applicable to the weld seam material removal amount prediction and grinding method based on the Preston equation as described above. The device includes:

[0082] An acquisition module 201, configured to acquire processing parameter data during the historical weld seam grinding process, where the processing parameter data at least includes grinding force, inclination angle, and grinding wheel linear velocity;

[0083] A construction module 202, configured to construct a mathematical relationship model between the material removal amount and the processing parameter data based on the Preston equation;

[0084] A determination module 203, configured to train the processing parameter data through a machine learning algorithm, optimize the mathematical relationship model, and determine a weld seam material removal amount prediction model;

[0085] The determination module 203 is further configured to determine the predicted weld seam material removal amount based on the weld seam material removal amount prediction model;

[0086] The determination module 203 is further configured to determine the initial processing parameters according to the predicted weld seam material removal amount;

[0087] The regulation module 204 is used to perform weld grinding based on the initial processing parameters and regulate the processing parameters in real time. In a possible implementation, in step S2, the Preston equation includes:

[0088] G = K·P·v·cos(θ);

[0089] where G is the weld material removal amount, K is the proportionality constant, P is the grinding force, v is the linear speed of the grinding wheel, and θ is the inclination angle; the proportionality constant K is obtained by performing regression analysis on historical grinding data and reflects the removal rate characteristics between a specific grinding wheel material and the material to be ground; the grinding force P is dynamically adjusted to the optimal value according to the grinding effect feedback in real time; the linear speed v of the grinding wheel includes a preset multi-level speed setting to adapt to weld materials with different hardnesses and thicknesses; the measurement of the inclination angle θ is obtained in real time through an inclination angle sensor to ensure precise control and adjustment of the inclination angle θ.

[0090] In a possible implementation, in steps S1 to S2, the processing parameter data further includes a grinding wheel wear compensation factor and an environmental compensation factor, and the Preston equation includes:

[0091] G = K·P·v·cos(θ)·C wear ·C env ;

[0092] where C wear is the grinding wheel wear compensation factor, indicating the influence of grinding wheel wear on the weld material removal amount; C env is the environmental compensation factor, indicating the influence of environmental factors on the weld material removal amount;

[0093] C wear = 1 - α·W(t);

[0094] where α is the wear sensitivity coefficient; W(t) is the grinding wheel wear degree, indicating the wear amount of the grinding wheel within the usage time t;

[0095] C env = 1 + β T ·(T - T 0 ) + β H ·(H - H 0 );

[0096] where β T is the temperature sensitivity coefficient; β H is the humidity sensitivity coefficient; T is the ambient temperature monitored in real time; H is the ambient humidity monitored in real time; T 0 is the temperature under the reference environmental conditions; H 0 is the humidity under the reference environmental conditions.

[0097] In a possible implementation, in step S3:

[0098] The machine learning algorithm is a neural network algorithm. A neural network model is constructed using a deep learning framework. The neural network model includes an input layer, multiple hidden layers, and an output layer. The hidden layer uses an activation function, and historical grinding data is used to train the neural network model. The network weights are optimized through the backpropagation algorithm. After the neural network model is trained, a prediction model for the weld material removal amount is obtained.

[0099] The relationship between the input layer, hidden layer, and output layer of the neural network model is:

[0100] G = f(P, v, θ, C wear , C env );

[0101] where the input layer is the grinding force P, the grinding wheel linear velocity v, the inclination angle θ, the grinding wheel wear compensation factor C wear and the environmental compensation factor C env ; the hidden layer is to extract complex relationships using a non-linear activation function; the output layer is the weld material removal amount G.

[0102] In a possible implementation, in step S6:

[0103] In response to the weld grinding process, the weld material removal amount is monitored in real time, and error analysis is performed with the weld material removal amount predicted by the weld material removal amount prediction model. According to the error analysis results, the processing parameters are adjusted in real time.

[0104] The error analysis includes:

[0105] ΔG = G predicted - G actual ;

[0106] where G predicted and G actual are respectively the weld material removal amount predicted by the model and the weld material removal amount monitored in real time;

[0107] The adjustment of the processing parameters includes:

[0108] P new = P current + k p ·ΔG;

[0109] V new = V current + k v ·ΔG;

[0110] where k p and k vis the adjustment coefficient, which is used to adjust the grinding force and the linear speed of the grinding disc respectively; P current is the current grinding force; V current is the current linear speed of the grinding disc.

