Fan assembly production data dynamic analysis system for intelligent manufacturing

Abstract

The invention relates to the technical field of fan assembly production data analysis, in particular to a fan assembly production data dynamic analysis system for intelligent manufacturing. According to the invention, data acquisition and analysis are carried out on processing, assembling and testing links in the production process of the small fan assembly, so that the quality monitoring of the whole process is realized. In addition, the quality monitoring is realized through the quality prediction of the next link; meanwhile, the deviation correction value calculated in the link is fed back to the previous link to automatically correct working parameters, closed-loop monitoring and control are carried out on all links of small fan assembly production, the data analysis effect in the small fan assembly production process is improved, scientific support is provided for improving the quality control effect in the production process, and the production efficiency is improved. And therefore, the production quality of an intelligent manufacturing product taking the small fan assembly as a key accessory is ensured.

Description

Technical Field

[0001] The present invention relates to the technical field of wind turbine component production data analysis, and in particular to a wind turbine component production data dynamic analysis system oriented to intelligent manufacturing. Background Art

[0002] The key data in the production process of small fan components covers multiple types of data generated in the steps of component processing, assembly and performance testing. The data in these links will directly or indirectly affect the production quality of fan components and reflect their performance in fan products. For example, the processing technology in the processing data directly affects the dimensional accuracy and surface quality of the fan components, affects the accuracy of the assembly, and then reflects on the operating stability of the fan products; for another example, the precision control in the processing data can ensure that the parts are more precise, reduce machine losses, and then reflect on the operating balance of the fan products. Therefore, it is necessary to conduct efficient and high-quality analysis of the production data of small fan components.

[0003] The research on small fan components in the prior art mainly focuses on optimizing the hardware design thereof so that the fan products using it as accessories have better working quality. However, there is a lack of targeted consideration and analysis on the production process of the fan components themselves, which may lead to inaccurate dimensions or inaccurate matching of the core components of the fan components, thereby affecting the quality of the fan components and resulting in failure to meet the application requirements of subsequent products.

[0004] In order to improve the quality control effect in the production process of fan components, this application proposes a dynamic analysis system for fan component production data for intelligent manufacturing. Summary of the invention

[0005] The purpose of the present invention is to provide a dynamic analysis system for fan assembly production data for intelligent manufacturing, which realizes quality monitoring of the whole process by collecting and analyzing data of the processing, assembly and testing links in the production process of small fan assemblies. In addition, quality monitoring is also realized by predicting the quality of the next link; at the same time, the deviation correction amount calculated in this link is fed back to the previous link to automatically correct the working parameters, and closed-loop monitoring and control are performed on each link of the production of small fan assemblies, which improves the data analysis effect in the production process of small fan assemblies, provides scientific support for improving the quality control effect of the production process, and ensures the production quality of intelligent manufacturing products with small fan assemblies as key accessories.

[0006] To achieve the above-mentioned object, the present invention provides a wind turbine component production data dynamic analysis system for intelligent manufacturing, comprising: a data acquisition module: used to collect production data in the wind turbine component production process, including processing data, assembly data and performance test data; Feature extraction module: extracts processing features from processing data, extracts assembly features from assembly data, and obtains performance features from performance test data; Data analysis module: obtains predicted assembly quality data based on processing features, and judges processing anomalies based on the deviation between the predicted assembly quality data and the actual assembly quality data; obtains predicted performance quality data based on assembly features, and judges assembly anomalies based on the deviation between the predicted performance quality data and the actual performance quality data; obtains quality data of fan components based on performance features, and judges quality anomalies; Feedback control module: analyzing the regression coefficient to obtain the ranking of assembly-influencing features; calculating the correction amount of the processing data according to the ranking of the assembly-influencing features, and feeding back the processing correction amount to the processing equipment to adjust the processing data; analyzing the regression coefficient to obtain the ranking of performance-influencing features; calculating the correction amount of the assembly data according to the ranking of the performance-influencing features, and feeding back the assembly correction amount to the assembly equipment to adjust the assembly data; Quality monitoring module: triggers quality warning when an abnormality occurs, and transmits deviation, corresponding working parameters and actual quality data to the terminal device.

