Full-automatic detection method applied to flywheel machining

Abstract

The invention relates to a full-automatic detection method applied to flywheel processing, and the method comprises the steps: Q1, placing a processed flywheel on a detection production line through a manipulator, obtaining the image data information of the flywheel in real time based on an online camera, obtaining the point cloud data information of the flywheel in real time based on an online laser radar sensor, and carrying out the detection of the flywheel; distance data information of the flywheel is obtained in real time based on an on-line laser displacement sensor; and Q2, on the basis of the image data information of the flywheel and the point cloud data information of the flywheel, constructing a first detection parameter function F1 of flywheel processing, representing detection parameters of flywheel processing, and obtaining data information of a first detection parameter matrix of flywheel processing. According to the flywheel detection device, flywheel machining can be rapidly and comprehensively detected, it is guaranteed that the quality of flywheel products meets the requirement, the multiple sensors are adopted for detecting the flywheel, the flywheel detection precision is improved, and the defective rate is reduced.

Description

Technical Field

[0001] The invention relates to the technical field of flywheel processing detection, and in particular to a full-automatic detection method applied to flywheel processing. Background Art

[0002] With the improvement of the automation level of automobile flywheel processing, automobile flywheel production lines have begun to tend to be personalized. The detection process composed of instruments for single detection data requires a lot of manpower and material resources, is not suitable for rapid detection, and has low detection accuracy. At the same time, automobile flywheels, as one of the important components of automobile engines, play a vital role. As an energy storage device, when the cylinder work is discontinuous during the operation of the engine, resulting in fluctuations in the crankshaft speed, the flywheel can store excess kinetic energy and release it when the speed drops, effectively smoothing the operation of the engine and improving its stability and stability. If defects occur, it will affect the overall operation of the engine. Therefore, how to accurately detect flywheel processing has become a problem that we urgently need to solve. Summary of the invention

[0003] In view of the above problems, the present invention provides a fully automatic detection method applied to flywheel processing, which can not only quickly and comprehensively detect the flywheel processing to ensure that the quality of the flywheel product meets the requirements, but also use multiple sensors to detect the flywheel, thereby improving the accuracy of flywheel detection and reducing the occurrence of defective rates.

[0004] In order to achieve the above-mentioned purpose and other related purposes, the technical solution provided by the present invention is as follows:

[0005] A fully automatic detection method for flywheel processing, the method comprising:

[0006] Q1. The robot places the processed flywheel on the inspection production line, obtains the image data information of the flywheel in real time based on the online camera, obtains the point cloud data information of the flywheel in real time based on the online laser radar sensor, and obtains the distance data information of the flywheel in real time based on the online laser displacement sensor;

[0007] Q2. Based on the image data information of the flywheel and the point cloud data information of the flywheel, a first detection parameter function F1 of flywheel processing is constructed, the detection parameters of flywheel processing are characterized, and data information of the first detection parameter matrix of flywheel processing is obtained;

[0008] Q3. Based on the image data information of the flywheel and the distance data information of the flywheel, a second detection parameter function F2 of flywheel processing is constructed to characterize the detection parameters of flywheel processing to obtain data information of the second detection parameter matrix of flywheel processing;

[0009] Q4. Based on the point cloud data information of the flywheel and the distance data information of the flywheel, a third detection parameter function F3 of flywheel processing is constructed, the detection parameters of flywheel processing are characterized, and data information of the third detection parameter matrix of flywheel processing is obtained;

[0010] Q5. Based on the data information of the third detection parameter matrix, the data information of the second detection parameter matrix and the data information of the first detection parameter matrix of the flywheel processing, the processing quality of the flywheel is predicted by using an improved BP neural network regression prediction algorithm to obtain the predicted processing quality data information of the flywheel;

[0011] Q6. Based on the predicted flywheel processing quality data information, construct a flywheel processing quality evaluation model, evaluate the flywheel processing quality, and output the flywheel processing quality evaluation data information.

[0012] Further, in step Q2, the first detection parameter function F1 of the flywheel processing is,

[0013]

[0014] Among them, x1 is the image data information of the flywheel, x2 is the point cloud data information of the flywheel, f1 is the relationship function between the image and the point cloud of the flywheel, and α1 and α2 are the first detection factors of flywheel processing.

[0015] Furthermore, the relationship function f1 between the image of the flywheel and the point cloud is,

[0016]

[0017] Among them, x1 is the image data information of the flywheel, and x2 is the point cloud data information of the flywheel.