[0111] In a possible implementation, in step S6:

[0112] In response to the welding seam grinding process, the position of the welding seam is captured in real time through a visual welding seam tracker, and the grinding path is automatically adjusted to ensure full coverage of the welding seam;

[0113] The automatic adjustment of the grinding path includes:

[0114] ΔX = X predicted - X actual ;

[0115] where X predicted is the path generated by the model; X actual is the path monitored in real time;

[0116] X new = X current + ΔX;

[0117] where X current is the current path position; X new is the newly generated path position; The path optimization is realized through interpolation or curve fitting algorithm, so that the end of the grinding robot fits the shape of the welding seam.

[0118] It should be noted that: The above-mentioned welding seam material removal amount prediction and grinding device based on the Preston equation provided in the above embodiments are only illustrated by the division of the above functional modules. In actual applications, the above functions can be allocated to different functional modules according to needs, that is, the internal structure of the device is divided into different functional modules to complete all or part of the functions described above.

[0119] Figure 3 shows a schematic structural diagram of a computer device for executing the welding seam material removal amount prediction and grinding method based on the Preston equation provided by an exemplary embodiment of the present invention. The computer device includes:

[0120] The processor 301 includes one or more processing cores. The processor 301 executes various functional applications and data processing by running software programs and modules.

[0121] The receiver 302 and the transmitter 303 can be implemented as a communication component, which can be a communication chip. Optionally, the communication component can be implemented to include a signal transmission function. That is, the transmitter 303 can be used to transmit control signals to the image acquisition device and the scanning device, and the receiver 302 can be used to receive corresponding feedback instructions.

[0122] The memory 304 is connected to the processor 301 through the bus 305.

[0123] The memory 304 can be used to store at least one instruction, and the processor 301 is used to execute the at least one instruction to implement each step in the above method embodiments.

[0124] An embodiment of the present invention also provides a computer-readable storage medium, in which at least one instruction, at least one segment of program, code set or instruction set is stored, and is to be loaded and executed by a processor to implement the above-mentioned method for predicting the removal amount of weld material and grinding based on the Preston equation.

[0125] The present invention also provides a computer program product or a computer program, which includes computer instructions stored in a computer-readable storage medium. The processor of the computer device reads the computer instructions from the computer-readable storage medium, and the processor executes the computer instructions, so that the computer device executes the above-mentioned method for predicting the removal amount of weld material and grinding based on the Preston equation described in any one of the above embodiments.

[0126] Optionally, the computer-readable storage medium may include: read-only memory (ROM, Read Only Memory), random access memory (RAM, Random Access Memory), solid state drive (SSD, Solid State Drives), or optical disc, etc. Among them, the random access memory may include resistive random access memory (ReRAM, Resistance Random Access Memory) and dynamic random access memory (DRAM, Dynamic Random Access Memory). The serial numbers of the above embodiments of the present invention are only for description and do not represent the advantages and disadvantages of the embodiments.

[0127] It can be understood that the specific examples herein are only for helping those skilled in the art better understand the present disclosure, rather than limiting the scope of the present invention.

[0128] It can be understood that in various embodiments of this specification, the magnitude of the serial numbers of each process does not mean the sequence of execution, and the execution sequence of each process should be determined by its function and internal logic, and should not constitute any limitation to the implementation process of the present disclosure.

[0129] It is understood that the various embodiments described in this specification can be implemented alone or in combination, and the present disclosure does not limit this.

[0130] Unless otherwise specified, all technical and scientific terms used in this disclosure have the same meaning as commonly understood by those skilled in the technical field of this specification. The terms used in this specification are only for the purpose of describing specific embodiments and are not intended to limit the scope of this specification. The term "and / or" used in this specification includes any and all combinations of one or more of the related listed items. The singular forms "a", "above", and "the" used in this disclosure and the appended claims are also intended to include the plural forms unless the context clearly indicates otherwise.

[0131] It is understood that the processor of the present disclosure can be an integrated circuit chip with signal processing capabilities. In the implementation process, the steps of the above method embodiments can be completed by the integrated logic circuit in the hardware of the processor or instructions in software form. The above processor can be a general-purpose processor, a digital signal processor (DSP), an application specific integrated circuit (ASIC), a field programmable gate array (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components. It can implement or execute the various methods, steps and logic block diagrams disclosed in this disclosure. The general-purpose processor can be a microprocessor or the processor can also be any conventional processor, etc. The steps of the method disclosed in combination with this disclosure can be directly embodied as being executed and completed by a hardware decoding processor, or executed and completed by a combination of hardware and software modules in the decoding processor. The software module can be located in a mature storage medium in the art such as random access memory, flash memory, read-only memory, programmable read-only memory or electrically erasable programmable memory, register, etc. This storage medium is located in the memory, and the processor reads the information in the memory and combines its hardware to complete the steps of the above method.