[0007] Furthermore, the steps of obtaining the processing characteristics include: extracting the deviation between the processing size of the component and the required size; calculating the surface roughness of the component; recording the deviation and the surface roughness as processing accuracy characteristics; obtaining the cutting force fluctuation, vibration signal and temperature change during the processing, and recording them as processing dynamic characteristics; Integrate machining accuracy features and machining dynamic features into machining features.

[0008] Furthermore, the steps for obtaining assembly features include: extracting the fitting force and assembly error of parts and components, which are recorded as assembly accuracy features; extracting the assembly force calculation peak force and force fluctuation during the assembly process, which are recorded as assembly force features; extracting statistical features from the assembly time of parts and components, which are recorded as assembly time features; and integrating assembly accuracy features, assembly force features, and assembly time features into processing features.

[0009] Furthermore, the performance characteristics are used to obtain quality data of the fan assembly; the specific steps include: analyzing the vibration spectrum and noise level of the fan assembly to extract the vibration and noise characteristics; testing the operation of the fan assembly under different loads, extracting the load change and the power consumption under different loads, and recording them as load test characteristics; The vibration and noise characteristics and load test characteristics are integrated into performance characteristics, and the quality data of the fan components are obtained based on the performance characteristics; the quality abnormality of the fan components is judged by the deviation between the quality data and the quality requirements.

[0010] Furthermore, processing characteristics, assembly characteristics and performance characteristics are respectively input into the predictive regression model to obtain predicted assembly quality data, predicted performance quality data and quality data respectively; the construction of the predictive regression model is based on the correlation analysis of historical production data and historical actual quality data.

[0011] Furthermore, the step of determining the processing abnormality includes: Calculate predicted assembly quality data and actual assembly quality data The deviation between ;when Greater than threshold When the output is abnormal.

[0012] Furthermore, the step of determining the assembly abnormality includes: Compute predicted performance quality data and actual performance quality data The deviation between ;when Greater than threshold , the output assembly is abnormal.

[0013] Further, the step of adjusting the processing data according to the processing correction amount includes: the assembly influence feature sorting is expressed as ,in, For the The assembly affects the feature value. is the number of assembly-affecting features; the assembly-affecting features are sorted in descending order; Processing correction amount The calculation formula is: ; in, For the The weight of each assembly influencing feature is set according to the order of assembly influencing features; is a nonlinear transformation function, is the processing adjustment coefficient, To correct the offset; According to the processing correction amount Adjust processing data.

[0014] Further, the step of adjusting the assembly data according to the assembly correction amount includes: the assembly influence feature ranking is expressed as ,in, For the Each performance affects the characteristic value, is the number of features that affect performance; the sorting of features that affect performance is in descending order; the amount of assembly correction The calculation formula is: ; in, For the Each performance affects the weight of the feature, is the assembly adjustment coefficient, is a logarithmic function. According to the assembly correction Adjust assembly data.

[0015] Furthermore, the data analysis module is continuously monitored to trigger quality warnings when processing abnormalities, assembly abnormalities and quality abnormalities occur; the abnormality category, corresponding deviation, working parameters and actual quality data are output and transmitted to the terminal device.

[0016] Compared with the prior art, the present invention has the following beneficial effects: 1. The present invention realizes full-process monitoring of the production process of small fan components through the data acquisition module, covering the three key links of processing, assembly and performance testing. By collecting various types of production data, including processing accuracy, assembly accuracy, vibration and noise data, and combining the analysis of the feature extraction module, the extracted key features are used as the data basis for subsequent abnormal judgment, and the production status of each production link can be accurately and comprehensively understood. This full-process monitoring can dynamically control the quality of small fan components in each production link, and provide auxiliary decision-making support for improving the production quality of small fan components.

[0017] 2. The present invention mines and analyzes the correlation between historical production data and historical quality data through a predictive regression model, and predicts the quality data of the next link based on this correlation. Then, according to the deviation between the predicted quality and the actual quality / demand quality, it can identify possible quality abnormalities in advance and provide accurate quality warnings. If an abnormality occurs in the processing link, it can be identified and warned in the assembly link to prevent quality defects from being transmitted to downstream links. It can also find out the influencing factors that may cause the abnormality based on the situation in which the abnormality occurs.

