Thermal error modeling and compensating method for blade five-axis machining center

Abstract

The invention discloses a thermal error modeling and compensation method for a blade five-axis machining center, and relates to the technical field of temperature compensation. A blade five-axis machining three-dimensional model is constructed through a blade structure and a machining process, and a plurality of monitoring points are uniformly arranged in the constructed blade five-axis machining three-dimensional model; the method comprises the following steps: setting a plurality of monitoring points, arranging temperature acquisition equipment based on the plurality of set monitoring points, acquiring historical temperature data of the corresponding monitoring points based on the arranged temperature acquisition equipment, processing the acquired historical temperature data of the corresponding monitoring points after acquisition is completed, and acquiring thermal deformation data corresponding to the processed historical temperature data at the same time; the method comprises the steps of processing historical temperature data of a blade five-axis machining center, building a thermal error prediction model based on the processed historical temperature data, finally collecting temperature data of a corresponding monitoring point in real time, verifying the thermal error prediction model, and determining thermal error compensation of the blade five-axis machining center based on the thermal error prediction model after verification is passed. And the thermal error compensation accuracy of the chip five-axis machining center is improved.

Description

Technical Field

[0001] This invention relates to the field of temperature compensation technology, specifically to a method for thermal error modeling and compensation for five-axis machining centers for blades. Background Technology

[0002] As a key component of wind power generation, the machining accuracy of blades directly determines the quality and performance of the product. The blade spindle, as one of the core components of wind power generation, undertakes the tasks of rotating and positioning the workpiece or cutting tool during machining, serving as the positional and motion reference in actual processing. The accuracy and stability of the spindle have a crucial impact on the machining accuracy of the workpiece. However, during the actual machining process of blades, the spindle is affected by various factors, among which temperature has a particularly significant impact. The influence of temperature on the blade spindle is mainly reflected in thermal errors. Thermal errors are changes in the size, shape, and relative position of various blade components caused by temperature variations, thus leading to machining errors. Summary of the Invention

[0003] The purpose of this invention is to provide a thermal error modeling and compensation method for five-axis machining centers for blades, which solves the problems existing in the background art.

[0004] To solve the above-mentioned technical problems, the present invention adopts the following technical solution: The present invention provides a method for thermal error modeling and compensation for five-axis machining centers for blades, specifically including the following steps: S1. Construct a three-dimensional model of blade five-axis machining based on blade structure and machining process. After the model is constructed, set multiple monitoring points evenly in the constructed three-dimensional model of blade five-axis machining. S2. Deploy temperature acquisition devices based on multiple set monitoring points, and collect historical temperature data of the corresponding monitoring points based on the deployed temperature acquisition devices; S3. The historical temperature data of the corresponding monitoring points collected are processed through data processing methods to obtain the processed historical temperature data. S4. Collect and process the thermal deformation data corresponding to the historical temperature data, and construct a thermal error prediction model based on the processed historical temperature data through data simulation. S5. Real-time acquisition of temperature data at corresponding monitoring points, verification of the constructed thermal error prediction model through multiple verification methods, and determination of thermal error compensation for the five-axis machining center of the blade based on the thermal error prediction model after verification.

[0005] Preferably, the step of constructing a five-axis machining three-dimensional model of the blade based on the blade structure and machining process, and then uniformly setting multiple monitoring points in the constructed five-axis machining three-dimensional model of the blade after construction includes the following steps: Collect blade structure and processing flow data, input the blade structure and processing flow data into 3D model building software, and obtain the corresponding 3D model of the blade; Based on the obtained three-dimensional model of the corresponding leaf, the leaf is segmented by slicing to obtain the leaf image after the three-dimensional model of the corresponding leaf is segmented. Based on the segmented leaf image, multiple monitoring points are uniformly set on the obtained segmented leaf image.

[0006] Preferably, the step of deploying temperature acquisition devices based on multiple set monitoring points and collecting historical temperature data for the corresponding monitoring points based on the deployed temperature acquisition devices includes the following steps: Based on the positions of all monitoring points on the segmented leaf image, the relative position of the temperature acquisition device is fixed so that all monitoring points on the segmented leaf image are evenly distributed in the field of view of the temperature acquisition device. Set the acquisition cycle of the temperature acquisition device, and collect historical temperature data of the corresponding monitoring point based on the set acquisition cycle.