[0018] Further, in step Q3, the second detection parameter function F2 of the flywheel processing is,

[0019]

[0020] Among them, x1 is the image data information of the flywheel, x3 is the distance data information of the flywheel, f2 is the relationship function between the image and the distance of the flywheel, and β1 and β2 are the second detection factors of the flywheel processing.

[0021] Further, in step Q4, the third detection parameter function F3 of the flywheel processing is,

[0022]

[0023]

[0024] Among them, x2 is the point cloud data information of the flywheel, x3 is the distance data information of the flywheel, f3 is the relationship function between the point cloud and the distance of the flywheel, and λ1 and λ2 are the third detection factors of flywheel processing.

[0025] Furthermore, the relationship function f3 between the point cloud and the distance of the flywheel is:

[0026]

[0027] Among them, x2 is the point cloud data information of the flywheel, and x3 is the distance data information of the flywheel.

[0028] Furthermore, in step Q5, the use of the improved BP neural network regression prediction algorithm to predict the processing quality of the flywheel includes:

[0029] Q51. Based on the data information of the third detection parameter matrix of the flywheel processing, the data information of the second detection parameter matrix and the data information of the first detection parameter matrix, establish the detection parameter fusion function G of the flywheel processing,

[0030]

[0031] Among them, a is the data information of the third detection parameter matrix of flywheel processing, b is the data information of the second detection parameter matrix of flywheel processing, c is the data information of the first detection parameter matrix of flywheel processing, ω1, ω2 and ω3 are the data fusion factors of flywheel processing, and the detection parameter matrix of flywheel processing is fused to obtain the data information of the fused detection parameter matrix of flywheel processing;

[0032] Q52. Input the data information of the fusion detection parameter matrix of the flywheel processing into the improved BP neural network regression prediction model for training and learning, and determine the improved neuron kernel function R,

[0033]

[0034] Among them, y is the data information of the fusion detection parameter matrix of flywheel processing, ρ1, ρ2 and ρ3 are the learning factors of flywheel processing detection, and the trained improved BP neural network regression prediction model is obtained;

[0035] Q53. Based on the trained improved BP neural network regression prediction model, the data information of the fusion detection parameter matrix of the flywheel processing is input to predict the processing quality of the flywheel, and obtain the predicted flywheel processing quality data information.

[0036] Furthermore, the learning factor ρ1 of the flywheel processing detection is,

[0037]

[0038] The learning factor ρ2 of the flywheel processing detection is,

[0039]

[0040] The learning factor ρ3 of the flywheel processing detection is,

[0041]

[0042] Among them, y is the data information of the fusion detection parameter matrix of flywheel processing.

[0043] Furthermore, in step Q6, the building of the flywheel processing quality evaluation model and the evaluation of the flywheel processing quality include:

[0044] Q61. Obtain the predicted flywheel processing quality data information, construct a flywheel processing quality data set, and divide it into a flywheel processing training data set and a flywheel processing test data set;

[0045] Q62. Input the flywheel processing training data set into the flywheel processing quality evaluation model for training and learning, and determine the evaluation function W of the flywheel processing quality,

[0046]

[0047] Among them, g is the flywheel processing training data set, η1 and η2 are the evaluation factors of flywheel processing quality, and the evaluation model of flywheel processing quality after training is obtained;

[0048] Q63. Based on the trained flywheel processing quality evaluation model, the flywheel processing test data set is input for optimization to obtain a trained flywheel processing quality evaluation model;

[0049] Q64. Based on the trained flywheel processing quality evaluation model, the predicted flywheel processing quality data information is input, the flywheel processing quality is evaluated, and the flywheel processing quality evaluation data information is output.

[0050] Furthermore, the method further comprises:

[0051] Q7. Based on the evaluation data information of the processing quality of the flywheel, a preset threshold is set. If the evaluation data of the processing quality of the flywheel is less than the preset threshold, the quality requirements are not met and the flywheel is unqualified. If the evaluation data of the processing quality of the flywheel is greater than the preset threshold, the quality requirements are met and the flywheel is qualified.