[0018] 3. The present invention calculates the corresponding correction amount when quality abnormalities are detected based on the regression coefficient in the prediction regression model, and feeds it back to the previous link. When the current processing abnormality is detected based on the deviation between the predicted assembly quality and the actual assembly quality, the present invention sorts the processing features that cause the assembly deviation according to the regression coefficient through the feedback control module, and then calculates the correction amount of each processing feature based on the sorting, and feeds the correction amount back to the processing equipment to adjust the processing data. The handling method of assembly abnormalities is similar. The closed-loop control mechanism of the present invention ensures that each link in the production process can be adjusted in time according to the actual situation, avoiding the accumulation and transmission of deviations, optimizing the quality control of the entire production process, and thus ensuring the quality of subsequent intelligent manufacturing fan products. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] Figure 1 This is a schematic diagram of the workflow of the dynamic analysis system for production data of small wind turbine components; Figure 2 Dynamic analysis of structural diagrams for wind turbine component production data; Figure 3 Data flow diagram for processing data; Figure 4 A data flow diagram for assembly data. DETAILED DESCRIPTION

[0020] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0021] In order to improve the overall production quality of small fan components and optimize production efficiency, a smart fan component manufacturer introduced a fan component production data dynamic analysis system for smart manufacturing to control the quality of the production process of small fan components. Figure 1 Given, Figure 1 The following is a schematic diagram of the workflow of a small wind turbine component production data dynamic analysis system. The specific implementation of the wind turbine component production data dynamic analysis system will be described below through Example 1 and Example 2.

[0022] Embodiment 1 This embodiment sets the key steps of fan assembly production as processing, assembly and performance testing; the data acquisition module is used to collect production data in the fan assembly production process, including processing data, assembly data and performance test data. Processing data refers to information related to processing quality during the processing of small fan assemblies, assembly data refers to information related to assembly quality during the assembly of small fan assemblies, and performance test data refers to information that can reflect the actual operating status of the fan during the performance test stage. Table 1 gives some examples of production data.

[0023] Table 1 Some examples of production data

[0024] It should be noted that the data listed in Table 1 are not all the data based on which the subsequent data processing and analysis of this embodiment is based. All the data in the production process of small fan components can be used as the basis for the subsequent processing and analysis of this embodiment. The collection of various types of production data in Table 1 can be achieved through a variety of mature existing technologies (such as sensors), which will not be described in detail in this embodiment.

[0025] Reference Figure 2 , Figure 2 A structural diagram of the dynamic analysis of wind turbine component production data provided for this implementation.

[0026] Further, refer to Figure 3 and Figure 4 , Figure 3 and Figure 4 They are respectively the processing data flow diagram and the assembly data flow diagram provided in this embodiment.

[0027] This embodiment uses a feature extraction module to extract processing features from processing data, extract assembly features from assembly data, and obtain performance features from performance test data. Specifically: Measure the deviation between the actual processing size of the parts and the required size; scan the surface of the parts and calculate the deviation according to the integral formula Calculate the surface roughness of the component; wherein, is the surface roughness; To measure length, is the surface height of the component. Both the measured length and surface height are obtained by scanning.

[0028] The deviation amount and the surface roughness are recorded as processing accuracy characteristics.

[0029] The cutting force fluctuation, vibration signal and temperature change during the machining process are obtained and recorded as the machining dynamic characteristics; the specific steps include: continuously monitoring the change of cutting force over time during the machining process to obtain the cutting force fluctuation; then The cutting force fluctuation at the moment is: ; in, for Cutting force fluctuations at all times, is the average cutting force during monitoring time, for Cutting force at all times.

[0030] Continuously monitor the vibration signals of processing equipment and obtain the vibration spectrum; monitor the temperature changes of processing equipment and parts; record the cutting force fluctuations, vibration spectrum and temperature changes as processing dynamic characteristics; The machining accuracy features and machining dynamic features are integrated into a machining feature matrix. Table 2 gives some examples of machining feature data of multiple fan components during machining.

[0031] Table 2 Some examples of processing feature data

[0032] This embodiment uses a feature extraction module to obtain processing features during the fan component processing, including processing accuracy and other fluctuation data generated over time, which can provide accurate key data for quality analysis and process optimization in subsequent links, reducing the complexity of subsequent data processing. Processing accuracy features and processing dynamic features extract key quality parameters in the fan component processing process, providing a reliable basis for quality monitoring and abnormality identification.