[0007] Preferably, the step of processing the historical temperature data of the corresponding monitoring points to obtain the processed historical temperature data includes the following steps: S31. Filter the historical temperature data of the corresponding monitoring points by the method of measuring point screening to obtain the filtered historical temperature data; Since temperature changes can cause thermal deformation of the blades, thermal deformation data of the blades under corresponding historical temperature data is collected. The shared information between historical temperature data of corresponding monitoring points is analyzed by calculating the maximum mutual information coefficient. The formula for calculating the maximum mutual information coefficient is as follows: ; in, These represent the values ​​of the thermal deformation data variable and the historical temperature data variable, respectively. Representing variables Mutual information between them Representing variables The joint probability density function, Representing variables marginal probability density, Representing variables marginal probability density, Representing variables The differential; Based on the shared information among the historical temperature data of corresponding monitoring points, the corresponding monitoring points are filtered to obtain the filtered historical temperature data; S32. The filtered historical temperature data is processed using data processing methods to obtain processed historical temperature data.

[0008] Preferably, the step of processing the filtered historical temperature data to obtain processed historical temperature data includes the following steps: The standardized calculation formula is as follows: ; in, This represents the minimum value among the filtered historical temperature data. This represents the maximum value among the filtered historical temperature data. This represents the historical temperature data after the Sth filter. This represents the Sth historical temperature data after standardization.

[0009] Preferably, the process of collecting and processing historical temperature data corresponding to thermal deformation data, and constructing a thermal error prediction model based on the processed historical temperature data through data simulation, includes the following steps: S41. Based on the processed historical temperature data, construct the thermal error relationship through data simulation. Thermal deformation data at corresponding locations were recorded by controlling variable method and data simulation method for every 1℃ increase in temperature; A thermal error relationship is constructed based on the recorded thermal deformation data at the corresponding locations; The thermal error relationship is as follows: ; in, Indicates the thermal error relationship. This indicates the thermal deformation data at the corresponding location. This represents the temperature change data at the corresponding location; S42. Based on the constructed thermal error relationship, a thermal error prediction model is constructed using a neural network.

[0010] Preferably, the step of constructing a thermal error prediction model using a neural network based on the constructed thermal error relationship includes the following steps: S421. Set the neural network structure and collect the corresponding processed historical temperature data, thermal deformation data at the corresponding location, and thermal error relationship. The LSTM network is configured to include input gates, output gates, and forget gates; The input set of the input gate is set as follows ,in Indicates the first The historical temperature data processed for the corresponding location is output from the output layer. ,in This represents the first [unclear] based on the thermal error relationship and the output. The predicted thermal deformation data for the corresponding location; S422. Input the collected and processed historical temperature data into the LSTM network, and input the thermal error relationship and the corresponding thermal deformation data as candidate values ​​into the LSTM network. S423. Process the current input and the corresponding thermal deformation data from the previous time step using the forget gate, and update them accordingly; S424. The corresponding position thermal deformation data after being processed by the forget gate is output through the output gate to obtain the predicted corresponding position thermal deformation data at the next moment. S425. Iteratively use the LSTM network to determine and obtain the trained LSTM network; Set an error threshold between the thermal deformation data at the corresponding location predicted by the LSTM network and the collected thermal deformation data at the corresponding location. When the error between the thermal deformation data at the corresponding location predicted by the LSTM network and the collected thermal deformation data at the corresponding location is within the set error threshold range, the iteration stops, and the trained LSTM network is obtained. S426. Set an LSTM network that meets the error threshold range as the thermal error prediction model, and output the corresponding thermal deformation data at the next time step after prediction.

[0011] Preferably, the real-time acquisition of temperature data at the corresponding monitoring points, and the verification of the constructed thermal error prediction model through multiple verification methods, after successful verification, the determination of thermal error compensation for the five-axis machining center of the blade based on the thermal error prediction model includes the following steps: Temperature data at the corresponding monitoring points are collected in real time, and the temperature data at the corresponding monitoring points are input into the constructed thermal error prediction model to obtain the predicted thermal deformation data at the corresponding location. A validation population was constructed by summarizing the predicted thermal deformation data at the corresponding locations; The constructed thermal error prediction model was validated using the antlion algorithm; The solution space is constructed based on the constructed verification population. Let the position of each ant in the solution space be: ; Each ant is set to randomly choose to surround its prey or use a trap to drive it away, and the probability of each ant choosing these two behaviors is equal. The position of each ant in the solution space is defined as a set of temperature data and corresponding thermal deformation data for a monitoring point; The process of surrounding the prey includes the following steps: When ants surround prey, they will choose to move towards the ant in the best position or towards a random ant. When the ant moves towards the optimal position: The formula for updating the ant's position is as follows:

[0012] in, Let represent the optimal ant position at time t. It is a random number. This represents the position of the z-th ant at time t+1. This represents the position of the z-th ant at time t; The use of traps to drive away prey includes the following steps: When ants use traps to drive away prey, they also constantly update their own location; When using a trap, the ant's position is updated using the following formula:

[0013] in, The default value for a constant is 1. Let e ​​be a random number uniformly distributed in the range [-1, 1], and let e represent the natural constant. The ant optimization algorithm is used iteratively until the position of each ant in the solution space converges, and the temperature data and corresponding thermal deformation data of the corresponding monitoring point are output. The temperature data and corresponding thermal deformation data of the corresponding monitoring point are set to be used as the thermal error compensation data of the five-axis machining center of the blade at the next moment.

[0014] The present invention also provides a method for thermal error modeling and compensation for a five-axis machining center for blades, comprising: a data acquisition module, a data processing module, a thermal error prediction module, and a thermal error compensation module; The data acquisition module is used to collect temperature data and thermal deformation data at the corresponding monitoring points; The data processing module is used to process the collected temperature data to obtain processed temperature data; The thermal error prediction module is used to construct a thermal error prediction module based on the processed temperature data and the collected thermal deformation data. The thermal error compensation module is used to determine thermal error compensation data based on the constructed thermal error prediction module.

[0015] The beneficial effects of this invention are as follows: (1) This invention constructs a three-dimensional model of blade five-axis machining through blade structure and processing flow, and uniformly sets multiple monitoring points in the constructed three-dimensional model of blade five-axis machining. At the same time, temperature acquisition devices are set up based on the set multiple monitoring points, and historical temperature data of the corresponding monitoring points are collected based on the set temperature acquisition devices. After the collection is completed, the historical temperature data of the corresponding monitoring points is processed, and the thermal deformation data corresponding to the processed historical temperature data is collected. Based on the processed historical temperature data, a thermal error prediction model is constructed. Finally, the temperature data of the corresponding monitoring points is collected in real time, and the thermal error prediction model is verified. After the verification is passed, the thermal error compensation of the blade five-axis machining center is determined based on the thermal error prediction model, which improves the accuracy of thermal error compensation of the blade five-axis machining center.

[0016] (2) This invention records the thermal deformation data of the corresponding position for every 1°C increase in temperature by using the controlled variable method and data simulation based on the processed historical temperature data. At the same time, it constructs the thermal error relationship based on the recorded thermal deformation data of the corresponding position. After the construction is completed, the thermal error prediction model is determined by the neural network. After the thermal error prediction model is determined, the temperature data of the corresponding monitoring point is collected in real time and the thermal error prediction model is verified. At the same time, the thermal error compensation of the five-axis machining center of the blade is determined based on the verified thermal error prediction model, which improves the rationality of thermal error compensation. Attached Figure Description

[0017] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0018] Figure 1 This is a schematic diagram of the method flow of the present invention. Detailed Implementation

[0019] 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.

