CN111230740B - Method and device for predicting grinding burn of aero-engine blade robot - Google Patents
Method and device for predicting grinding burn of aero-engine blade robot Download PDFInfo
- Publication number
- CN111230740B CN111230740B CN202010034441.9A CN202010034441A CN111230740B CN 111230740 B CN111230740 B CN 111230740B CN 202010034441 A CN202010034441 A CN 202010034441A CN 111230740 B CN111230740 B CN 111230740B
- Authority
- CN
- China
- Prior art keywords
- grinding
- workpiece
- burn
- grinding burn
- parameters
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Images
Classifications
-
- B—PERFORMING OPERATIONS; TRANSPORTING
- B24—GRINDING; POLISHING
- B24B—MACHINES, DEVICES, OR PROCESSES FOR GRINDING OR POLISHING; DRESSING OR CONDITIONING OF ABRADING SURFACES; FEEDING OF GRINDING, POLISHING, OR LAPPING AGENTS
- B24B49/00—Measuring or gauging equipment for controlling the feed movement of the grinding tool or work; Arrangements of indicating or measuring equipment, e.g. for indicating the start of the grinding operation
- B24B49/003—Measuring or gauging equipment for controlling the feed movement of the grinding tool or work; Arrangements of indicating or measuring equipment, e.g. for indicating the start of the grinding operation involving acoustic means
-
- B—PERFORMING OPERATIONS; TRANSPORTING
- B24—GRINDING; POLISHING
- B24B—MACHINES, DEVICES, OR PROCESSES FOR GRINDING OR POLISHING; DRESSING OR CONDITIONING OF ABRADING SURFACES; FEEDING OF GRINDING, POLISHING, OR LAPPING AGENTS
- B24B21/00—Machines or devices using grinding or polishing belts; Accessories therefor
- B24B21/18—Accessories
-
- B—PERFORMING OPERATIONS; TRANSPORTING
- B24—GRINDING; POLISHING
- B24B—MACHINES, DEVICES, OR PROCESSES FOR GRINDING OR POLISHING; DRESSING OR CONDITIONING OF ABRADING SURFACES; FEEDING OF GRINDING, POLISHING, OR LAPPING AGENTS
- B24B49/00—Measuring or gauging equipment for controlling the feed movement of the grinding tool or work; Arrangements of indicating or measuring equipment, e.g. for indicating the start of the grinding operation
- B24B49/16—Measuring or gauging equipment for controlling the feed movement of the grinding tool or work; Arrangements of indicating or measuring equipment, e.g. for indicating the start of the grinding operation taking regard of the load
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01D—MEASURING NOT SPECIALLY ADAPTED FOR A SPECIFIC VARIABLE; ARRANGEMENTS FOR MEASURING TWO OR MORE VARIABLES NOT COVERED IN A SINGLE OTHER SUBCLASS; TARIFF METERING APPARATUS; MEASURING OR TESTING NOT OTHERWISE PROVIDED FOR
- G01D21/00—Measuring or testing not otherwise provided for
- G01D21/02—Measuring two or more variables by means not covered by a single other subclass
Abstract
The invention discloses a method for predicting grinding burn of an aircraft engine blade robot, which comprises the steps of establishing a grinding burn prediction model; the grinding burn prediction model is configured to take a grinding machining parameter as an input quantity, take a grinding burn characteristic value as a state quantity and take a grinding burn degree as an output quantity; inputting the grinding parameters of the workpiece into the grinding burn prediction model to obtain the grinding burn degree of the workpiece; and judging the grinding burn degree of the workpiece, comparing the grinding processing parameter of the workpiece with the grinding burn critical threshold value according to the judgment result, and adjusting the grinding processing parameter of the workpiece according to the comparison result. The invention uses a plurality of sensors such as an acoustic emission sensor, a force sensor, an acceleration sensor, a temperature sensor, a current and voltage sensor and the like to form a blade robot abrasive belt grinding multi-sensor monitoring system, can acquire signals such as acoustic emission, force, vibration and the like on line in real time, and realizes comprehensive monitoring on the machining of a robot with a complex curved surface.
Description
Technical Field
The invention belongs to the field of abrasive belt grinding processing of an aircraft engine blade robot, and particularly relates to a method and a device for predicting grinding burn of an aircraft engine blade robot.
