A method and system for optimizing and controlling gear production lines based on intelligent algorithms

Abstract

This invention discloses an optimized control method and system for gear production lines based on intelligent algorithms, specifically relating to the field of gear production control technology. The method includes the following steps: Step S1, using an industrial camera to record and process the surface of the produced gears, obtaining feature values, and then calculating the power spectral density of the gear surface roughness; Step S2, acquiring screening aperture data, grinding accuracy data, and heating temperature data, combining them to obtain a total correction coefficient, and comparing the calculated power spectral density of the gear surface roughness with the total correction coefficient to obtain a sub-correction coefficient; Step S3, using the sub-correction coefficients to correct the threshold of the total correction coefficient. By using online recording with an industrial camera and calculating the power spectral density of the roughness surface, the limitations of traditional single-point measurement are overcome to achieve precise correlation between multi-process parameters and surface quality. Simultaneously, the lag in offline detection is eliminated, enabling immediate triggering of corrections to avoid batch defects and significantly reducing production losses.

Description

Technical Field

[0001] This invention relates to the field of gear production control technology, and more specifically, to a gear production line optimization control method and system based on intelligent algorithms. Background Technology

[0002] As a core component of mechanical transmission systems, gears are widely used in key fields such as aerospace, shipbuilding, automobiles, and agricultural machinery. Their surface quality and dimensional accuracy directly determine transmission efficiency, operational stability, and service life. With the rapid development of high-end equipment manufacturing, the market's quality requirements for gears are constantly upgrading. Precision gears need to have surface roughness controlled at the micron or even nanometer level, and must possess stable metallographic structure and mechanical properties. This poses a stringent challenge to the coordinated control of multiple core processes such as screening, grinding, and heating.

[0003] Traditional gear surface quality assessment relies heavily on contact roughness meters to measure single-point values, which only reflect local roughness and cannot quantify the distribution of morphological features at different spatial frequencies. This makes it difficult to accurately correlate the intrinsic relationship between multiple process parameters such as sieve aperture fluctuations, grinding energy changes, and heating temperature deviations and surface quality. Furthermore, manual sampling or offline inspection methods are inherently lagging; by the time quality defects are discovered, a batch of non-conforming products has already been generated, significantly increasing rework costs and production losses.

[0004] To address the aforementioned problems, a technical solution is provided. Summary of the Invention

[0005] In order to overcome the above-mentioned defects of the prior art, embodiments of the present invention provide a gear production line optimization control method and system based on intelligent algorithms to solve the problems mentioned in the background art.

[0006] To achieve the above objectives, the present invention provides the following technical solution:

[0007] The gear production line optimization control method based on intelligent algorithms includes the following steps:

[0008] Step S1: Use an industrial camera to record the surface of the gear after production, obtain feature values, and then calculate the power spectral density of the gear surface roughness.

[0009] Step S2: Obtain sieve aperture data, grinding accuracy data, and heating temperature data, combine them to obtain the total correction coefficient, and compare the calculated power spectral density of the gear surface roughness with the total correction coefficient to obtain the sub-correction coefficient.

[0010] Step S3: Correct the total correction coefficient threshold using the sub-correction coefficients;

[0011] Step S4: Sort the correction order according to the correction results, and process according to the sorting results.

[0012] In a preferred embodiment, step S1 includes the following specific contents:

[0013] Calculate the independent coefficients of the surface roughness of the gear, denoted as... ;

[0014] The calculation formula is:

[0015] ;

[0016] in, The root mean square of the height undulation of the rough surface. This represents the distance between two points on a rough surface. Indicates the overall length of the rough surface. Represents the natural constant;

[0017] The initial surface roughness power spectral density of the gear is calculated using the following formula:

[0018] ;

[0019] Combining the above two equations, we obtain the final surface roughness power spectral density of the gear surface, calculated as follows:

[0020] .

[0021] In a preferred embodiment, step S2 includes the following specific contents:

[0022] The continuous aperture size sequence acquired by the laser diameter measurement system is divided into equal parts. The data segment is used to calculate the... Median value of segment aperture size If the aperture size deviates from the specified value for three consecutive sampling points within a certain time period... Exceeding the standard deviation of this segment If so, then mark that time period as the data segment for filtering out interference;

[0023] The grinding wheel feed rate and vibration amplitude of the grinding machine are collected, multiplied to obtain the grinding energy sequence, and then divided into equal parts. The data segment is used to calculate the... Median value of grinding energy If, within a certain time period, the grinding energy is higher than the standard value for five consecutive sampling points... If it is 1.2 times that, it is marked as a data segment for grinding interference;

[0024] The temperature curves of the multi-zone heat treatment furnace are divided into equal parts according to time. The data segment; calculate the first data segment. Median temperature of the segment If the temperature deviates from the normal range for 3 consecutive minutes within a certain period of time... If the temperature exceeds ±10℃, it is marked as a temperature interference data segment;

[0025] For the screening interference data segment, statistically analyze the relationship between each aperture size and the screening interference data segment. The absolute values ​​of the deviations are summed and then divided by the number of data points to obtain the mean deviation per unit point, which is... ;

[0026] For the grinding interference data segment, calculate the grinding energy and its relationship within each grinding interference data segment. The percentage deviation is calculated using the following formula:

[0027] The average of all deviation percentages is... ;

[0028] For the temperature interference data segment, take the temperature within the temperature interference data segment and... The maximum value of the absolute value of the deviation is... .

[0029] In a preferred embodiment, the median time of the segment filtering out interfering data is used as the center time point. ;turn up The most recent non-interference data segment, denoted by its median time as And use this non-interference segment as the reference segment;

[0030] extract arrive The aperture data between the two sections are used to calculate the size change trend rate as the trend intensity through linear fitting; the aperture data of the interference section and the reference section are sorted by time and the difference is calculated point by point, and then multiplied by the trend intensity to obtain the screening stability offset, which is then used to form the screening stability offset sequence. Similarly, the grinding stability offset sequence and the temperature stability offset sequence are generated.

[0031] If you grind down the interfering data segments If the intensity of the grinding interference is greater than the average of all grinding interference segments, and the grinding energy within the segment exhibits an oscillating characteristic of first rising, then falling, and then rising again, it is determined to be external oscillation interference. Using the difference between the maximum and minimum vibration amplitudes within the segment as a benchmark, the ratio of the vibration amplitude at each sampling point to the benchmark value is calculated, multiplied by a preset oscillation coefficient, to obtain the grinding oscillation offset, forming a grinding oscillation offset sequence.

[0032] If the disturbance is stable, the stable offset sequence of screening is added point by point to the original aperture data of the disturbance section of screening. If the disturbance is oscillating, the oscillating offset sequence of grinding is superimposed to obtain the corrected screening aperture data. Similarly, the feed rate and vibration amplitude data of the disturbance section of grinding are corrected point by point using the stable offset sequence of grinding or the oscillating offset sequence of grinding to obtain the corrected grinding accuracy data. The temperature curve of the disturbance section of temperature is corrected point by point using the stable offset sequence of temperature or the oscillating offset sequence of temperature to obtain the corrected heating temperature data.

[0033] The corrected screening aperture data, corrected grinding accuracy data, and corrected heating temperature data were normalized, and the normalized data were weighted and summed according to their weights to reflect the comprehensive correction requirements of the three types of process parameters on the rough surface of the gear.

[0034] The deviation rate between the corrected sieve aperture and the standard aperture, the conformity of the grinding accuracy with the standard grinding parameters, and the matching degree of the heating temperature with the standard temperature curve were calculated separately. Based on the weights of the three types of data, the sieve correction coefficient, grinding correction coefficient, and heating correction coefficient were calculated.

[0035] In a preferred embodiment, step S3 includes the following specific contents:

[0036] From the historical production database, select consecutive qualified batches whose power spectral density of gear surface roughness completely meets the preset quality standard within the past 3 months, and extract the total correction coefficient corresponding to these batches.

[0037] Remove extreme values ​​caused by sudden equipment failure or abnormal raw material batches, and take the arithmetic mean of the remaining data as the initial threshold baseline for the total correction coefficient. This baseline reflects the basic correction requirement level when the gear surface quality is stable and qualified.

[0038] The process deviation analysis corresponding to the correction coefficient is conducted for the three processes of screening, grinding, and heating. The deviation characteristics between the current process parameters and the standard parameters are analyzed independently based on their respective correction coefficients.

[0039] Using the sieving correction coefficient as a reference, the actual measured value of the corrected sieve aperture is compared with the tolerance range of the standard aperture, and the frequency of aperture values ​​exceeding the tolerance centerline is statistically analyzed.

[0040] Based on the grinding correction coefficient, analyze the degree of difference between the corrected grinding accuracy parameters and the standard parameters, and record the number of surface scratches and raised defects through a visual inspection system;

[0041] Using the heating correction factor as a reference, compare the temperature difference between the corrected temperature curve and the standard temperature curve at key nodes, and calculate the cumulative sum of the differences.

[0042] In a preferred embodiment, the initial threshold baseline is directly added to the correction amounts of the above three processes to obtain the updated total correction coefficient threshold. The effective range is set according to the quality grade of gear production, with the threshold range for precision gears being [0.2, 0.7] and for ordinary gears being [0.1, 0.8].

[0043] If the updated threshold is lower than the lower limit of the interval, it will be automatically adjusted to the lower limit value; if it is higher than the upper limit, it will be adjusted to the upper limit value. When the real-time total correction coefficient exceeds the threshold, the parameter adjustment process of the corresponding process will be triggered.

[0044] In a preferred embodiment, step S4 includes the following specific details:

[0045] When a correction order trigger instruction is received, it is first integrated to form a set of correction tasks to be processed, including historical incomplete correction tasks and their corresponding existing order sequences, as well as correction tasks newly added in the current production process.

[0046] When determining the sorting priority of newly added correction tasks, it is necessary to first retrieve the historical correction task database, extract historical task data of the same process type as the new task, and classify the screening process, grinding process, and heating process into different types. The focus is on statistically analyzing the average execution cycle of historical correction tasks of the same type and the trigger frequency of historical correction tasks of the same type within the same production cycle, and then further calculate and determine the sorting priority of each new correction task.

[0047] When calculating the average execution cycle of historical correction tasks of the same type, it is necessary to first obtain the actual execution time of a single correction task of the same type in the historical database, and then eliminate the influence of extreme abnormal durations by weighted averaging to finally obtain the standard average execution cycle of this type of task.

[0048] In a preferred embodiment, when determining the sorting priority of new correction tasks by combining historical data, the comprehensive priority index of each process type correction task is first calculated based on the average execution cycle and historical trigger frequency of historical similar tasks; then, according to the value of the comprehensive priority index, a corresponding sorting priority level is set for each type of process correction task; finally, according to the process type to which the new correction task belongs, the corresponding sorting priority level is matched for it.

[0049] When calculating the comprehensive priority index of each process type of correction task, a weighted calculation method is adopted: multiply the average execution cycle of the task type by the first weight factor to obtain the first priority component; multiply the historical trigger frequency of the task type by the second weight factor to obtain the second priority component; and then add the two priority components to finally obtain the comprehensive priority index of the task type.