[0020] In a specific embodiment of the present invention, Reference Figure 1 As shown, this invention provides a method for thermal error modeling and compensation for five-axis machining centers for blades, including: S1. Construct a three-dimensional model of blade five-axis machining based on blade structure and machining process. After the model is constructed, set multiple monitoring points evenly in the constructed three-dimensional model of blade five-axis machining. S2. Deploy temperature acquisition devices based on multiple set monitoring points, and collect historical temperature data of the corresponding monitoring points based on the deployed temperature acquisition devices; S3. The historical temperature data of the corresponding monitoring points collected are processed through data processing methods to obtain the processed historical temperature data. S4. Collect and process the thermal deformation data corresponding to the historical temperature data, and construct a thermal error prediction model based on the processed historical temperature data through data simulation. S5. Real-time acquisition of temperature data at corresponding monitoring points, verification of the constructed thermal error prediction model through multiple verification methods, and determination of thermal error compensation for the five-axis machining center of the blade based on the thermal error prediction model after verification.

[0021] Furthermore, referring to Figure 1 As shown, a five-axis machining 3D model of the blade is constructed based on the blade structure and machining process. After the model is constructed, multiple monitoring points are uniformly set in the constructed five-axis machining 3D model of the blade, including the following steps: Collect blade structure and processing flow data, input the blade structure and processing flow data into 3D model building software, and obtain the corresponding 3D model of the blade; Furthermore, based on the obtained three-dimensional model of the corresponding leaf, the leaf is segmented by slicing to obtain the leaf image after the three-dimensional model of the corresponding leaf is segmented; Furthermore, based on the segmented leaf image, multiple monitoring points are uniformly set on the obtained segmented leaf image; Furthermore, referring to Figure 1 As shown, the process of deploying temperature acquisition devices at multiple designated monitoring points and collecting historical temperature data for the corresponding monitoring points using these devices includes the following steps: Based on the positions of all monitoring points on the segmented leaf image, the relative position of the temperature acquisition device is fixed so that all monitoring points on the segmented leaf image are evenly distributed in the field of view of the temperature acquisition device. Furthermore, the acquisition cycle of the temperature acquisition device is set, and historical temperature data of the corresponding monitoring point is collected based on the set acquisition cycle; Furthermore, referring to Figure 1 As shown, the historical temperature data collected from the corresponding monitoring points is processed using data processing methods to obtain the processed historical temperature data, including the following steps: S31. Filter the historical temperature data of the corresponding monitoring points by the method of measuring point screening to obtain the filtered historical temperature data; Since temperature changes can cause thermal deformation of the blades, thermal deformation data of the blades under corresponding historical temperature data is collected. The shared information between historical temperature data of corresponding monitoring points is analyzed by calculating the maximum mutual information coefficient. The formula for calculating the maximum mutual information coefficient is as follows: ; in, These represent the values ​​of the thermal deformation data variable and the historical temperature data variable, respectively. Representing variables Mutual information between them Representing variables The joint probability density function, Representing variables marginal probability density, Representing variables marginal probability density, Representing variables The differential; Furthermore, based on the shared information among the historical temperature data of the corresponding monitoring points, the corresponding monitoring points are filtered to obtain the filtered historical temperature data; S32. The filtered historical temperature data is processed using data processing methods to obtain processed historical temperature data. The standardized calculation formula is as follows: ; in, This represents the minimum value among the filtered historical temperature data. This represents the maximum value among the filtered historical temperature data. This represents the historical temperature data after the Sth filter. This represents the Sth historical temperature data after standardization. Furthermore, referring to Figure 1 As shown, the process involves collecting and processing historical temperature data corresponding to thermal deformation data, and then constructing a thermal error prediction model based on this processed historical temperature data using data simulation. The steps include: S41. Based on the processed historical temperature data, construct the thermal error relationship through data simulation. Thermal deformation data at corresponding locations were recorded by controlling variable method and data simulation method for every 1℃ increase in temperature; Furthermore, a thermal error relationship is constructed based on the recorded thermal deformation data at the corresponding locations; The thermal error relationship is as follows: ; in, Indicates the thermal error relationship. This indicates the thermal deformation data at the corresponding location. This represents the temperature change data at the corresponding location; S42. Based on the constructed thermal error relationship, a thermal error prediction model is built using a neural network; S421. Set the neural network structure and collect the corresponding processed historical temperature data, thermal deformation data at the corresponding location, and thermal error relationship. The LSTM network is configured to include input gates, output gates, and forget gates; The input set of the input gate is set as follows ,in Indicates the first The historical temperature data processed for the corresponding location is output from the output layer. ,in This represents the first [unclear] based on the thermal error relationship and the output. The predicted thermal deformation data for the corresponding location; S422. Input the collected and processed historical temperature data into the LSTM network, and input the thermal error relationship and the corresponding thermal deformation data as candidate values ​​into the LSTM network. S423. Process the current input and the corresponding thermal deformation data from the previous time step using the forget gate, and update them accordingly; S424. The corresponding position thermal deformation data after being processed by the forget gate is output through the output gate to obtain the predicted corresponding position thermal deformation data at the next moment. S425. Iteratively use the LSTM network to determine and obtain the trained LSTM network; Set an error threshold between the thermal deformation data at the corresponding location predicted by the LSTM network and the collected thermal deformation data at the corresponding location. When the error between the thermal deformation data at the corresponding location predicted by the LSTM network and the collected thermal deformation data at the corresponding location is within the set error threshold range, the iteration stops, and the trained LSTM network is obtained. S426. Set an LSTM network that meets the error threshold range as the thermal error prediction model, and output the thermal deformation data at the corresponding position after the prediction at the next time step. Furthermore, referring to Figure 1 As shown, temperature data at corresponding monitoring points are collected in real time. The constructed thermal error prediction model is validated through multiple validation methods. After successful validation, the thermal error compensation for the five-axis machining center of the blade is determined based on the thermal error prediction model, including the following steps: Temperature data at the corresponding monitoring points are collected in real time, and the temperature data at the corresponding monitoring points are input into the constructed thermal error prediction model to obtain the predicted thermal deformation data at the corresponding location. A validation population was constructed by summarizing the predicted thermal deformation data at the corresponding locations; Furthermore, the constructed thermal error prediction model was validated using the antlion algorithm; The solution space is constructed based on the constructed verification population. Let the position of each ant in the solution space be: ; Each ant is set to randomly choose to surround its prey or use a trap to drive it away, and the probability of each ant choosing these two behaviors is equal. The position of each ant in the solution space is defined as a set of temperature data and corresponding thermal deformation data for a monitoring point; The process of surrounding the prey includes the following steps: When ants surround prey, they will choose to move towards the ant in the best position or towards a random ant. When the ant moves towards the optimal position: The formula for updating the ant's position is as follows:

[0022] in, Let represent the optimal ant position at time t. It is a random number. This represents the position of the z-th ant at time t+1. This represents the position of the z-th ant at time t; The use of traps to drive away prey includes the following steps: When ants use traps to drive away prey, they also constantly update their own location; When using a trap, the ant's position is updated using the following formula:

[0023] in, The default value for a constant is 1. Let e ​​be a random number uniformly distributed in the range [-1, 1], and let e represent the natural constant. Furthermore, the ant optimization algorithm is used iteratively until the position of each ant in the solution space converges, and the temperature data and corresponding thermal deformation data of the corresponding monitoring point are output. The temperature data and thermal deformation data of the corresponding monitoring points are set to be the thermal error compensation data of the five-axis machining center of the blade at the next moment. In one specific embodiment, the thermal error modeling and compensation method for a five-axis machining center for blades further includes: a data acquisition module, a data processing module, a thermal error prediction module, and a thermal error compensation module. The data acquisition module is used to collect temperature data and thermal deformation data at the corresponding monitoring points; The data processing module is used to process the collected temperature data to obtain processed temperature data; The thermal error prediction module is used to construct a thermal error prediction module based on the processed temperature data and the collected thermal deformation data. The thermal error compensation module is used to determine thermal error compensation data based on the constructed thermal error prediction module.

[0024] It should be noted that, The above content is merely an example and illustration of the concept 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 concept of the invention or exceed the scope defined by the present invention, and all such modifications and additions should fall within the protection scope of the present invention.