Background
The blade is a core power component of an aircraft engine power device, and the quality and the profile precision of the machined surface directly influence the working efficiency and the service life of the engine. The aero-engine blade usually operates in a high-pressure, high-temperature and high-speed environment, so that the new-generation aero-engine blade adopts titanium alloy, nickel-based high-temperature alloy and other difficult-to-process materials with high alloying degree as raw materials, is manufactured into a blank through a forging process, and is manufactured into the blade through a milling process, the surface precision of the blade at the moment still cannot meet the use requirement, and a grinding process is generally required to meet the manufacturing precision requirement.
The grinding processing technology is mainly divided into grinding wheel grinding and abrasive belt grinding according to the type of a processing tool, and compared with the grinding wheel grinding, the abrasive belt grinding has the advantages of high grinding efficiency, small system vibration, lower grinding temperature, elastic contact and the like, so that the abrasive belt grinding technology is widely applied to the field of processing of aeroengine blades; the processing method mainly comprises the following steps: manual grinding and polishing, numerical control machine tool grinding and polishing and robot grinding and polishing. Compared with the former two processing modes, the robot grinding and polishing has the characteristics of good flexibility, large operation space, strong expansibility and the like, and is very suitable for grinding and polishing complex curved surfaces. In summary, the abrasive belt grinding and polishing of the robot becomes one of the popular research directions in the field of grinding and processing of the aero-engine at present, and the defects of the traditional manual grinding and polishing in the aspects of processing precision, product consistency, processing efficiency, processing flexibility of multi-axis numerical control machine tool grinding and polishing and the like are hopefully solved. However, due to the unique functions and operating characteristics of aircraft engine blades, the structure of the aircraft engine blades is generally thin-walled, bent and twisted parts, and the aircraft engine blades have complex shapes, large machining spans and complex stresses. This exacerbates the complexity of the robotic machining and results in less than well controlled surface quality of the machined blade, particularly with respect to the inhibition of grinding burns. Grinding burn can seriously affect the service life of the blade, and the prediction and inhibition of the grinding burn are one of the problems to be solved urgently in the field of grinding of the blade of the aeroengine.
At present, the prediction and inhibition methods of grinding burn mainly comprise a statistical characteristic method, artificial neural network pattern recognition, a support vector machine and other prediction and inhibition methods of grinding burn. And the methods are less applied to the abrasive belt grinding and polishing of the robot for the blade-class complex parts. The statistical characteristic method is to select characteristics which are beneficial to classification from sample characteristics of known categories, and learn statistical characteristic values of different categories in a classification manner so as to realize the prediction of grinding burn, and the method depends on the quality of characteristic selection, is greatly influenced by human subjective factors and has unstable classification effect; the support vector machine SVM is based on a statistical learning theory and a structure risk minimum principle, and the classification capability of the SVM is improved as much as possible on the premise that the recognition error of a training sample is minimum. The SVM is good at solving small samples and is not suitable for large-batch production and processing of large samples such as aviation engine blades; the traditional neural network establishes a mapping relation between input and output through learning input and output samples of a system, has been applied to the field of grinding burn prediction, but has the defects of strong dependence on a processing environment, slow convergence of a learning algorithm, easy falling into local optimum in a training process and the like.
Disclosure of Invention
In view of the above drawbacks and needs of the prior art, the present invention provides a method for predicting grinding burn of an aircraft engine blade. The method takes grinding parameters as input quantity, grinding burn characteristic values as state quantity and grinding burn degree as output quantity, establishes a grinding burn model, and realizes accurate prediction of the grinding burn state; and obtaining the grinding burn degree under different processing parameter combinations according to the grinding burn prediction model to obtain a grinding burn critical domain, comparing the grinding processing parameter with a grinding burn critical threshold, and adjusting the grinding processing parameter, thereby reducing the grinding burn probability of the workpiece.
In order to achieve the above object, according to one aspect of the present invention, there is provided an aircraft engine blade robot grinding burn prediction method, including:
establishing a grinding burn prediction model;
the grinding burn prediction model is configured to take a grinding machining parameter as an input quantity, take a grinding burn characteristic value as a state quantity and take a grinding burn degree as an output quantity;
inputting the grinding parameters of the workpiece into the grinding burn prediction model to obtain the grinding burn degree of the workpiece; and judging the grinding burn degree of the workpiece, comparing the grinding processing parameter of the workpiece with the grinding burn critical threshold value according to the judgment result, and adjusting the grinding processing parameter of the workpiece according to the comparison result.