[0033] Measure the fit force between various parts of the components during the assembly process; calculate the assembly error based on the actual measured assembly value and the target value, and record the fit force and assembly error as assembly accuracy characteristics; continuously monitor the assembly force during the assembly process, calculate the force peak value and force fluctuation, and record them as assembly force characteristics; the method for obtaining the force peak value is expressed as ,in is the peak force, To obtain the maximum value, For the moment assembly force.

[0034] Statistical features are extracted from the assembly time of parts and components and recorded as assembly time features. The specific steps include: calculating the average assembly time, the longest assembly time and the shortest assembly time by recording the assembly time; integrating the assembly accuracy features, assembly force features and assembly time features into an assembly feature matrix. Table 3 gives some examples of assembly feature data of multiple fan components during the assembly process.

[0035] Table 3 Partial example of assembly characteristic data

[0036] The assembly features obtained in this embodiment can comprehensively and accurately monitor the assembly process of the fan assembly. The matching force and assembly error reflect the accuracy of the assembly process, the assembly force peak and fluctuation reflect the mechanical stability during the assembly process, and the assembly time characteristics can reflect the assembly efficiency; these assembly features provide key data support for subsequent quality prediction and production optimization.

[0037] Monitor the vibration spectrum and noise level of the fan assembly during operation to extract vibration and noise characteristics. Test the operation of the fan assembly under different loads, extract the load change and power consumption under different loads, and record them as load test characteristics; The vibration and noise characteristics and the load test characteristics are integrated into the performance characteristic matrix. Table 4 gives a partial example of performance characteristic data of multiple wind turbine groups during the performance test process.

[0038] Table 4 Some examples of performance characteristic data

[0039] The vibration noise characteristics and load test characteristics are weighted to obtain the quality data of the fan components; the weighted formula is: ; in, For quality data, is the vibration noise characteristic, For load testing characteristics; and They are the weights of noise vibration characteristics and load test characteristics, which are set by technicians in this field according to actual needs.

[0040] The deviation between the quality data and the quality requirements can be used to determine the quality anomalies of the fan components. and quality requirements The deviation between them is the difference between the two; when the absolute value of the difference Greater than the set threshold When , it is determined that the current fan component has quality abnormalities; it is expressed as: ; in, is the result of quality abnormality judgment. If there is quality abnormality, then On the contrary, .

[0041] This embodiment grasps the performance of the fan assembly during the performance test phase from the perspectives of vibration, noise, load, and power consumption. These performances can reflect the quality of the fan assembly in actual production. These performance characteristics provide real-time feedback on the quality of the fan assembly and can also serve as a basis for judging whether the quality is abnormal. By judging the deviation between the performance characteristics and the quality requirements, the quality abnormality of the fan assembly can be identified in a timely manner, providing real-time decision-making reference for the management party.

[0042] Furthermore, this embodiment uses a data analysis module to obtain predicted assembly quality data based on processing characteristics, and judge processing abnormalities based on the deviation between the predicted assembly quality data and the actual assembly quality data; obtain predicted performance quality data based on assembly characteristics, and judge assembly abnormalities based on the deviation between the predicted performance quality data and the actual performance quality data. Specifically: The processing features and assembly features are respectively input into the prediction regression model to obtain the predicted assembly quality data and the predicted performance quality data respectively; the prediction regression model is constructed based on the correlation analysis of historical production data and historical actual quality data.

[0043] The training steps of the prediction regression model are: clean the historical production data and the historical actual quality results and then normalize them; the normalization method used in this embodiment is the Min-Max normalization method. The linear relationship between the historical processing characteristics and the historical assembly quality, and the historical assembly characteristics and the historical performance quality in the historical production data is established through the linear regression model. The linear regression model is expressed as: ; in, To predict assembly quality data, To predict performance quality data; and Respectively Historical processing features and their corresponding weight coefficients, and Respectively Historical assembly features and their corresponding weight coefficients; and are the number of historical processing features and historical assembly features, and are the biases of predicted assembly quality data and predicted performance quality data, respectively.

[0044] This embodiment iteratively optimizes the established linear regression model through mean square error.

[0045] This embodiment obtains the correlation between historical production data and historical quality data through a predictive regression method, and constructs a predictive regression model; the assembly quality and performance quality of the fan component are predicted from the processing characteristics and assembly characteristics according to the predictive regression model, which can improve the predictive ability in the production process of the fan component and provide data for feedback optimization of subsequent working parameters.