[0052] The present invention has the following positive effects:

[0053] 1. The present invention constructs a first detection parameter function F1, a second detection parameter function F2 and a third detection parameter function F3 of flywheel processing to characterize the overall detection parameters of flywheel processing from all aspects. It can not only quickly and comprehensively detect the flywheel processing to ensure that the quality of the flywheel product meets the requirements, but also use multiple sensors to detect the flywheel, thereby improving the accuracy of flywheel detection and reducing the occurrence of defective rates.

[0054] 2. The present invention predicts the flywheel processing quality by adopting an improved BP neural network regression prediction algorithm, and evaluates the flywheel processing quality in combination with the flywheel processing quality evaluation model. It can not only continuously change according to the demand for flywheel processing quality and improve the detection accuracy of flywheel detection, but also during the flywheel detection process, no human participation is required, thereby reducing labor costs and improving the efficiency of flywheel processing. BRIEF DESCRIPTION OF THE DRAWINGS

[0055] Figure 1 It is a schematic diagram of the method flow of the present invention;

[0056] Figure 2 It is a schematic diagram of the flow chart of the improved BP neural network regression prediction algorithm of the present invention;

[0057] Figure 3 A schematic diagram of a flow chart of constructing an evaluation model for flywheel machining quality according to the present invention;

[0058] Figure 4 It is a structural schematic diagram of the present invention.

[0059] Explanation of the numbers in the figure: 1-manipulator, 2-detection production line, 3-camera, 4-lidar sensor, 5-laser displacement sensor, 6-flywheel. DETAILED DESCRIPTION

[0060] The following is a description of exemplary embodiments of the present disclosure in conjunction with the accompanying drawings, including various details of the embodiments of the present disclosure to facilitate understanding, which should be considered as merely exemplary. Therefore, it should be recognized by those of ordinary skill in the art that various changes and modifications may be made to the embodiments described herein without departing from the scope and spirit of the present disclosure. Similarly, for the sake of clarity and conciseness, descriptions of well-known functions and structures are omitted in the following description.

[0061] Example 1: Figure 1 or Figure 4 As shown, a fully automatic detection method applied to flywheel processing, the method comprising:

[0062] Q1. The robot 1 places the processed flywheel 6 on the inspection production line 2, obtains the image data information of the flywheel in real time based on the online camera 3, obtains the point cloud data information of the flywheel in real time based on the online laser radar sensor 4, and obtains the distance data information of the flywheel in real time based on the online laser displacement sensor 5;

[0063] Q2. Based on the image data information of the flywheel and the point cloud data information of the flywheel, a first detection parameter function F1 of flywheel processing is constructed, the detection parameters of flywheel processing are characterized, and data information of the first detection parameter matrix of flywheel processing is obtained;

[0064] Q3. Based on the image data information of the flywheel and the distance data information of the flywheel, a second detection parameter function F2 of flywheel processing is constructed to characterize the detection parameters of flywheel processing to obtain data information of the second detection parameter matrix of flywheel processing;

[0065] Q4. Based on the point cloud data information of the flywheel and the distance data information of the flywheel, a third detection parameter function F3 of flywheel processing is constructed, the detection parameters of flywheel processing are characterized, and data information of the third detection parameter matrix of flywheel processing is obtained;

[0066] Q5. Based on the data information of the third detection parameter matrix, the data information of the second detection parameter matrix and the data information of the first detection parameter matrix of the flywheel processing, the processing quality of the flywheel is predicted by using an improved BP neural network regression prediction algorithm to obtain the predicted processing quality data information of the flywheel;

[0067] Q6. Based on the predicted flywheel processing quality data information, construct a flywheel processing quality evaluation model, evaluate the flywheel processing quality, and output the flywheel processing quality evaluation data information.

[0068] In this embodiment, in step Q2, the first detection parameter function F1 of the flywheel machining is,

[0069]

[0070] Among them, x1 is the image data information of the flywheel, x2 is the point cloud data information of the flywheel, f1 is the relationship function between the image and the point cloud of the flywheel, and α1 and α2 are the first detection factors of flywheel processing.

[0071] In this embodiment, the relationship function f1 between the image of the flywheel and the point cloud is:

[0072]

[0073] Among them, x1 is the image data information of the flywheel, and x2 is the point cloud data information of the flywheel.

[0074] In this embodiment, in step Q3, the second detection parameter function F2 of the flywheel machining is,

[0075]

[0076] Among them, x1 is the image data information of the flywheel, x3 is the distance data information of the flywheel, f2 is the relationship function between the image and the distance of the flywheel, and β1 and β2 are the second detection factors of the flywheel processing.