Further, the establishment of the grinding burn prediction model comprises the following steps:
installing a plurality of sensors for measurement, and establishing a multi-sensor monitoring system for the abrasive belt grinding process of the blade robot;
carrying out blade grinding processing experiments under different processing parameters, detecting the grinding burn degree in each experiment, and acquiring multi-sensor information such as grinding force and acoustic emission signals through a multi-sensor monitoring system;
fusing the multi-sensing information acquired in each experiment by adopting a fusion algorithm to obtain corresponding multi-sensing fusion information; analyzing and processing the multi-sensor fusion information corresponding to each experiment by using a feature extraction algorithm to obtain a grinding burn feature vector of the multi-sensor information, wherein the grinding burn feature vector comprises a plurality of feature values;
extracting a plurality of burn and non-burn samples and corresponding eigenvectors from the experimental result to form a training set of a grinding burn prediction model;
establishing a grinding burn prediction model based on deep learning by taking grinding parameters as input quantities, grinding burn characteristic values as state quantities and grinding burn degrees as output quantities, and training the model by using a training set;
and predicting the grinding burn degree under the condition of the combination of the grinding machining parameters as much as possible by using the trained grinding burn prediction model to obtain the grinding burn critical threshold.
Further, the grinding burn degree of the workpiece is divided into a high grade, a medium grade and a low grade, if the burn degree of the workpiece is judged to be in the high grade or the medium grade, the grinding processing parameter of the workpiece is compared with the grinding burn critical threshold value, the grinding processing parameter of the workpiece is adjusted according to the comparison result, and if the grinding processing parameter of the workpiece is larger than the grinding burn critical threshold value, the grinding processing parameter of the workpiece is modified.
Further, the method comprises the steps of acquiring a grinding force signal, a vibration signal and an acoustic emission signal in the workpiece processing process, and processing the grinding force signal, the vibration signal and the acoustic emission signal to obtain multi-sensing fusion information of the workpiece;
and analyzing and processing the multi-sensor fusion information of the workpiece to obtain the grinding burn characteristic value of the workpiece.
And further, judging the grinding burn degree of the workpiece under different processing parameters according to the grinding burn prediction model to obtain the grinding burn critical parameters.
Further, according to the burn degree judgment result obtained by inputting the grinding parameters of the workpiece into the grinding burn prediction model, the grinding parameters of the workpiece which is just not burnt are obtained, and the grinding parameters are the grinding burn critical threshold.
According to another aspect of the invention, an aircraft engine blade robot grinding burn prediction device is provided, which comprises:
an acquisition unit: acquiring grinding parameters of a workpiece;
a processing unit: inputting the grinding parameters of the workpiece into a grinding burn prediction model to obtain the grinding burn degree of the workpiece; judging the grinding burn degree of the workpiece, and comparing the grinding processing parameter of the workpiece with a grinding burn critical threshold value according to the judgment result;
a determination unit: and determining the optimal grinding parameters of the workpiece according to the comparison result of the grinding parameters of the workpiece and the grinding burn critical threshold value.
Further, the grinding burn prediction model is configured to take the grinding machining parameters as input quantities, the grinding burn characteristic values as state quantities, and the grinding burn degree as output quantities.
Further, if the grinding parameters of the workpiece are larger than the grinding burn critical threshold, the grinding parameters of the workpiece are modified.
Further, a grinding force signal, a vibration signal and an acoustic emission signal in the workpiece processing process are obtained, the grinding force signal, the vibration signal and the acoustic emission signal are processed to obtain multi-sensor fusion information of the workpiece, and the multi-sensor fusion information of the workpiece is analyzed to obtain a grinding burn characteristic value of the workpiece.
In general, compared with the prior art, the above technical solution contemplated by the present invention can achieve the following beneficial effects:
1. according to the method for predicting the grinding burn of the blade robot, the multi-sensor monitoring system for the abrasive belt grinding of the blade robot is formed by the acoustic emission sensor, the force sensor, the acceleration sensor, the temperature sensor, the current and voltage sensor and the like, signals such as acoustic emission, force, vibration and the like can be acquired on line in real time, comprehensive monitoring on the machining of the robot with the complex curved surface is achieved, and therefore the machining quality of the blade robot is effectively guaranteed.