[0046] The steps for judging processing abnormalities include: Calculate the predicted assembly quality data and the actual assembly quality data The deviation between ;when Greater than threshold When the output is abnormal.

[0047] The steps for determining assembly anomalies include: Calculate the predicted performance quality data and the actual performance quality data The deviation between ;when Greater than threshold , the output assembly is abnormal.

[0048] In this embodiment, the threshold and threshold They are set to 0.030 and 0.050 respectively; Table 5 gives some examples of judging abnormalities in processing and assembly; Table 5 Examples of judging abnormal processing and assembly

[0049] This embodiment determines whether there is an abnormality in the current processing process by calculating the deviation between the predicted assembly quality and the actual assembly quality; if there is an abnormality in the processing process, it will affect the accuracy of the subsequent assembly combination. If there is an abnormality in the processing size, it will cause the components to not fit tightly during the assembly process; if the processing temperature is too high or too low, it will cause assembly expansion and contraction problems. This step is conducive to improving the accuracy of the small fan component processing process and reducing quality problems caused by processing errors.

[0050] This embodiment determines whether there is an abnormality in the current assembly process by calculating the deviation between the predicted performance quality and the actual performance; if there is an abnormality in the assembly process, it will affect the subsequent operating performance of the fan assembly. If the assembly is inaccurate or the assembly force is uneven, it may cause misalignment between the parts of the fan assembly. This step is conducive to ensuring that the assembly of the fan assembly can meet the expected performance standards and prevent the final quality of the product from being affected by assembly abnormalities.

[0051] Furthermore, in this embodiment, a feedback control module is used to analyze the regression coefficient to obtain the assembly influence feature ranking; the correction amount of the processing data is calculated according to the assembly influence feature ranking, and the processing correction amount is fed back to the processing equipment to adjust the processing data; Analyze the regression coefficient to obtain the ranking of performance-affecting features; calculate the correction amount of the assembly data according to the ranking of performance-affecting features, and feed back the assembly correction amount to the assembly equipment to adjust the assembly data. Specifically: According to the linear regression model calculated in the previous step and , sorted in descending order, obtain the ranking of assembly influencing features and performance influencing features respectively, and discard the influencing features whose weight coefficients are lower than the sensitive value.

[0052] The step of adjusting the processing data according to the processing correction amount includes: the assembly-affecting feature sorting is expressed as: ,in, For the The assembly affects the feature value. is the number of assembly-affecting features; the assembly-affecting features are sorted in descending order; the processing correction amount The calculation formula is: ; in, For the The weight of each assembly influencing feature is set according to the order of the assembly influencing features; is a nonlinear transformation function, is the processing adjustment coefficient, To correct the offset; According to the processing correction amount The processing data is adjusted.

[0053] The step of adjusting the assembly data according to the assembly correction amount includes: the assembly influencing feature ranking is expressed as: ,in, For the Each performance affects the characteristic value, is the number of performance-affecting features; the performance-affecting features are sorted in descending order; the assembly correction amount The calculation formula is: ; in, For the Each performance affects the weight of the feature, is the assembly adjustment coefficient, is a logarithmic function.

[0054] According to the assembly correction amount The assembly data is adjusted.

[0055] This embodiment is aimed at the production process of this batch of fan components of smart fan component manufacturer A, and some examples of processing correction amounts and assembly correction amounts are given in Table 6.

[0056] Table 6 Partial example of machining correction amount

[0057] It should be noted that the present embodiment uses processing parameters and assembly parameters as examples for calculating processing correction amounts and assembly correction amounts, and Table 6 only provides examples of some processing parameters and assembly parameters. Based on the understanding of the present application by those skilled in the art, other processing data and assembly data can be used to calculate the correction amounts through the above method and simple variations of the above method.

[0058] This embodiment can prioritize the processing data or steps that have a greater impact on the assembly quality by sorting the features that affect the assembly quality, timely and accurately adjust the processing process, and ensure the accuracy and stability of the processing process. The processing correction amount calculated based on the nonlinear relationship between the processing data and the assembly quality accurately reflects the degree to which the processing data needs to be adjusted in the current actual assembly situation from a local perspective, thereby improving the accuracy and flexibility of the quality monitoring of the processing process.