[0077] In this embodiment, in step Q4, the third detection parameter function F3 of the flywheel processing is,

[0078]

[0079] Among them, x2 is the point cloud data information of the flywheel, x3 is the distance data information of the flywheel, f3 is the relationship function between the point cloud and the distance of the flywheel, and λ1 and λ2 are the third detection factors of flywheel processing.

[0080] In this embodiment, the relationship function f3 between the point cloud and the distance of the flywheel is:

[0081]

[0082] Among them, x2 is the point cloud data information of the flywheel, and x3 is the distance data information of the flywheel.

[0083] Embodiment 2: Based on the fully automatic detection method applied to flywheel processing in Embodiment 1, the present invention is further illustrated and described below.

[0084] like Figure 1 or Figure 4 As shown, a fully automatic detection method applied to flywheel processing, the method comprising:

[0085] Q1. The robot 1 places the processed flywheel 6 on the inspection production line 2, obtains the image data information of the flywheel in real time based on the online camera 3, obtains the point cloud data information of the flywheel in real time based on the online laser radar sensor 4, and obtains the distance data information of the flywheel in real time based on the online laser displacement sensor 5;

[0086] Q2. Based on the image data information of the flywheel and the point cloud data information of the flywheel, a first detection parameter function F1 of flywheel processing is constructed, the detection parameters of flywheel processing are characterized, and data information of the first detection parameter matrix of flywheel processing is obtained;

[0087] Q3. Based on the image data information of the flywheel and the distance data information of the flywheel, a second detection parameter function F2 of flywheel processing is constructed to characterize the detection parameters of flywheel processing to obtain data information of the second detection parameter matrix of flywheel processing;

[0088] Q4. Based on the point cloud data information of the flywheel and the distance data information of the flywheel, a third detection parameter function F3 of flywheel processing is constructed, the detection parameters of flywheel processing are characterized, and data information of the third detection parameter matrix of flywheel processing is obtained;

[0089] Q5. Based on the data information of the third detection parameter matrix, the data information of the second detection parameter matrix and the data information of the first detection parameter matrix of the flywheel processing, the processing quality of the flywheel is predicted by using an improved BP neural network regression prediction algorithm to obtain the predicted processing quality data information of the flywheel;

[0090] Q6. Based on the predicted flywheel processing quality data information, construct a flywheel processing quality evaluation model, evaluate the flywheel processing quality, and output the flywheel processing quality evaluation data information.

[0091] In this embodiment, if Figure 2 As shown, in step Q5, the use of the improved BP neural network regression prediction algorithm to predict the processing quality of the flywheel includes:

[0092] Q51. Based on the data information of the third detection parameter matrix of the flywheel processing, the data information of the second detection parameter matrix and the data information of the first detection parameter matrix, establish the detection parameter fusion function G of the flywheel processing,

[0093]

[0094] Among them, a is the data information of the third detection parameter matrix of flywheel processing, b is the data information of the second detection parameter matrix of flywheel processing, c is the data information of the first detection parameter matrix of flywheel processing, ω1, ω2 and ω3 are the data fusion factors of flywheel processing, and the detection parameter matrix of flywheel processing is fused to obtain the data information of the fused detection parameter matrix of flywheel processing;

[0095] Q52. Input the data information of the fusion detection parameter matrix of the flywheel processing into the improved BP neural network regression prediction model for training and learning, and determine the improved neuron kernel function R,

[0096]

[0097] Among them, y is the data information of the fusion detection parameter matrix of flywheel processing, ρ1, ρ2 and ρ3 are the learning factors of flywheel processing detection, and the trained improved BP neural network regression prediction model is obtained;

[0098] Q53. Based on the trained improved BP neural network regression prediction model, the data information of the fusion detection parameter matrix of the flywheel processing is input to predict the processing quality of the flywheel, and obtain the predicted flywheel processing quality data information.

[0099] In this embodiment, the learning factor ρ1 of the flywheel processing detection is,

[0100]

[0101] The learning factor ρ2 of the flywheel processing detection is,

[0102]

[0103] The learning factor ρ3 of the flywheel processing detection is,

[0104]

[0105] Among them, y is the data information of the fusion detection parameter matrix of flywheel processing.

[0106] In this embodiment, if Figure 3 As shown, in step Q6, the flywheel processing quality evaluation model is constructed to evaluate the flywheel processing quality, including:

[0107] Q61. Obtain the predicted flywheel processing quality data information, construct a flywheel processing quality data set, and divide it into a flywheel processing training data set and a flywheel processing test data set;

[0108] Q62. Input the flywheel processing training data set into the flywheel processing quality evaluation model for training and learning, and determine the evaluation function W of the flywheel processing quality,

[0109]

[0110] Among them, g is the flywheel processing training data set, η1 and η2 are the evaluation factors of flywheel processing quality, and the evaluation model of flywheel processing quality after training is obtained;

[0111] Q63. Based on the trained flywheel processing quality evaluation model, the flywheel processing test data set is input for optimization to obtain a trained flywheel processing quality evaluation model;

[0112] Q64. Based on the trained flywheel processing quality evaluation model, the predicted flywheel processing quality data information is input, the flywheel processing quality is evaluated, and the flywheel processing quality evaluation data information is output.

[0113] In this embodiment, the method further includes:

[0114] Q7. Based on the evaluation data information of the processing quality of the flywheel, a preset threshold is set. If the evaluation data of the processing quality of the flywheel is less than the preset threshold, the quality requirements are not met and the flywheel is unqualified. If the evaluation data of the processing quality of the flywheel is greater than the preset threshold, the quality requirements are met and the flywheel is qualified.

[0115] In this embodiment, the present invention provides a computer-readable storage medium, on which is stored a computer program programmed or configured to execute any one of the fully automatic detection methods for flywheel processing.

[0116] Any reference to memory, storage, database or other media used in the embodiments provided in the present application may include non-volatile and / or volatile memory. Non-volatile memory may include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM) or flash memory. Volatile memory may include random access memory (RAM) or external cache memory. As an illustration and not limitation, RAM is available in various forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link (Synchlink) DRAM (SLDRAM), memory bus (Rambus) direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM) and memory bus dynamic RAM (RDRAM).

[0117] In summary, the present invention can not only quickly and comprehensively detect flywheel processing to ensure that the quality of the flywheel product meets the requirements, but also uses multiple sensors to detect the flywheel, thereby improving the accuracy of flywheel detection and reducing the occurrence of defective rates.

[0118] The above specific implementations do not constitute a limitation on the protection scope of the present disclosure. It should be understood by those skilled in the art that various modifications, combinations, sub-combinations and substitutions can be made according to design requirements and other factors. Any modification, equivalent substitution and improvement made within the spirit and principle of the present disclosure shall be included in the protection scope of the present disclosure.

Claims

1. A fully automatic detection method for flywheel processing, characterized in that: The method comprises: Q1. The robot places the processed flywheel on the inspection production line, obtains the image data information of the flywheel in real time based on the online camera, obtains the point cloud data information of the flywheel in real time based on the online laser radar sensor, and obtains the distance data information of the flywheel in real time based on the online laser displacement sensor; Q2. Based on the image data information of the flywheel and the point cloud data information of the flywheel, a first detection parameter function F1 of flywheel processing is constructed, the detection parameters of flywheel processing are characterized, and data information of the first detection parameter matrix of flywheel processing is obtained; Q3. Based on the image data information of the flywheel and the distance data information of the flywheel, a second detection parameter function F2 of flywheel processing is constructed to characterize the detection parameters of flywheel processing to obtain data information of the second detection parameter matrix of flywheel processing; Q4. Based on the point cloud data information of the flywheel and the distance data information of the flywheel, a third detection parameter function F3 of flywheel processing is constructed, the detection parameters of flywheel processing are characterized, and data information of the third detection parameter matrix of flywheel processing is obtained; Q5. Based on the data information of the third detection parameter matrix, the data information of the second detection parameter matrix and the data information of the first detection parameter matrix of the flywheel processing, the processing quality of the flywheel is predicted by using an improved BP neural network regression prediction algorithm to obtain the predicted processing quality data information of the flywheel; Q6. Based on the predicted flywheel processing quality data information, construct a flywheel processing quality evaluation model, evaluate the flywheel processing quality, and output the flywheel processing quality evaluation data information.

2. The fully automatic detection method for flywheel processing according to claim 1 is characterized in that: In step Q2, the first detection parameter function F1 of the flywheel machining is, Among them, x1 is the image data information of the flywheel, x2 is the point cloud data information of the flywheel, f1 is the relationship function between the image and the point cloud of the flywheel, and α1 and α2 are the first detection factors of flywheel processing.