2. The method for predicting the grinding burn of the blade robot has the advantages that the multi-sensor information is efficiently fused, the redundant information in the multi-sensor information can be eliminated, the quality of the information is improved, and the expenditure on data storage is saved.
3. The leaf robot grinding burn prediction method can effectively extract grinding burn invariance characteristics in multi-sensor fusion information, and provides powerful guarantee for accurate identification of a leaf burn prediction model on grinding burn.
4. The method for predicting the grinding burn of the blade robot is based on the blade grinding burn prediction model established by deep learning, has strong real-time performance, can quickly analyze and process large-scale data, has strong generalization capability, improves the robustness of the system, and improves the accuracy and the adaptability of the blade front and rear edge robot grinding burn prediction.
5. According to the blade robot grinding burn prediction method, the grinding burn critical area of the blade is visually displayed in the form of the radar map, the reasonability of grinding machining parameters is conveniently and quickly identified, and the grinding burn phenomenon in the blade machining process is effectively inhibited while the machining efficiency is ensured.
Drawings
FIG. 1 is a schematic diagram of a multi-sensor information fusion and feature extraction structure in an embodiment of the present invention;
FIG. 2 is a schematic flow chart of a method for predicting grinding burn of an aircraft engine blade robot in an embodiment of the invention;
FIG. 3 is a flow chart of a blade grinding burn suppression method in an embodiment of the present invention;
FIG. 4 is a schematic diagram of a grinding burn threshold region in an embodiment of the present invention;
fig. 5 is a schematic structural diagram of a device for predicting grinding burn of an aircraft engine blade robot according to an embodiment of the invention.
Detailed Description
In order to make the objects, technical solutions and advantages of the present invention more apparent, the present invention is described in further detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention. In addition, the technical features involved in the embodiments of the present invention described below may be combined with each other as long as they do not conflict with each other.
Example 1
In order to achieve the above object, embodiment 1 of the present invention provides a method for predicting grinding burn of an aircraft engine blade, including:
establishing a grinding burn prediction model;
the grinding burn prediction model is configured to take a grinding machining parameter as an input quantity, take a grinding burn characteristic value as a state quantity and take a grinding burn degree as an output quantity;
inputting the grinding parameters of the workpiece into a grinding burn prediction model to obtain the grinding burn degree of the workpiece; inputting the grinding parameters of the workpiece into a grinding burn prediction model to obtain the grinding burn degree of the workpiece; and judging the grinding burn degree of the workpiece, comparing the grinding processing parameter of the workpiece with the grinding burn critical threshold value according to the judgment result, and adjusting the grinding processing parameter of the workpiece according to the comparison result.
Further, the burn degree of the workpiece is divided into a high gear, a middle gear and a low gear, the grinding processing parameter of the workpiece is input into the grinding burn prediction model to obtain the grinding burn degree of the workpiece, if the burn degree of the workpiece is judged to be in the high gear or the middle gear, the grinding processing parameter of the workpiece needs to be compared with a grinding burn critical threshold value, and the grinding processing parameter of the workpiece is adjusted according to a comparison result, so that the burn of the engine blade in the processing process can be effectively predicted and restrained.
Further, if the grinding parameters of the workpiece are larger than the grinding burn critical threshold, the grinding parameters of the workpiece are modified.
And further, acquiring a grinding force signal, a vibration signal and an acoustic emission signal in the workpiece processing process, and processing the grinding force signal, the vibration signal and the acoustic emission signal to obtain multi-sensing fusion information of the workpiece.
And further, analyzing and processing the multi-sensor fusion information of the workpiece to obtain a grinding burn characteristic value of the workpiece.
Further, according to the grinding burn prediction model, the grinding burn degree of the workpiece under different processing parameters is judged, and the grinding burn critical parameter is obtained.
And further, obtaining a grinding burn critical threshold value according to the grinding machining parameters of the workpiece and the grinding burn critical parameters.
Example 2
The embodiment 2 of the invention provides a device for predicting grinding burn of an aircraft engine blade, which comprises:
an acquisition unit: acquiring grinding parameters of a workpiece;
a processing unit: inputting the grinding parameters of the workpiece into a grinding burn prediction model to obtain the grinding burn degree of the workpiece; judging the grinding burn degree of the workpiece, and comparing the grinding processing parameter of the workpiece with the grinding burn critical threshold value according to the judgment result;
a determination unit: and determining the grinding parameters of the workpiece according to the comparison result of the grinding parameters of the workpiece and the grinding burn critical threshold value.