[0050] In a preferred embodiment, when sorting the newly added correction tasks to generate corresponding sorting results, the tasks are first sorted from high to low according to their sorting priority level to form an initial candidate sorting sequence. Then, the initial candidate sorting sequence is cleaned up by merging duplicate tasks. If there are duplicate correction task entries for the same process in the sequence, the one with the highest priority is retained and the other duplicate entries are deleted to finally obtain the sorting result of the newly added correction tasks.

[0051] The existing sorting sequences of historical tasks that have not been completed are integrated and merged with the sorting results of newly added tasks to form the final target task sorting sequence; at the same time, the original historical task sorting sequence is replaced with this target sorting sequence.

[0052] The gear production line optimization control system based on intelligent algorithms is used to implement the control method described above, including a recording and calculation module, a comparison and analysis module, a correction module, and a sorting module; the modules are connected by signals.

[0053] The camera and calculation module uses an industrial camera to record the surface of the gear after production, and calculates the power spectral density of the gear surface roughness after obtaining the feature values.

[0054] The comparative analysis module is used to obtain screening aperture data, grinding accuracy data, and heating temperature data. After combining them, the total correction coefficient is obtained. The calculated power spectral density of the gear surface roughness is compared with the total correction coefficient to obtain the sub-correction coefficient.

[0055] The correction module uses sub-correction coefficients to correct the total correction coefficient threshold;

[0056] The sorting module sorts the correction order according to the correction results and processes the results accordingly.

[0057] The technical effects and advantages of the intelligent algorithm-based gear production line optimization control method and system of this invention are as follows:

[0058] 1. Excessive aperture tolerance in the screening process can cause periodic large-scale undulations on the gear surface, corresponding to energy peaks in the low-frequency band of the power spectral density; abnormal vibration amplitude in the grinding process can cause micro-scratches in the high-frequency band, manifested as an energy rise in the high-frequency band of the power spectral density; deviations in heating temperature can indirectly change the surface roughness after grinding by affecting the material hardness, manifested as an energy shift across the entire frequency band of the power spectral density. This precise correlation allows production personnel to quickly locate which process parameter abnormality is causing the quality problem by observing changes in the power spectral density, significantly improving the accuracy of process optimization.

[0059] 2. The correction process is triggered the moment a quality deviation occurs in a single gear. If the power spectral density of a gear shows abnormal energy in the high-frequency band, the system immediately identifies grinding interference data through step S2, corrects the total threshold through step S3, and sorts grinding correction tasks through step S4. The grinding parameters are adjusted before the next gear is produced, avoiding batch replication of defects. This design greatly improves the first-pass yield of gears, while reducing the loss of raw materials and time due to rework and scrap, thus reducing the production loss rate and significantly improving the stability of the production line. Attached Figure Description

[0060] Figure 1 This is a flowchart illustrating the gear production line optimization control method and system based on intelligent algorithms of the present invention.

[0061] Figure 2 This is a schematic diagram of the structure of the gear production line optimization control method and system based on intelligent algorithms of the present invention. Detailed Implementation

[0062] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0063] Example 1

[0064] This invention relates to an optimized control method and system for gear production lines based on intelligent algorithms.

[0065] Figure 1 The present invention provides an optimized control method for a gear production line based on intelligent algorithms, which includes the following steps:

[0066] Step S1: Use an industrial camera to record the surface of the gear after production, obtain feature values, and then calculate the power spectral density of the gear surface roughness.

[0067] Step S2: Obtain sieve aperture data, grinding accuracy data, and heating temperature data, combine them to obtain the total correction coefficient, and compare the calculated power spectral density of the gear surface roughness with the total correction coefficient to obtain the sub-correction coefficient.

[0068] Different frequency bands of power spectral density reflect the morphological characteristics of gear surfaces at different scales. Low-frequency bands correspond to large-scale undulations, mainly related to the screening process; high-frequency bands correspond to microscopic defects such as tiny scratches, mainly related to the grinding process; and overall changes across the entire frequency band reflect the uniformity of material properties, mainly related to the heating process. This correspondence provides a basis for subsequent responsibility allocation. Based on the test data of qualified gears, normal energy distribution ranges for different frequency bands are set as a benchmark for judging whether the current surface quality is qualified. By comparing the current gear's power spectral density with the standard range, the energy deviations of each frequency band are identified, clarifying which processes may have problems reflected by these deviations. Based on the proportion of each frequency band deviation in the total deviation, combined with the influence weight of different processes on surface quality, the contribution of the screening, grinding, and heating processes to surface quality issues is determined. The total correction coefficient is then decomposed according to the contribution, which is a quantitative representation of the comprehensive deviation of each process. Based on the calculated contribution of each process, the total correction coefficient is decomposed into sub-correction coefficients corresponding to each process, thus clarifying the specific degree of adjustment required for each process.

[0069] Step S3: Correct the total correction coefficient threshold using the sub-correction coefficients;

[0070] Step S4: Sort the correction order according to the correction results, and process according to the sorting results.

[0071] Step S1 includes the following specific contents:

[0072] Calculate the independent coefficients of the surface roughness of the gear, denoted as... ;

[0073] The calculation formula is:

[0074] ;

[0075] in, The root mean square of the height undulation of the rough surface. This represents the distance between two points on a rough surface. Indicates the overall length of the rough surface. Represents the natural constant;

[0076] The initial surface roughness power spectral density of the gear is calculated using the following formula:

[0077] ;

[0078] Combining the above two equations, we obtain the final surface roughness power spectral density of the gear surface, calculated as follows:

[0079] .

[0080] Image recognition algorithms are used to extract feature values ​​of the micro-morphology of the gear surface from the captured images, including information such as the height of the surface concave and convex structures and the spacing between adjacent concave and convex points.