Claims

1. A method for thermal error modeling and compensation in a five-axis machining center for blades, characterized in that, Includes the following steps: S1. Construct a three-dimensional model of the blade's five-axis machining based on the blade structure and machining process. After the model is constructed, set multiple monitoring points evenly in the constructed three-dimensional model of the blade's five-axis machining. S2. Deploy temperature acquisition devices based on multiple set monitoring points, and collect historical temperature data of the corresponding monitoring points based on the deployed temperature acquisition devices; S3. The historical temperature data of the corresponding monitoring points collected are processed through data processing methods to obtain the processed historical temperature data. S4. Collect and process the thermal deformation data corresponding to the historical temperature data, and construct a thermal error prediction model based on the processed historical temperature data through data simulation. S5. Real-time acquisition of temperature data at corresponding monitoring points, verification of the constructed thermal error prediction model through multiple verification methods, and determination of thermal error compensation for the five-axis machining center of the blade based on the thermal error prediction model after verification.

2. The method for thermal error modeling and compensation for a five-axis machining center for blades according to claim 1, characterized in that, The process of constructing a five-axis machining 3D model of the blade based on its structure and manufacturing process, and then uniformly setting multiple monitoring points in the constructed five-axis machining 3D model of the blade, includes the following steps: Collect blade structure and processing flow data, input the blade structure and processing flow data into 3D model building software, and obtain the corresponding 3D model of the blade; Based on the obtained three-dimensional model of the corresponding leaf, the leaf is segmented by slicing to obtain the leaf image after the three-dimensional model of the corresponding leaf is segmented. Based on the segmented leaf image, multiple monitoring points are uniformly set on the obtained segmented leaf image.

3. The method for thermal error modeling and compensation for a five-axis machining center for blades according to claim 1, characterized in that, The process of deploying temperature acquisition devices at multiple designated monitoring points and collecting historical temperature data for the corresponding monitoring points using these devices includes the following steps: Based on the positions of all monitoring points on the segmented leaf image, the relative position of the temperature acquisition device is fixed so that all monitoring points on the segmented leaf image are evenly distributed in the field of view of the temperature acquisition device. Set the acquisition cycle of the temperature acquisition device, and collect historical temperature data of the corresponding monitoring point based on the set acquisition cycle.

4. The method for thermal error modeling and compensation for a five-axis machining center for blades according to claim 1, characterized in that, The process of processing the historical temperature data collected from the corresponding monitoring points to obtain the processed historical temperature data includes the following steps: S31. Filter the historical temperature data of the corresponding monitoring points by the measurement point screening method to obtain the filtered historical temperature data; Since temperature changes can cause thermal deformation of the blades, thermal deformation data of the blades under corresponding historical temperature data is collected. The shared information between historical temperature data of corresponding monitoring points is analyzed by calculating the maximum mutual information coefficient. The formula for calculating the maximum mutual information coefficient is as follows: ； in, These represent the values ​​of the thermal deformation data variable and the historical temperature data variable, respectively. Representing variables Mutual information between them Representing variables The joint probability density function, Representing variables marginal probability density, Representing variables marginal probability density, Representing variables The differential; Based on the shared information among the historical temperature data of corresponding monitoring points, the corresponding monitoring points are filtered to obtain the filtered historical temperature data; S32. The filtered historical temperature data is processed using data processing methods to obtain processed historical temperature data.

5. The method for thermal error modeling and compensation for a five-axis machining center for blades according to claim 4, characterized in that, The process of processing the filtered historical temperature data to obtain processed historical temperature data includes the following steps: The standardized calculation formula is as follows: ； in, This represents the minimum value among the filtered historical temperature data. This represents the maximum value among the filtered historical temperature data. This represents the historical temperature data after the Sth filter. This represents the Sth historical temperature data after standardization.

6. The method for thermal error modeling and compensation for a five-axis machining center for blades according to claim 1, characterized in that, The process involves collecting and processing historical temperature data corresponding to thermal deformation data, and then constructing a thermal error prediction model based on this processed historical temperature data using data simulation. This includes the following steps: S41. Based on the processed historical temperature data, construct the thermal error relationship through data simulation. Thermal deformation data at corresponding locations were recorded by controlling variable method and data simulation method for every 1℃ increase in temperature; A thermal error relationship is constructed based on the recorded thermal deformation data at the corresponding locations; The thermal error relationship is as follows: ； in, Indicates the thermal error relationship. This indicates the thermal deformation data at the corresponding location. This represents the temperature change data at the corresponding location; S42. Based on the constructed thermal error relationship, a thermal error prediction model is constructed using a neural network.

7. The method for thermal error modeling and compensation for a five-axis machining center for blades according to claim 6, characterized in that, The process of constructing a thermal error prediction model based on the established thermal error relationship and using a neural network includes the following steps: S421. Set the neural network structure and collect the corresponding processed historical temperature data, thermal deformation data at the corresponding location, and thermal error relationship. The LSTM network is configured to include input gates, output gates, and forget gates; Set the input set of the input gate as follows ,in Indicates the first The historical temperature data processed for the corresponding location is output from the output layer. ,in This represents the first [unclear] based on the thermal error relationship and the output. The predicted thermal deformation data for the corresponding location; S422. Input the collected and processed historical temperature data into the LSTM network, and input the thermal error relationship and the corresponding thermal deformation data as candidate values ​​into the LSTM network. S423. Process the current input and the corresponding thermal deformation data from the previous time step using the forget gate, and update them accordingly; S424. The corresponding position thermal deformation data after being processed by the forget gate is output through the output gate to obtain the predicted corresponding position thermal deformation data at the next moment. S425. Iteratively use the LSTM network to determine and obtain the trained LSTM network; Set an error threshold between the thermal deformation data at the corresponding location predicted by the LSTM network and the collected thermal deformation data at the corresponding location. When the error between the thermal deformation data at the corresponding location predicted by the LSTM network and the collected thermal deformation data at the corresponding location is within the set error threshold range, the iteration stops, and the trained LSTM network is obtained. S426. Set an LSTM network that meets the error threshold range as the thermal error prediction model, and output the corresponding thermal deformation data at the next time step after prediction.

8. The method for thermal error modeling and compensation for a five-axis machining center for blades according to claim 1, characterized in that, The real-time acquisition of temperature data at corresponding monitoring points is used to validate the constructed thermal error prediction model through multiple validation methods. After successful validation, the thermal error compensation for the five-axis machining center of the blade is determined based on the thermal error prediction model, including the following steps: Temperature data at the corresponding monitoring points are collected in real time, and the temperature data at the corresponding monitoring points are input into the constructed thermal error prediction model to obtain the predicted thermal deformation data at the corresponding location. A validation population was constructed by summarizing the predicted thermal deformation data at the corresponding locations; The constructed thermal error prediction model was validated using the antlion algorithm; The solution space is constructed based on the constructed verification population. Let the position of each ant in the solution space be: ; Each ant is set to randomly choose to surround its prey or use a trap to drive it away, and the probability of each ant choosing these two behaviors is equal. The position of each ant in the solution space is defined as a set of temperature data and corresponding thermal deformation data for a monitoring point; The process of surrounding the prey includes the following steps: When ants surround prey, they will choose to move towards the ant in the best position or towards a random ant. When the ant moves towards the optimal position: The formula for updating the ant's position is as follows: ； in, Let represent the optimal ant position at time t. It is a random number. This represents the position of the z-th ant at time t+1. This represents the position of the z-th ant at time t; The use of traps to drive away prey includes the following steps: When ants use traps to drive away prey, they also constantly update their own location; When using a trap, the ant's position is updated using the following formula: ； in, The default value for a constant is 1. Let e ​​be a random number uniformly distributed in the range [-1, 1], and let e represent the natural constant. The ant optimization algorithm is used iteratively until the position of each ant in the solution space converges, and the temperature data and corresponding thermal deformation data of the corresponding monitoring point are output. The temperature data and corresponding thermal deformation data of the corresponding monitoring point are set to be used as the thermal error compensation data of the five-axis machining center of the blade at the next moment.

9. A method for thermal error modeling and compensation for a five-axis machining center for blades as described in claim 1, characterized in that, Also includes: The module includes a data acquisition module, a data processing module, a thermal error prediction module, and a thermal error compensation module. The data acquisition module is used to collect temperature data and thermal deformation data at the corresponding monitoring points; The data processing module is used to process the collected temperature data to obtain processed temperature data; The thermal error prediction module is used to construct a thermal error prediction module based on the processed temperature data and the collected thermal deformation data. The thermal error compensation module is used to determine thermal error compensation data based on the constructed thermal error prediction module.

Images