Further, the grinding burn prediction model is configured to take the grinding process parameters as input quantities, the grinding burn characteristic values as state quantities, and the grinding burn degree as output quantities.
Further, if the grinding parameters of the workpiece are larger than the grinding burn critical threshold, the grinding parameters of the workpiece are modified.
Further, a grinding force signal, a vibration signal and an acoustic emission signal in the workpiece processing process are obtained, the grinding force signal, the vibration signal and the acoustic emission signal are processed to obtain multi-sensor fusion information of the workpiece, the multi-sensor fusion information of the workpiece is analyzed, and a grinding burn characteristic value of the workpiece is obtained.
Example 3
As shown in fig. 2, a schematic flow chart of a method for predicting grinding burn of an aircraft engine blade robot in an embodiment of the present invention includes the following steps:
carrying out blade grinding burn experiments under different processing parameters (such as abrasive belt linear velocity, robot feeding speed, grinding force and the like), wherein in each experiment, a multi-sensor is used for acquiring signals of grinding force, vibration, acoustic emission and the like in the blade processing process;
further, a specific information fusion algorithm is used for the collected signals to obtain multi-sensing fusion information.
And further, analyzing and processing the multi-sensor fusion information by using a corresponding feature extraction method to obtain the grinding burn characteristic value of the multi-sensor fusion information.
Further, a plurality of burn and non-burn samples are extracted from the grinding experiment result to form a training set of a grinding burn prediction model;
further, a grinding burn prediction model based on deep learning is established by taking the processing parameters as input quantities, the grinding burn characteristic values as state quantities and the grinding burn degree as output quantities, and the model is trained by using a training set.
Further, the blade burn degree under different processing parameters is judged through a grinding burn prediction model, and grinding burn critical parameters are obtained.
And further, drawing a radar map with multiple grinding processing parameters, and drawing a grinding burn critical area on the radar map according to the grinding burn critical parameters.
Fig. 1 shows a structure diagram of multi-sensor information fusion and feature extraction in an embodiment of the present invention, in which a blade robot belt grinding multi-sensor monitoring system is composed of a plurality of sensors, such as a six-dimensional force sensor, an acoustic emission sensor, a temperature sensor, an acceleration sensor, a current and voltage sensor, which are all connected to a computer terminal through a data acquisition card, the acquired information is processed at the computer terminal, and a specific information fusion algorithm is used to fuse the multi-sensor information. And analyzing and processing the fused information by adopting a combined characteristic analysis method combining wavelet transform and Hilbert-Huang transform to obtain a grinding burn characteristic vector (comprising a plurality of characteristic values).
As shown in fig. 3, which is a flowchart of a method for suppressing a blade grinding burn in an embodiment of the present invention, the steps of specifically implementing the blade grinding burn suppression include the following steps:
the blade grinding parameters to be used (such as belt line speed, robot feed speed and grinding force, etc.) are input.
Further, judging whether the blade grinding parameters to be used are in a critical area or not according to the grinding burn critical area; if the blade grinding parameters are in the critical area, modifying the processing parameters; if not, the grinding process can be continued.
Fig. 4 shows a radar chart with multiple grinding parameters in an embodiment of the present invention, wherein the area enclosed by the broken line on the chart is a grinding burn critical area. The processing parameter combination located in the critical area does not cause the blade processing process to generate the grinding burn phenomenon, and the processing parameter combination on the boundary represents the grinding burn critical parameter.