[0132] It can be understood that the memory in the present disclosure can be a volatile memory or a non-volatile memory, or can include both volatile and non-volatile memories. Among them, the non-volatile memory can be a read-only memory (ROM), a programmable read-only memory (PROM), an erasable programmable read-only memory (EPROM), an electrically erasable programmable read-only memory (EEPROM), or a flash memory. The volatile memory can be a random access memory (RAM). It should be noted that the memories of the systems and methods described herein are intended to include but are not limited to these and any other suitable types of memories.

[0133] Those of ordinary skill in the art can realize that the units and algorithm steps of each example described in combination with the embodiments disclosed herein can be implemented by electronic hardware, or by a combination of computer software and electronic hardware. Whether these functions are executed in a hardware or software manner depends on the specific application and design constraints of the technical solution. Professional technicians can use different methods to implement the described functions for each specific application, but such implementation should not be considered to exceed the scope of this specification.

[0134] Those skilled in the art can clearly understand that for the convenience and brevity of description, the specific working processes of the systems, devices, and units described above can refer to the corresponding processes in the foregoing method embodiments and will not be elaborated herein.

[0135] In the several embodiments provided in this specification, it should be understood that the disclosed systems, devices, and methods can be implemented in other ways. For example, the device embodiments described above are merely illustrative. For example, the division of the units is only a logical function division, and there can be other division methods in actual implementation. For example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the displayed or discussed couplings or direct couplings or communication connections to each other can be through some interfaces. The indirect couplings or communication connections of the devices or units can be in an electrical, mechanical, or other form.

[0136] The units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they can be located in one place, or can be distributed to multiple network units. Some or all of the units can be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0137] In addition, the functional units in each embodiment of this specification can be integrated into one processing unit, or each unit can exist physically alone, or two or more units can be integrated into one unit.

[0138] When the above-mentioned function is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this specification, in essence, or the part that contributes to the prior art, or part of the technical solution, can be embodied in the form of a software product. The computer software product is stored in a storage medium and includes several instructions for causing a computer device (which may be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods described in various embodiments of this specification. The foregoing storage medium includes: various media such as USB flash drives, mobile hard disks, read-only memories (ROMs), random access memories (RAMs), magnetic disks, or optical discs that can store program codes.

[0139] As described above, the foregoing are only the specific embodiments of this specification, but the protection scope of the present invention is not limited thereto. Any person skilled in the art within the technical scope disclosed in this specification can easily think of changes or substitutions, which should all be covered by the protection scope of this specification. Therefore, the protection scope of the present invention should be subject to the protection scope of the claims.

Claims

1. A weld material removal prediction and grinding method based on Preston equation, characterized in that: The method adopts a grinding robot for processing, and the method comprises: S1. Acquire processing parameter data in a historical weld grinding process, wherein the processing parameter data at least includes grinding force, inclination angle, and grinding disc linear speed; S2, constructing a mathematical relationship model between the material removal amount and the processing parameter data based on the Preston equation; S3, training the processing parameter data through a machine learning algorithm, optimizing the mathematical relationship model, and determining a weld material removal amount prediction model; S4. Determine a predicted weld material removal amount based on the weld material removal amount prediction model; S5. Determine initial processing parameters according to the predicted weld material removal amount; S6. Perform weld grinding based on the initial processing parameters, and adjust the processing parameters in real time.

2. The weld material removal prediction and grinding method based on Preston equation according to claim 1, characterized in that: In step S2, the Preston equation includes: G = K·P·v·cos(θ); Wherein, G is the amount of weld material removed, K is the proportional constant, P is the grinding force, v is the grinding disc linear speed, and θ is the inclination angle; the proportional constant K is obtained by regression analysis of historical grinding data, reflecting the removal rate characteristics between a specific grinding disc material and a grinded material; the grinding force P is dynamically adjusted and adjusted to the optimal value according to the real-time feedback of the grinding effect; the grinding disc linear speed v includes a preset multi-level speed setting to adapt to weld materials of different hardness and thickness; the measurement of the inclination angle θ is obtained in real time through an inclination sensor to ensure the precise control and adjustment of the inclination angle θ.

3. The weld material removal prediction and grinding method based on Preston equation according to claim 2, characterized in that: In the steps S1 to S2, the processing parameter data also includes a grinding disc wear compensation factor and an environmental compensation factor, and the Preston equation includes: G=K·P·v·cos(θ)·C wear ·C env ; Among them, C wear is the grinding wheel wear compensation factor, which indicates the effect of grinding wheel wear on the weld material removal amount; C env is the environmental compensation factor, which indicates the influence of environmental factors on the amount of weld material removed; C wear =1-α·W(t); Where α is the wear sensitivity coefficient; W(t) is the wear degree of the grinding disc, which means the wear amount of the grinding disc in the use time t; C env =1+β T ·(T-T0)+β H ·(H-H0); Among them, β T is the temperature sensitivity coefficient; β H is the humidity sensitivity coefficient; T is the real-time monitored ambient temperature; H is the real-time monitored ambient humidity; T0 is the temperature under reference ambient conditions; H0 is the humidity under reference ambient conditions.

4. The weld material removal prediction and grinding method based on Preston equation according to claim 1, characterized in that: In step S3: The machine learning algorithm is a neural network algorithm, and a neural network model is constructed using a deep learning framework. The neural network model includes an input layer, multiple hidden layers, and an output layer. The hidden layer uses an activation function, and the neural network model is trained using historical grinding data. The network weight is optimized by a back propagation algorithm. After the neural network model is trained, a weld material removal amount prediction model is obtained; The relationship between the input layer, hidden layer and output layer of the neural network model is: G1f(P,v,θ,C wear ,C env )4 Among them, the input layer is the grinding force P, the grinding disc linear speed v, the inclination angle θ, and the grinding disc wear compensation factor C wear and environmental compensation factor C env ; The hidden layer uses nonlinear activation functions to extract complex relationships; the output layer is the weld material removal amount G.

5. The weld material removal prediction and grinding method based on Preston equation according to claim 1, characterized in that: In step S6: In response to the weld grinding process, the weld material removal amount is monitored in real time, and an error analysis is performed with the weld material removal amount predicted by the weld material removal amount prediction model, and the processing parameters are adjusted in real time according to the error analysis results; The error analysis includes: ΔG=G predicted -G actual ; Among them, G predicted and G actual They are the weld material removal amount predicted by the model and the real-time monitoring; The control processing parameters include: P new =P current +k p ·ΔG; V new =V current +k v ·ΔG; Among them, k p and k v is the adjustment coefficient, which is used to adjust the grinding force and the grinding disc linear speed respectively; P current is the current grinding force; V current is the current grinding disc linear speed.

6. The weld material removal prediction and grinding method based on Preston equation according to claim 1, characterized in that: In step S6: In response to the weld grinding process, the weld position is captured in real time through the visual weld tracker, and the grinding path is automatically adjusted to ensure full coverage of the weld; The automatic adjustment of the grinding path comprises: ΔX=X predicted -X actual ; Among them, X predicted The path generated for the model; X actual The path for real-time monitoring; X new =X current +ΔX; Among them, X current is the current path position; X new is the newly generated path position; path optimization is achieved through interpolation or curve fitting algorithm to make the end of the grinding robot fit the weld shape.

7. A weld material removal prediction and grinding device based on Preston equation, characterized in that: The device is applicable to the weld material removal amount prediction and grinding method based on the Preston equation as described in any one of claims 1 to 6, and the device comprises: An acquisition module, used to acquire processing parameter data in a historical weld grinding process, wherein the processing parameter data at least includes grinding force, inclination angle, and grinding disc linear speed; A construction module, for constructing a mathematical relationship model between the material removal amount and the processing parameter data based on the Preston equation; A determination module, used to train the processing parameter data through a machine learning algorithm, optimize the mathematical relationship model, and determine a weld material removal amount prediction model; The determination module is further used to determine the predicted weld material removal amount based on the weld material removal amount prediction model; The determination module is further used to determine initial processing parameters according to the predicted weld material removal amount; The control module is used to perform weld grinding processing based on the initial processing parameters and to control the processing parameters in real time.

8. A computer device, characterized in that: The computer device includes a processor and a memory, wherein the memory stores at least one instruction, at least one program, a code set or an instruction set, and the at least one instruction, at least one program, a code set or an instruction set is loaded and executed by the processor to implement the weld material removal amount prediction and grinding method based on the Preston equation as described in any one of claims 1 to 6.

9. A computer-readable storage medium, characterized in that: The readable storage medium stores at least one instruction, at least one program, code set or instruction set, and the at least one instruction, at least one program, code set or instruction set is loaded and executed by the processor to implement the weld material removal amount prediction and grinding method based on the Preston equation as described in any one of claims 1 to 6.

Images