[0059] This embodiment can prioritize the assembly data or steps that have a greater impact on performance quality by sorting the features that affect performance quality, and timely compensate and correct the defects in the assembly process that are fed back, thereby improving the accuracy of wind turbine assembly and ensuring the stability of subsequent product quality. Through nonlinear changes and logarithmic functions, different performance-related assembly data are adaptively adjusted from a global quality perspective to reduce errors in the assembly process.

[0060] Furthermore, this embodiment uses a quality monitoring module to trigger a quality warning when an abnormality occurs, and transmits the deviation, corresponding working parameters and actual quality data to the terminal device. Specifically: The data analysis module is continuously monitored to trigger a quality warning when the processing abnormality, the assembly abnormality and the quality abnormality occur; the abnormality category, the corresponding deviation, the working parameters and the actual quality data are output and transmitted to the terminal device.

[0061] This embodiment can continuously monitor and timely detect abnormalities in the entire process of processing, assembly and performance testing of small fan components through the quality monitoring module. Once an abnormality is detected, an early warning is triggered and the abnormality category, deviation, working parameters and actual quality data are output and presented to relevant staff as a reference for decision-making, providing support for the quality control of small fan components in the production process.

[0062] Embodiment 2 This embodiment will explain the influence of processing data on assembly quality and the influence of assembly data on performance quality based on the production process of small fan components; the main data explanations are given below.

[0063] Analysis of the impact of processing data on assembly quality; Processing size error: affects the tightness of fit and assembly stability, and determines whether the fan assembly meets the design tightness requirements after assembly. Improper fit may cause loose components or abnormal bite during equipment operation. Surface roughness: has a direct impact on the friction coefficient and wear resistance of the contact surface. Larger surface roughness may increase the fit resistance or reduce the tightness effect; Hole-shaft coaxiality: It will affect the concentricity of rotating parts. Excessive deviation may cause excessive vibration amplitude after assembly.

[0064] Analysis of the impact of assembly data on performance quality; Assembly force: affects the sealing and load transfer capacity of the contact parts. Too small an assembly force may cause the fan assembly to loosen, while too large an assembly force may cause irreversible permanent deformation or contact stress concentration. Assembly clearance: determines the friction coefficient and sliding resistance of the relative movement of parts. Too large a clearance will cause vibration and noise to exceed reasonable standards, while too small a clearance may cause jamming or friction overheating. Torque uniformity: Uneven torque can cause speed fluctuations and reduce the smooth operation of the transmission in the fan assembly; Assembly angle: The impact of the assembly angle of parts is directly related to the alignment and coordination efficiency between wind turbine components.

[0065] In view of the problem that the existing technology ignores the data analysis in the production process of small fan components, this application proposes a dynamic analysis system for fan component production data for intelligent manufacturing, aiming to improve the quality control effect in the production process of fan components. This application first collects and analyzes data on the processing, assembly and testing links in the production process of small fan components to achieve quality monitoring of the entire process. In addition, quality monitoring is achieved by predicting the quality of the next link; at the same time, the deviation correction amount calculated in this link is fed back to the previous link to automatically correct the working parameters, and closed-loop monitoring and control are performed on each link of the small fan component, which improves the data analysis effect in the production process of small fan components, provides scientific support for improving the quality control effect of the production process, and thus ensures the production quality of intelligent manufacturing products with small fan components as key accessories.

[0066] Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. A wind turbine component production data dynamic analysis system for intelligent manufacturing, characterized in that: include: Data acquisition module: used to collect production data during the production process of fan components, including processing data, assembly data and performance test data; Feature extraction module: extracting processing features from the processing data, extracting assembly features from the assembly data, and acquiring performance features from the performance test data; Data analysis module: obtaining predicted assembly quality data according to the processing characteristics, and judging processing abnormality according to the deviation between the predicted assembly quality data and the actual assembly quality data; Acquire predicted performance quality data according to the assembly features, and determine assembly abnormality according to a deviation between the predicted performance quality data and actual performance quality data; Acquiring quality data of the fan component according to the performance characteristics, and determining quality abnormalities; Feedback control module: Analyze regression coefficients to obtain the order of assembly-influencing features; Calculate the correction amount of the processing data according to the assembly influencing feature sorting, and feed back the processing correction amount to the processing equipment to adjust the processing data; Analyzing the regression coefficients to obtain a ranking of performance-influencing features; Calculate the correction amount of the assembly data according to the ranking of the performance-influencing features, and feed back the assembly correction amount to the assembly equipment to adjust the assembly data; Quality monitoring module: triggers quality warning when the processing abnormality, assembly abnormality and quality abnormality occur, and transmits abnormality-related data to the terminal device.