Example 4
The embodiment 4 of the invention provides a method for predicting grinding burn of an aircraft engine blade, which comprises the following steps:
installing a plurality of types of sensors for measurement, and establishing a multi-sensor monitoring system for the abrasive belt grinding process of the blade robot;
further, blade grinding processing experiments under different processing parameters are carried out, the grinding burn degree needs to be detected in each experiment, and multiple sensing information such as grinding force, acoustic emission signals and the like is collected through a multi-sensor monitoring system;
further, fusing the multi-sensing information acquired in each experiment by using a specific fusion algorithm to obtain corresponding multi-sensing fusion information; analyzing and processing the multi-sensor fusion information corresponding to each experiment by using a specific feature extraction algorithm to obtain a grinding burn feature vector of the multi-sensor information, wherein the grinding burn feature vector comprises a plurality of feature values;
further, extracting a plurality of burn and non-burn samples and corresponding characteristic vectors from experimental results to form a training set of a grinding burn prediction model;
further, establishing a grinding burn prediction model based on deep learning by taking grinding parameters as input quantities, grinding burn characteristic values as state quantities and grinding burn degrees as output quantities, and training the model by using a training set;
further, predicting the grinding burn degree under the condition of the grinding machining parameter combination as much as possible by using the trained grinding burn prediction model to obtain a grinding burn critical threshold;
further, drawing a radar map with multiple grinding processing parameters, and drawing a grinding burn critical area on the radar map according to the grinding burn critical threshold;
further, according to the grinding burn critical area, the grinding burn in the actual blade machining process is reduced;
preferably, the blade robot belt sanding process multi-sensor monitoring system includes, but is not limited to, the following sensors: the device comprises a six-dimensional force sensor, an acoustic emission sensor, a temperature sensor, an acceleration sensor and a current and voltage sensor.
Preferably, the specific fusion algorithm includes a mainstream information fusion algorithm, such as: kernel principal component analysis.
Preferably, the specific feature extraction algorithm adopts a combined feature analysis method combining wavelet transform and Hilbert-Huang transform.
Preferably, the specific step of suppressing the grinding burn in the actual blade machining process is to judge whether the currently used grinding parameter is in a critical area, if not, the machining parameter needs to be modified to avoid the grinding burn, otherwise, the machining can be continued.
1. According to the blade robot grinding burn prediction method provided by the embodiment of the invention, the abrasive belt grinding multi-sensor monitoring system of the blade robot is formed by the acoustic emission sensor, the force sensor, the acceleration sensor, the temperature sensor, the current and voltage sensor and other multi-sensors, signals of acoustic emission, force, vibration, temperature and the like can be acquired on line in real time, and comprehensive monitoring on the complex curved surface robot processing is realized, so that the processing quality of the blade robot is effectively ensured.
2. According to the blade robot grinding burn prediction method provided by the embodiment of the invention, the information fusion algorithm can be used for efficiently fusing multi-sensor information, redundant information in the multi-sensor information can be eliminated, the information quality is improved, and the expenditure on data storage is saved.
3. According to the leaf robot grinding burn prediction method provided by the embodiment of the invention, the characteristic extraction algorithm is adopted, grinding burn invariance characteristics in multi-sensor fusion information can be effectively extracted, and powerful guarantee is provided for accurate identification of a leaf burn prediction model on grinding burn.
4. The embodiment of the invention provides a blade robot grinding burn prediction method, a blade grinding burn prediction model established based on deep learning has strong real-time performance, large-scale data can be rapidly analyzed and processed, meanwhile, the generalization capability of the model is strong, the robustness of the system is improved, and the accuracy and the adaptability of the blade front and rear edge robot grinding burn prediction are improved.
5. According to the method for inhibiting the grinding burn of the blade robot provided by the embodiment of the invention, the grinding burn critical domain of the blade is visually displayed in the form of the radar map, so that the reasonability of the grinding machining parameter is conveniently and quickly identified, and the grinding burn phenomenon in the blade machining process is effectively inhibited while the machining efficiency is ensured.
6. The method for predicting the grinding burn of the blade robot provided by the embodiment of the invention has the advantages of low application cost, high automation degree, strong expandability and the like, and can add or delete the input of a sensor and a model according to the actual processing requirement, thereby enhancing the flexibility of an abrasive belt grinding processing system of the blade robot.
It will be understood by those skilled in the art that the foregoing is only a preferred embodiment of the present invention, and is not intended to limit the invention, and that any modification, equivalent replacement, or improvement made within the spirit and principle of the present invention should be included in the scope of the present invention.