3. The fully automatic detection method for flywheel processing according to claim 2 is characterized in that: The relationship function f1 between the image of the flywheel and the point cloud is, Among them, x1 is the image data information of the flywheel, and x2 is the point cloud data information of the flywheel.

4. The fully automatic detection method for flywheel processing according to claim 1 is characterized in that: In step Q3, the second detection parameter function F2 of the flywheel machining is, Among them, x1 is the image data information of the flywheel, x3 is the distance data information of the flywheel, f2 is the relationship function between the image and the distance of the flywheel, and β1 and β2 are the second detection factors of the flywheel processing.

5. The fully automatic detection method for flywheel processing according to claim 1 is characterized in that: In step Q4, the third detection parameter function F3 of the flywheel machining is, Among them, x2 is the point cloud data information of the flywheel, x3 is the distance data information of the flywheel, f3 is the relationship function between the point cloud and the distance of the flywheel, and λ1 and λ2 are the third detection factors of flywheel processing.

6. The fully automatic detection method for flywheel processing according to claim 5 is characterized in that: The relationship function f3 between the point cloud and distance of the flywheel is: Among them, x2 is the point cloud data information of the flywheel, and x3 is the distance data information of the flywheel.

7. The fully automatic detection method for flywheel processing according to claim 1 is characterized in that: In step Q5, the use of the improved BP neural network regression prediction algorithm to predict the processing quality of the flywheel includes: Q51. Based on the data information of the third detection parameter matrix of the flywheel processing, the data information of the second detection parameter matrix and the data information of the first detection parameter matrix, establish the detection parameter fusion function G of the flywheel processing, Among them, a is the data information of the third detection parameter matrix of flywheel processing, b is the data information of the second detection parameter matrix of flywheel processing, c is the data information of the first detection parameter matrix of flywheel processing, ω1, ω2 and ω3 are the data fusion factors of flywheel processing, and the detection parameter matrix of flywheel processing is fused to obtain the data information of the fused detection parameter matrix of flywheel processing; Q52. Input the data information of the fusion detection parameter matrix of the flywheel processing into the improved BP neural network regression prediction model for training and learning, and determine the improved neuron kernel function R, Among them, y is the data information of the fusion detection parameter matrix of flywheel processing, ρ1, ρ2 and ρ3 are the learning factors of flywheel processing detection, and the trained improved BP neural network regression prediction model is obtained; Q53. Based on the trained improved BP neural network regression prediction model, the data information of the fusion detection parameter matrix of the flywheel processing is input to predict the processing quality of the flywheel, and obtain the predicted flywheel processing quality data information.

8. The fully automatic detection method for flywheel processing according to claim 7 is characterized in that: The learning factor ρ1 of the flywheel processing detection is, The learning factor ρ2 of the flywheel processing detection is, The learning factor ρ3 of the flywheel processing detection is, Among them, y is the data information of the fusion detection parameter matrix of flywheel processing.

9. The fully automatic detection method for flywheel processing according to claim 1, characterized in that: In step Q6, the building of the flywheel processing quality evaluation model and the evaluation of the flywheel processing quality include: Q61. Obtain the predicted flywheel processing quality data information, construct a flywheel processing quality data set, and divide it into a flywheel processing training data set and a flywheel processing test data set; Q62. Input the flywheel processing training data set into the flywheel processing quality evaluation model for training and learning, and determine the evaluation function W of the flywheel processing quality, Among them, g is the flywheel processing training data set, η1 and η2 are the evaluation factors of flywheel processing quality, and the evaluation model of flywheel processing quality after training is obtained; Q63. Based on the trained flywheel processing quality evaluation model, the flywheel processing test data set is input for optimization to obtain a trained flywheel processing quality evaluation model; Q64. Based on the trained flywheel processing quality evaluation model, the predicted flywheel processing quality data information is input, the flywheel processing quality is evaluated, and the flywheel processing quality evaluation data information is output.

10. The fully automatic detection method for flywheel processing according to claim 1, characterized in that: The method further comprises: Q7. Based on the evaluation data information of the processing quality of the flywheel, a preset threshold is set. If the evaluation data of the processing quality of the flywheel is less than the preset threshold, the quality requirements are not met and the flywheel is unqualified. If the evaluation data of the processing quality of the flywheel is greater than the preset threshold, the quality requirements are met and the flywheel is qualified.

Images