2. The wind turbine component production data dynamic analysis system for intelligent manufacturing according to claim 1 is characterized in that: The step of acquiring the processing features comprises: Extract the deviation between the machining size of the component and the required size; calculate the surface roughness of the component; record the deviation and the surface roughness as machining accuracy features; obtain the cutting force fluctuation, vibration signal and temperature change during the machining process, and record them as machining dynamic features; The machining accuracy feature and the machining dynamic feature are integrated into the machining feature.

3. The wind turbine component production data dynamic analysis system for intelligent manufacturing according to claim 1 is characterized in that: The step of obtaining the assembly feature comprises: The matching force and assembly error of parts are extracted and recorded as assembly accuracy features. The assembly force calculation peak value and force fluctuation during the assembly process are extracted and recorded as assembly force features. Statistical features are extracted from the assembly time of the parts and recorded as assembly time features. The assembly accuracy features, the assembly force features and the assembly time features are integrated into the assembly features.

4. The wind turbine component production data dynamic analysis system for intelligent manufacturing according to claim 1 is characterized in that: The performance characteristics are used to obtain the quality data of the fan component; the specific steps include: Analyze the vibration spectrum and noise level of the fan assembly to extract vibration and noise characteristics; test the operation of the fan assembly under different loads, extract load changes and power consumption under different loads, and record them as load test characteristics; The vibration noise characteristics and the load test characteristics are integrated into the performance characteristics, and the quality data of the fan component is obtained according to the performance characteristics; and the quality abnormality of the fan component is judged by the deviation between the quality data and the quality requirements.

5. The wind turbine component production data dynamic analysis system for intelligent manufacturing according to claim 1 is characterized in that: The processing features and the assembly features are respectively input into a prediction regression model to obtain the predicted assembly quality data and the predicted performance quality data respectively; the prediction regression model is constructed based on the correlation analysis of historical production data and historical actual quality data.

6. The wind turbine component production data dynamic analysis system for intelligent manufacturing according to claim 1 is characterized in that: The step of determining the processing abnormality comprises: pass Calculate the predicted assembly quality data and the actual assembly quality data The deviation between ;when Greater than threshold When the output is abnormal.

7. The wind turbine component production data dynamic analysis system for intelligent manufacturing according to claim 1 is characterized in that: The step of determining the assembly abnormality comprises: pass Calculate the predicted performance quality data and the actual performance quality data The deviation between ;when Greater than threshold , the output assembly is abnormal.

8. The wind turbine component production data dynamic analysis system for intelligent manufacturing according to claim 1 is characterized in that: The step of adjusting the processing data according to the processing correction amount comprises: The assembly-affecting feature ranking is expressed as ,in, For the The assembly affects the feature value. The number of features that affect assembly, is a sorting function; the assembly influencing feature sorting is in descending order; The processing correction amount The calculation formula is: ； in, For the The weight of each assembly influencing feature is set according to the order of the assembly influencing features; is a nonlinear transformation function, is the processing adjustment coefficient, To correct the offset; According to the processing correction amount The processing data is adjusted.

9. The wind turbine component production data dynamic analysis system for intelligent manufacturing according to claim 1, characterized in that: The step of adjusting the assembly data according to the assembly correction amount comprises: The assembly-affecting feature ranking is expressed as ,in, For the Each performance affects the characteristic value, is the number of performance-influencing features; the performance-influencing features are sorted in descending order; The assembly correction amount The calculation formula is: ； in, For the Each performance affects the weight of the feature, is the assembly adjustment coefficient, is a logarithmic function; According to the assembly correction amount The assembly data is adjusted.

10. The wind turbine component production data dynamic analysis system for intelligent manufacturing according to claim 1, characterized in that: The data analysis module is continuously monitored, and a quality warning is triggered when the processing abnormality, the assembly abnormality, and the quality abnormality occur; the abnormality category, the corresponding deviation, the working parameters, and the actual quality data are output and transmitted to the terminal device.

Images