Claims (9)
1. A method for predicting grinding burn of an aircraft engine blade robot is characterized by comprising the following steps:
establishing a grinding burn prediction model, comprising the following steps:
installing a plurality of sensors for measurement, and establishing a multi-sensor monitoring system for the abrasive belt grinding process of the blade robot;
carrying out blade grinding processing experiments under different processing parameters, detecting the grinding burn degree in each experiment, and acquiring grinding force and acoustic emission signal multi-sensing information through a multi-sensor monitoring system;
fusing the multi-sensing information acquired in each experiment by adopting a fusion algorithm to obtain corresponding multi-sensing fusion information; analyzing and processing the multi-sensor fusion information corresponding to each experiment by using a feature extraction algorithm to obtain a grinding burn feature vector of the multi-sensor information, wherein the grinding burn feature vector comprises a plurality of feature values;
extracting a plurality of burn and non-burn samples and corresponding eigenvectors from the experimental result to form a training set of a grinding burn prediction model;
establishing a grinding burn prediction model based on deep learning by taking grinding parameters as input quantities, grinding burn characteristic values as state quantities and grinding burn degrees as output quantities, and training the model by using a training set;
predicting the grinding burn degree under the condition of the grinding machining parameter combination as much as possible by using the trained grinding burn prediction model to obtain a grinding burn critical threshold;
the grinding burn prediction model is configured to take a grinding machining parameter as an input quantity, take a grinding burn characteristic value as a state quantity and take a grinding burn degree as an output quantity;
inputting the grinding parameters of the workpiece into the grinding burn prediction model to obtain the grinding burn degree of the workpiece; and judging the grinding burn degree of the workpiece, comparing the grinding processing parameter of the workpiece with the grinding burn critical threshold value according to the judgment result, and adjusting the grinding processing parameter of the workpiece according to the comparison result.
2. The aircraft engine blade robot grinding burn prediction method of claim 1, wherein: the grinding burn degree of the workpiece is divided into a high grade, a medium grade and a low grade, if the burn degree of the workpiece is judged to be in the high grade or the medium grade, the grinding machining parameter of the workpiece is compared with a grinding burn critical threshold value, the grinding machining parameter of the workpiece is adjusted according to a comparison result, and if the grinding machining parameter of the workpiece is larger than the grinding burn critical threshold value, the grinding machining parameter of the workpiece is modified.
3. The aircraft engine blade robot grinding burn prediction method of claim 1, wherein: the method further comprises the steps of acquiring a grinding force signal, a vibration signal and an acoustic emission signal in the workpiece processing process, and processing the grinding force signal, the vibration signal and the acoustic emission signal to obtain multi-sensor fusion information of the workpiece;
and analyzing and processing the multi-sensor fusion information of the workpiece to obtain the grinding burn characteristic value of the workpiece.
4. The aircraft engine blade robot grinding burn prediction method of claim 3, wherein: and judging the grinding burn degree of the workpiece under different processing parameters according to the grinding burn prediction model to obtain the grinding burn critical parameters.
5. The aircraft engine blade robot grinding burn prediction method of claim 4, wherein: and inputting the burn degree judgment result obtained by the grinding burn prediction model according to the grinding parameters of the workpiece to obtain the grinding parameters of the workpiece which is just not burnt, namely the grinding burn critical threshold.
6. An apparatus for implementing the aircraft engine blade robot grinding burn prediction method according to any one of claims 1 to 5, characterized by comprising:
an acquisition unit: acquiring grinding parameters of a workpiece;
a processing unit: inputting the grinding parameters of the workpiece into a grinding burn prediction model to obtain the grinding burn degree of the workpiece; judging the grinding burn degree of the workpiece, and comparing the grinding processing parameter of the workpiece with a grinding burn critical threshold value according to the judgment result;
a determination unit: and determining the optimal grinding parameters of the workpiece according to the comparison result of the grinding parameters of the workpiece and the grinding burn critical threshold value.
7. An apparatus according to claim 6, characterized in that: the grinding burn prediction model is configured to take a grinding machining parameter as an input quantity, a grinding burn characteristic value as a state quantity and a grinding burn degree as an output quantity.
8. An apparatus according to claim 7, characterized in that: and if the grinding parameters of the workpiece are larger than the grinding burn critical threshold, modifying the grinding parameters of the workpiece.