[0081] By performing integral operations, the morphological characteristics of the rough surface in the spatial domain are transformed into the wavenumber domain to reflect the energy distribution law of the rough components at different spatial frequencies. The final power spectral density quantitatively describes the roughness distribution of the gear surface at different spatial frequencies, such as the frequencies corresponding to large-scale undulations and micro-scratches. It can accurately reflect the influence of processes such as screening, grinding, and heating on the surface morphology.

[0082] Step S2 includes the following specific contents:

[0083] The continuous aperture size sequence acquired by the laser diameter measurement system is divided into equal parts. The data segment is used to calculate the... Median value of segment aperture size If the aperture size deviates from the specified value for three consecutive sampling points within a certain time period... Exceeding the standard deviation of this segment If so, then mark that time period as the data segment for filtering out interference;

[0084] The grinding wheel feed rate and vibration amplitude of the grinding machine are collected, multiplied to obtain the grinding energy sequence, and then divided into equal parts. The data segment is used to calculate the... Median value of grinding energy If, within a certain time period, the grinding energy is higher than the standard value for five consecutive sampling points... If it is 1.2 times that, it is marked as a data segment for grinding interference;

[0085] The temperature curves of the multi-zone heat treatment furnace are divided into equal parts according to time. The data segment; calculate the first data segment. Median temperature of the segment If the temperature deviates from the normal range for 3 consecutive minutes within a certain period of time... If the temperature exceeds ±10℃, it is marked as a temperature interference data segment;

[0086] For the screening interference data segment, statistically analyze the relationship between each aperture size and the screening interference data segment. The absolute values ​​of the deviations are summed and then divided by the number of data points to obtain the mean deviation per unit point, which is... ;

[0087] For the grinding interference data segment, calculate the grinding energy and its relationship within each grinding interference data segment. The percentage deviation is calculated using the following formula:

[0088] The average of all deviation percentages is... ;

[0089] For the temperature interference data segment, take the temperature within the temperature interference data segment and... The maximum value of the absolute value of the deviation is... .

[0090] Centered on the median time of the data segment used to filter out interference. ;turn up The most recent non-interference data segment, denoted by its median time as And use this non-interference segment as the reference segment;

[0091] extract arrive The aperture data between the two sections are used to calculate the size change trend rate as the trend intensity through linear fitting; the aperture data of the interference section and the reference section are sorted by time and the difference is calculated point by point, and then multiplied by the trend intensity to obtain the screening stability offset, which is then used to form the screening stability offset sequence. Similarly, the grinding stability offset sequence and the temperature stability offset sequence are generated.

[0092] If you grind down the interfering data segments If the intensity of the grinding interference is greater than the average of all grinding interference segments, and the grinding energy within the segment exhibits an oscillating characteristic of first rising, then falling, and then rising again, it is determined to be external oscillation interference. Using the difference between the maximum and minimum vibration amplitudes within the segment as a benchmark, the ratio of the vibration amplitude at each sampling point to the benchmark value is calculated, multiplied by a preset oscillation coefficient, to obtain the grinding oscillation offset, forming a grinding oscillation offset sequence.

[0093] If the disturbance is stable, the stable offset sequence of screening is added point by point to the original aperture data of the disturbance section of screening. If the disturbance is oscillating, the oscillating offset sequence of grinding is superimposed to obtain the corrected screening aperture data. Similarly, the feed rate and vibration amplitude data of the disturbance section of grinding are corrected point by point using the stable offset sequence of grinding or the oscillating offset sequence of grinding to obtain the corrected grinding accuracy data. The temperature curve of the disturbance section of temperature is corrected point by point using the stable offset sequence of temperature or the oscillating offset sequence of temperature to obtain the corrected heating temperature data.

[0094] The corrected screening aperture data, corrected grinding accuracy data, and corrected heating temperature data were normalized, and the normalized data were weighted and summed according to their weights to reflect the comprehensive correction requirements of the three types of process parameters on the rough surface of the gear.

[0095] The deviation rate between the corrected sieve aperture and the standard aperture, the conformity of the grinding accuracy with the standard grinding parameters, and the matching degree of the heating temperature with the standard temperature curve were calculated separately. Based on the weights of the three types of data, the sieve correction coefficient, grinding correction coefficient, and heating correction coefficient were calculated.

[0096] Screening uses 3 points deviating from the median value, grinding uses 5 points with energy exceeding 1.2 times the median value, and temperature uses a 3-minute deviation of ±10℃ from the median value to accurately distinguish between normal process fluctuations and abnormal interference, avoiding misjudgment or omission. Correction sequences for stable and oscillating interference are designed to bring the interference data back to the ideal process state, ensuring the accuracy of production line optimization decisions. The stable offset sequence is based on trend fitting of normal process segments, while the oscillating offset sequence is used to quantitatively cancel out periodic oscillations.

[0097] By combining interference intensity quantification, data correction, and correction coefficient calculation for the three independent processes of screening, grinding, and heating, the total correction coefficient integrates the comprehensive impact of the three processes on the rough surface, allowing production personnel to grasp the overall correction needs. The individual correction coefficients, namely the correction coefficients for screening, grinding, and heating, precisely break down the deviations and contributions of each process, transforming process adjustment needs into quantifiable numerical indicators. This provides a basis for prioritizing subsequent processes and allows gear production line optimization to shift towards data and algorithm-driven approaches, significantly improving the operability of process optimization.

[0098] Step S3 includes the following specific contents:

[0099] From the historical production database, select consecutive qualified batches whose power spectral density of gear surface roughness completely meets the preset quality standard within the past 3 months, and extract the total correction coefficient corresponding to these batches.

[0100] Remove extreme values ​​caused by sudden equipment failure or abnormal raw material batches, and take the arithmetic mean of the remaining data as the initial threshold baseline for the total correction coefficient. This baseline reflects the basic correction requirement level when the gear surface quality is stable and qualified.

[0101] The process deviation analysis corresponding to the correction coefficient is conducted for the three processes of screening, grinding, and heating. The deviation characteristics between the current process parameters and the standard parameters are analyzed independently based on their respective correction coefficients.

[0102] Using the sieving correction coefficient as a reference, compare the actual measured value of the corrected sieve aperture with the tolerance range of the standard aperture, and count the frequency of aperture values ​​exceeding the tolerance centerline. The higher the frequency, the more significant the deviation of the sieving parameters.

[0103] Based on the grinding correction coefficient, the degree of difference between the corrected grinding accuracy parameters and the standard parameters is analyzed. The number of surface scratches and protrusions is recorded through a visual inspection system. The larger the ratio of the number of defects to the grinding correction coefficient, the stronger the necessity for grinding process correction.

[0104] By referring to the heating correction factor, the temperature difference between the corrected temperature curve and the standard temperature curve at key nodes is compared, and the cumulative sum of the differences is calculated. The larger the sum, the more prominent the contribution of the heating parameters to the total correction requirement.

[0105] Multiply the frequency of aperture deviation by the preset aperture sensitivity coefficient to obtain the correction amount of the total threshold for the screening process; multiply the ratio of the number of surface defects to the grinding correction coefficient by the roughness influence coefficient to obtain the threshold correction amount for the grinding process; multiply the cumulative sum of temperature differences by the temperature aging coefficient to obtain the threshold correction amount for the heating process.

[0106] The aperture sensitivity coefficient is used in the screening process to reflect the sensitivity of gear material hardness and aperture machining accuracy to aperture deviations. The frequency of aperture deviations is converted into a correction amount for the total threshold, ensuring the impact of screening deviations aligns with the total threshold calculation. The roughness influence coefficient is used in the grinding process to reflect the influence of grinding tool characteristics on the correlation between surface defects and correction requirements. The ratio of the number of surface defects to the grinding correction coefficient is converted into a correction amount for the total threshold, amplifying or reducing the weight of grinding defects on the total correction requirement. It is mainly related to the grinding wheel grit; surfaces ground with fine-grained grinding wheels are more sensitive to defects, with a coefficient of 0.9-1.1; for coarse-grained grinding wheels, it is 0.5-0.7. The coefficient is also referenced to the quality standards for gear surface roughness; the higher the requirement, the larger the coefficient. The temperature aging coefficient is used in the heating process to characterize the influence of heating time on the cumulative effect of temperature deviations. The cumulative sum of temperature differences is converted into a correction amount for the total threshold, reflecting the actual impact of temperature deviations under different heating times. Longer heating times result in a greater cumulative impact of temperature deviations on gear performance. The coefficient is mainly determined based on the duration of the heating process. For long-duration heating processes, the coefficient is 0.8-1.0; for short-duration heating processes, the coefficient is 0.4-0.6. In addition, considering the criticality of the gear heat treatment process, the quenching and tempering process is sensitive to temperature, so the coefficient is appropriately increased.

[0107] By combining the process characteristics of each process with the gear quality requirements, the deviation indicators of different dimensions are uniformly quantified into the correction amount for the total correction coefficient threshold.

[0108] The initial threshold baseline is directly added to the correction amount of the above three processes to obtain the updated total correction coefficient threshold. The effective range is set according to the quality grade of gear production. The threshold range for precision gears is [0.2, 0.7], and for ordinary gears it is [0.1, 0.8].

[0109] If the updated threshold is lower than the lower limit of the interval, it will be automatically adjusted to the lower limit value; if it is higher than the upper limit, it will be adjusted to the upper limit value. When the real-time total correction coefficient exceeds the threshold, the parameter adjustment process of the corresponding process will be triggered.

[0110] The frequency of aperture deviations in the screening process is a count-type indicator. Its impact on the total threshold needs to be combined with the gear precision level and differentiated weighted by an aperture sensitivity coefficient. The coefficient value is higher for high-precision gears, so that even a small deviation can significantly drive the threshold adjustment, avoiding assembly problems caused by the accumulation of small aperture deviations. The ratio of the number of surface defects to the correction coefficient in the grinding process is a relative indicator, while the roughness influence coefficient is dynamically adjusted according to the grinding wheel grit size to ensure that the impact of micro-defects such as surface scratches on the total threshold matches the actual quality requirements. The cumulative sum of temperature differences in the heating process is an absolute indicator, while the temperature aging coefficient is related to the heating time. This allows the correction amount of temperature fluctuations to the total threshold to accurately reflect its potential impact on the gear metallographic structure, so that the correction of the total threshold is no longer a simple numerical accumulation, but a precise mapping of the real impact of each process.

[0111] Three coefficients, through preset rules, allow the total threshold adjustment to flexibly adapt to different gear types and production conditions. This provides a traceable and objective basis for threshold adjustment, significantly improving the reliability of quality control. The coefficient values ​​are dynamically adjusted based on gear quality grade, material characteristics, and tool parameters. When producing precision gears, the bore sensitivity coefficient and roughness influence coefficient are appropriately increased to ensure strict control over minute deviations. When producing ordinary gears, the coefficients are appropriately decreased to reduce unnecessary adjustment frequency while ensuring quality. This allows the same optimization logic to adapt to diverse production needs, avoiding resource waste or quality risks. The coefficient settings are based on historical production data and process theory. When the total threshold triggers an adjustment, production personnel can clearly identify key processes affecting quality by tracing the coefficient values ​​and deviation data.

[0112] Step S4 includes the following specific contents:

[0113] When a correction order trigger instruction is received, it is first integrated to form a set of correction tasks to be processed, including historical incomplete correction tasks and their corresponding existing order sequences, as well as correction tasks newly added in the current production process.

[0114] When determining the sorting priority of newly added correction tasks, it is necessary to first retrieve the historical correction task database, extract historical task data of the same process type as the new task, and classify the screening process, grinding process, and heating process into different types. The focus is on statistically analyzing the average execution cycle of historical correction tasks of the same type and the trigger frequency of historical correction tasks of the same type within the same production cycle, and then further calculate and determine the sorting priority of each new correction task.

[0115] When calculating the average execution cycle of historical correction tasks of the same type, it is necessary to first obtain the actual execution time of a single correction task of the same type in the historical database, and then eliminate the influence of extreme abnormal durations by weighted averaging to finally obtain the standard average execution cycle of this type of task.

[0116] When determining the priority of new correction tasks by combining historical data, the comprehensive priority index of each process type correction task (screening, grinding, heating) is first calculated based on the average execution cycle and historical trigger frequency of similar tasks in the past. Then, according to the value of the comprehensive priority index, a corresponding priority level is set for each type of process correction task. Finally, the corresponding priority level is matched according to the process type to which the new correction task belongs.

[0117] When calculating the comprehensive priority index of each process type of correction task, a weighted calculation method is adopted: multiply the average execution cycle of the task type by the first weight factor to obtain the first priority component; multiply the historical trigger frequency of the task type by the second weight factor to obtain the second priority component; and then add the two priority components to finally obtain the comprehensive priority index of the task type.

[0118] When sorting the newly added correction tasks to generate the corresponding sorting results, the tasks are first sorted from high to low according to their sorting priority to form an initial candidate sorting sequence. Then, the initial candidate sorting sequence is cleaned up by merging duplicate tasks. If there are duplicate correction task entries for the same process in the sequence, the one with the highest priority is retained and the other duplicate entries are deleted to finally obtain the sorting result of the newly added correction tasks.

[0119] The existing sorting sequences of historical unfinished correction tasks are integrated and merged with the sorting results of newly added correction tasks to form the final target correction task sorting sequence; at the same time, the original historical task sorting sequence is replaced by this target sorting sequence as the basis for the execution of subsequent correction tasks.

[0120] The average execution cycle of similar historical tasks is weighted to eliminate extreme values, reflecting the typical time consumption for corrections in different processes. Grinding, which requires wheel replacement, may take longer, resulting in a longer average cycle than screening. Combined with the first weighting factor, this ensures that short-duration, easily executed tasks are prioritized, reducing production line downtime. Historical trigger frequency reflects the failure tendency of processes. Heating processes, due to their sensitive temperature control systems, are more prone to frequent corrections. Combined with the second weighting factor, processes prone to problems receive priority attention. This approach ensures that urgent corrections for critical processes are not overlooked, while also preventing efficiency losses due to low-priority tasks consuming resources.

[0121] By merging historical and new tasks and deduplicating repetitive tasks, the system achieves high efficiency in task scheduling and continuity in the production process. The integration of existing sorting sequences of historically incomplete tasks with the sorting results of new tasks ensures that important historical tasks are not shelved. If there are historically unexecuted heating process correction tasks and a new grinding correction task is added, the system will sort them based on a combined priority, ensuring that important historical tasks are not delayed. Duplicate tasks are merged and cleaned up, retaining the highest priority items. If three consecutive screening process corrections are triggered within 10 minutes, the system will merge them into a single highest priority task, avoiding parameter fluctuations and equipment wear caused by repeated adjustments. Real-time replacement of historical sequences with the target sorting sequence ensures that task scheduling is always based on the latest correction requirements, guaranteeing that the production line can dynamically respond to process changes. This reduces ineffective labor and improves the production line's ability to respond quickly to quality fluctuations.

[0122] Figure 2 An intelligent algorithm-based optimization control system for a gear production line is presented to implement any of the control methods described above. The system includes a recording and calculation module, a comparison and analysis module, a correction module, and a sorting module; the modules are connected by signals.

[0123] The camera and calculation module uses an industrial camera to record the surface of the gear after production, and calculates the power spectral density of the gear surface roughness after obtaining the feature values.

[0124] The comparative analysis module is used to obtain screening aperture data, grinding accuracy data, and heating temperature data. After combining them, the total correction coefficient is obtained. The calculated power spectral density of the gear surface roughness is compared with the total correction coefficient to obtain the sub-correction coefficient.

[0125] The correction module uses sub-correction coefficients to correct the total correction coefficient threshold;

[0126] The sorting module sorts the correction order according to the correction results and processes the results accordingly.

[0127] The above embodiments can be implemented, in whole or in part, by software, hardware, firmware, or any other combination thereof. When implemented using software, the above embodiments can be implemented, in whole or in part, as a computer program product. The computer program product includes one or more computer instructions or computer programs. When the computer instructions or computer programs are loaded or executed on a computer, all or part of the processes or functions described in the embodiments of this application are generated. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, the computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via wired (e.g., infrared, wireless, microwave, etc.) means. The computer-readable storage medium can be any available medium that a computer can access or a data storage device such as a server or data center that includes one or more sets of available media. The available medium can be a magnetic medium (e.g., floppy disk, hard disk, magnetic tape), an optical medium (e.g., DVD), or a semiconductor medium. The semiconductor medium can be a solid-state drive.

[0128] Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0129] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working processes of the systems, devices, and units described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here.

[0130] In the several embodiments provided in this application, it should be understood that the disclosed systems, apparatuses, and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection between apparatuses or units may be electrical, mechanical, or other forms.

[0131] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0132] In addition, the functional units in the various embodiments of this application can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit.

[0133] If the aforementioned functions are implemented as software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or a portion of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this application. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0134] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

[0135] In conclusion, the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A gear production line optimization control method based on intelligent algorithms, characterized in that, Includes the following steps: Step S1: Use an industrial camera to record the surface of the gear after production, obtain feature values, and then calculate the power spectral density of the gear surface roughness. Step S2: Obtain screening aperture data, grinding accuracy data, and heating temperature data. Combine these to obtain the total correction coefficient. Compare the calculated power spectral density of the gear surface roughness with the total correction coefficient to obtain the sub-correction coefficients. Different frequency bands of the power spectral density reflect the morphological characteristics of the gear surface at different scales. The low-frequency band corresponds to large-scale undulations and is related to the screening process; the high-frequency band corresponds to microscopic defects such as tiny scratches and is related to the grinding process; the overall change across the entire frequency band reflects the uniformity of material properties and is related to the heating process. Compare the current power spectral density of the gear with the standard range to find the energy deviation of each frequency band. Clarify which processes these deviations reflect potential problems. Based on the proportion of each frequency band deviation in the total deviation, combined with the influence weight of different processes on surface quality, determine the contribution of each of the screening, grinding, and heating processes to surface quality issues. Decompose the total correction coefficient according to the contribution. The total correction coefficient is a quantitative representation of the comprehensive deviation of each process. According to the contribution of each process calculated above, the total correction coefficient is decomposed into sub-correction coefficients corresponding to each process, thereby clarifying the specific degree of adjustment required for each process. Step S3: Correct the total correction coefficient threshold using the sub-correction coefficients; Step S4: Sort the correction order according to the correction results, and process according to the sorting results.

2. The gear production line optimization control method based on intelligent algorithms according to claim 1, characterized in that: Step S1 includes the following specific contents: Calculate the independent coefficients of the surface roughness of the gear, denoted as... ; The calculation formula is: ； in, The root mean square of the height undulation of the rough surface. This represents the distance between two points on a rough surface. Indicates the overall length of the rough surface. Represents the natural constant; The initial surface roughness power spectral density of the gear is calculated using the following formula: ； Combining the above two equations, we obtain the final surface roughness power spectral density of the gear surface, calculated as follows: 。 3. The gear production line optimization control method based on intelligent algorithms according to claim 2, characterized in that: Step S2 includes the following specific contents: The continuous aperture size sequence acquired by the laser diameter measurement system is divided into equal parts. The data segment is used to calculate the... Median value of segment aperture size ; If the aperture size deviates from the specified value for three consecutive sampling points within a certain time period... Exceeding the standard deviation of this segment If so, then mark that time period as the data segment for filtering out interference; The grinding wheel feed rate and vibration amplitude of the grinding machine are collected, multiplied to obtain the grinding energy sequence, and then divided into equal parts. The data segment is used to calculate the... Median value of grinding energy If, within a certain time period, the grinding energy is higher than the standard value for five consecutive sampling points... If it is 1.2 times that, it is marked as a data segment for grinding interference; The temperature curves of the multi-zone heat treatment furnace are divided into equal parts according to time. The data segment; calculate the first data segment. Median temperature of the segment If the temperature deviates from the normal range for 3 consecutive minutes within a certain period of time... If the temperature exceeds ±10℃, it is marked as a temperature interference data segment; For the screening interference data segment, statistically analyze the relationship between each aperture size and the screening interference data segment. The absolute values ​​of the deviations are summed and then divided by the number of data points to obtain the mean deviation per unit point, which is... ; For the grinding interference data segment, calculate the grinding energy and its relationship within each grinding interference data segment. The percentage deviation is calculated using the following formula: The average of all deviation percentages is... ; For the temperature interference data segment, take the temperature within the temperature interference data segment and... The maximum value of the absolute value of the deviation is... .

4. The gear production line optimization control method based on intelligent algorithms according to claim 3, characterized in that: Centered on the median time of the data segment used to filter out interference. ;turn up The most recent non-interference data segment, denoted by its median time as And use this non-interference segment as the reference segment; extract arrive The aperture data between the two sections are used to calculate the size change trend rate as the trend intensity through linear fitting; the aperture data of the interference section and the reference section are sorted by time and the difference is calculated point by point, and then multiplied by the trend intensity to obtain the screening stability offset, which is then used to form the screening stability offset sequence. Similarly, the grinding stability offset sequence and the temperature stability offset sequence are generated. If you grind down the interfering data segments If the intensity of the grinding interference is greater than the average of all grinding interference segments, and the grinding energy within the segment exhibits an oscillating characteristic of first rising, then falling, and then rising again, it is determined to be external oscillation interference. Using the difference between the maximum and minimum vibration amplitudes within the segment as a benchmark, the ratio of the vibration amplitude at each sampling point to the benchmark value is calculated, multiplied by a preset oscillation coefficient, to obtain the grinding oscillation offset, forming a grinding oscillation offset sequence. If it is a stationary disturbance, the stationary offset sequence of the screening is added point by point to the original aperture data of the screening disturbance segment; If the interference is oscillation, the grinding oscillation offset sequence is superimposed to obtain the corrected screening aperture data; similarly, the feed rate and vibration amplitude data of the grinding interference section are corrected point by point using the grinding stable offset sequence or grinding oscillation offset sequence to obtain the corrected grinding accuracy data; the temperature curve of the temperature interference section is corrected point by point using the temperature stable offset sequence or temperature oscillation offset sequence to obtain the corrected heating temperature data. The corrected screening aperture data, corrected grinding accuracy data, and corrected heating temperature data were normalized, and the normalized data were weighted and summed according to their weights to reflect the comprehensive correction requirements of the three types of process parameters on the rough surface of the gear. The deviation rate between the corrected sieve aperture and the standard aperture, the conformity of the grinding accuracy with the standard grinding parameters, and the matching degree of the heating temperature with the standard temperature curve were calculated separately. Based on the weights of the three types of data, the sieve correction coefficient, grinding correction coefficient, and heating correction coefficient were calculated.

5. The gear production line optimization control method based on intelligent algorithms according to claim 4, characterized in that: Step S3 includes the following specific contents: From the historical production database, select consecutive qualified batches whose power spectral density of gear surface roughness completely meets the preset quality standard within the past 3 months, and extract the total correction coefficient corresponding to these batches. Remove extreme values ​​caused by sudden equipment failure or abnormal raw material batches, and take the arithmetic mean of the remaining data as the initial threshold baseline for the total correction coefficient. This baseline reflects the basic correction requirement level when the gear surface quality is stable and qualified. The process deviation analysis corresponding to the correction coefficient is conducted for the three processes of screening, grinding, and heating. The deviation characteristics between the current process parameters and the standard parameters are analyzed independently based on their respective correction coefficients. Using the sieving correction coefficient as a reference, the actual measured value of the corrected sieve aperture is compared with the tolerance range of the standard aperture, and the frequency of aperture values ​​exceeding the tolerance centerline is statistically analyzed. Based on the grinding correction coefficient, analyze the degree of difference between the corrected grinding accuracy parameters and the standard parameters, and record the number of surface scratches and raised defects through a visual inspection system; Using the heating correction factor as a reference, compare the temperature difference between the corrected temperature curve and the standard temperature curve at key nodes, and calculate the cumulative sum of the differences.

6. The gear production line optimization control method based on intelligent algorithms according to claim 5, characterized in that: The initial threshold baseline is directly added to the correction amount of the above three processes to obtain the updated total correction coefficient threshold. The effective range is set according to the quality grade of gear production. The threshold range for precision gears is [0.2, 0.7], and for ordinary gears it is [0.1, 0.8]. If the updated threshold is lower than the lower limit of the interval, it will be automatically adjusted to the lower limit value; if it is higher than the upper limit, it will be adjusted to the upper limit value. When the real-time total correction coefficient exceeds the threshold, the parameter adjustment process of the corresponding process will be triggered.

7. The gear production line optimization control method based on intelligent algorithms according to claim 6, characterized in that: Step S4 includes the following specific contents: When a correction order trigger instruction is received, it is first integrated to form a set of correction tasks to be processed, including historical incomplete correction tasks and their corresponding existing order sequences, as well as correction tasks newly added in the current production process. When determining the sorting priority of newly added correction tasks, it is necessary to first retrieve the historical correction task database, extract historical task data of the same process type as the new task, and classify the screening process, grinding process, and heating process into different types. The focus is on statistically analyzing the average execution cycle of historical correction tasks of the same type and the trigger frequency of historical correction tasks of the same type within the same production cycle, and then further calculate and determine the sorting priority of each new correction task. When calculating the average execution cycle of historical correction tasks of the same type, it is necessary to first obtain the actual execution time of a single correction task of the same type in the historical database, and then eliminate the influence of extreme abnormal durations by weighted averaging to finally obtain the standard average execution cycle of this type of task.

8. The gear production line optimization control method based on intelligent algorithms according to claim 7, characterized in that: When determining the priority of new correction tasks by combining historical data, the comprehensive priority index of each process type correction task is first calculated based on the average execution cycle and historical trigger frequency of similar tasks in the past. Then, according to the value of the comprehensive priority index, a corresponding priority level is set for each type of process correction task. Finally, the corresponding priority level is matched according to the process type to which the new correction task belongs. When calculating the comprehensive priority index of each process type of correction task, a weighted calculation method is adopted: multiply the average execution cycle of the task type by the first weight factor to obtain the first priority component; multiply the historical trigger frequency of the task type by the second weight factor to obtain the second priority component; and then add the two priority components to finally obtain the comprehensive priority index of the task type.

9. The gear production line optimization control method based on intelligent algorithms according to claim 8, characterized in that: When sorting the newly added correction tasks to generate the corresponding sorting results, the tasks are first sorted from high to low according to their sorting priority to form an initial candidate sorting sequence. Then, the initial candidate sorting sequence is cleaned up by merging duplicate tasks. If there are duplicate correction task entries for the same process in the sequence, the one with the highest priority is retained and the other duplicate entries are deleted to finally obtain the sorting result of the newly added correction tasks. The existing sorting sequences of historical tasks that have not been completed are integrated and merged with the sorting results of newly added tasks to form the final target task sorting sequence; at the same time, the original historical task sorting sequence is replaced with this target sorting sequence.

10. A gear production line optimization control system based on intelligent algorithms, used to implement the control method according to any one of claims 1-9, comprising a recording and calculation module, a comparison and analysis module, a correction module, and a sorting module; the modules are connected by signals. The camera and calculation module uses an industrial camera to record the surface of the gear after production, and calculates the power spectral density of the gear surface roughness after obtaining the feature values. The comparative analysis module is used to obtain screening aperture data, grinding accuracy data, and heating temperature data. After combining them, the total correction coefficient is obtained. The calculated power spectral density of the gear surface roughness is compared with the total correction coefficient to obtain the sub-correction coefficient. The correction module uses sub-correction coefficients to correct the total correction coefficient threshold; The sorting module sorts the correction order according to the correction results and processes the results accordingly.

Images