9. An apparatus according to claim 8, characterized in that: the method comprises the steps of obtaining a grinding force signal, a vibration signal and an acoustic emission signal in the process of processing a workpiece, processing the grinding force signal, the vibration signal and the acoustic emission signal to obtain multi-sensor fusion information of the workpiece, and analyzing the multi-sensor fusion information of the workpiece to obtain a grinding burn characteristic value of the workpiece.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010034441.9A CN111230740B (en) | 2020-01-14 | 2020-01-14 | Method and device for predicting grinding burn of aero-engine blade robot |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010034441.9A CN111230740B (en) | 2020-01-14 | 2020-01-14 | Method and device for predicting grinding burn of aero-engine blade robot |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN111230740A CN111230740A (en) | 2020-06-05 |
| CN111230740B true CN111230740B (en) | 2021-05-25 |
Family
ID=70877706
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202010034441.9A Active CN111230740B (en) | 2020-01-14 | 2020-01-14 | Method and device for predicting grinding burn of aero-engine blade robot |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN111230740B (en) |
Families Citing this family (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112946072A (en) * | 2021-01-27 | 2021-06-11 | 重庆大学 | Abrasive belt wear state monitoring method based on machine learning |
| CN115383514A (en) * | 2022-09-06 | 2022-11-25 | 陕西法士特齿轮有限责任公司 | Grinding burn monitoring method based on vibration signal time domain analysis |
| CN115647819B (en) * | 2022-09-20 | 2023-06-16 | 玉环仪表机床制造厂 | Turning and grinding integrated compound machine and control method thereof |
| CN116652823B (en) * | 2023-06-26 | 2024-03-22 | 浙江钱祥工具股份有限公司 | Automatic monitoring system and method for grinding machine |
Family Cites Families (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN85108547A (en) * | 1985-11-04 | 1987-05-20 | 中国纺织大学 | Grinding burning adaptive control system |
| US8353739B2 (en) * | 2009-12-08 | 2013-01-15 | Allison Transmission, Inc. | Method for detecting and/or preventing grind burn |
| CN106407683B (en) * | 2016-09-19 | 2019-01-15 | 上海理工大学 | Crush grinding process parameter optimizing method based on grinding removal rate model |
| CN107756250B (en) * | 2017-11-08 | 2019-05-24 | 山东理工大学 | A kind of grinding power and energy consumption intelligent monitor system and decision-making technique |
-
2020
- 2020-01-14 CN CN202010034441.9A patent/CN111230740B/en active Active
Also Published As
| Publication number | Publication date |
|---|---|
| CN111230740A (en) | 2020-06-05 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111230740B (en) | Method and device for predicting grinding burn of aero-engine blade robot | |
| CN110059442B (en) | Turning tool changing method based on part surface roughness and power information | |
| CN110153801A (en) | A kind of cutting-tool wear state discrimination method based on multi-feature fusion | |
| CN111633467B (en) | Cutter wear state monitoring method based on one-dimensional depth convolution automatic encoder | |
| CN114619292B (en) | Milling cutter wear monitoring method based on fusion of wavelet denoising and attention mechanism with GRU network | |
| CN111660141B (en) | Milling cutter wear state identification method based on spindle driving current and irrelevant to working conditions | |
| CN107891199B (en) | Spark discharge condition checkout gear and recognition methods | |
| CN103465107A (en) | Tool wear monitoring method | |
| Wang et al. | Multi-condition identification in milling Ti-6Al-4V thin-walled parts based on sensor fusion | |
| CN110434676A (en) | A kind of boring monitoring chatter method of multisensor time-frequency characteristics fusion | |
| CN113997122B (en) | Cutter wear monitoring method and system | |
| Oberle et al. | A use case to implement machine learning for life time prediction of manufacturing tools | |
| CN116871978A (en) | Drilling tool state monitoring method based on multi-sensor fusion | |
| Xu et al. | Intelligent recognition method of turning tool wear state based on information fusion technology and BP neural network | |
| Wang et al. | Artificial Intelligence for advanced non-conventional machining processes | |
| CN112207631A (en) | Method for generating tool detection model, method, system, device and medium for detecting tool detection model | |
| CN111331211B (en) | On-line penetration detection method and system for electric spark small hole machining | |
| Cheng et al. | Tool wear intelligent monitoring techniques in cutting: a review | |
| CN114227382A (en) | Cutter damage monitoring system and method based on novel capsule network | |
| CN112946072A (en) | Abrasive belt wear state monitoring method based on machine learning | |
| CN110744359A (en) | Numerical control lathe cutter wear monitoring system and method | |
| CN116628467A (en) | Turbine disc mortise broach wear identification method based on combination of space and time sequence | |
| CN112783138B (en) | Intelligent monitoring and abnormity diagnosis method and device for processing stability of production line equipment | |
| CN112884027B (en) | Cutting process real-time state monitoring method and device based on pattern recognition | |
| CN210878880U (en) | Numerical control lathe cutter wearing and tearing monitoring system |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |