Intelligent matching method for gantry assembly gasket

Abstract

The application discloses a portal frame assembly gasket intelligent matching method and relates to the technical field of forklift manufacturing. The method comprises the following steps: collecting the opening size data of outer portal frames and inner portal frames along the full stroke of the portal frames; performing outlier rejection cleaning on the collected data; performing linear fitting on the cleaned data by using the least square method to obtain the opening size change trend straight line of the outer portal frames and the inner portal frames along the stroke direction; calculating the actual assembly gap according to the fitted straight line, and combining the standard gap to calculate the required compensation gasket thickness; mapping the calculated continuous thickness value to discrete gasket specifications based on the gasket specification library, taking the arithmetic mean of adjacent specifications as the threshold; and outputting the matching result. The application realizes one-time accurate matching of the forklift portal frame assembly gasket, eliminates the low-efficiency process of traditional repeated disassembly and trial and error, and improves the portal frame assembly efficiency and quality consistency.

Description

Technical Field

[0001] This invention relates to the field of forklift manufacturing technology, specifically to a method for selecting and matching shims during the assembly process of a forklift mast, and in particular to an intelligent method for selecting and matching mast assembly shims based on full-stroke data acquisition and statistical fitting. Background Technology

[0002] The forklift mast is the core load-bearing and guiding component of a forklift, typically composed of an outer mast, a middle mast, and an inner mast nested in layers, with rollers enabling relative sliding between the masts. The quality of the mast assembly directly affects the forklift's lifting system's maneuverability, stability, and safety. Among these challenges, controlling the lateral assembly clearance between the inner and outer masts is crucial: excessive clearance leads to severe swaying of the mast at high positions, affecting cargo handling accuracy and operational safety; insufficient clearance causes the inner mast to jam during full-range pulling within the outer mast, and in severe cases, even results in abnormal wear between the rollers and the mast's channel steel.

[0003] Currently, adjusting the assembly clearance of forklift masts generally relies on manual experience. A typical process is as follows: First, the outer and inner masts are trial-assembled. The inner mast is manually pressed against one side of the outer mast, and a feeler gauge is used to measure the clearance on the other side. Then, the inner mast is removed from the outer mast, and the measured clearance value is converted into the thickness of the shims on both sides, which are then installed below the mast roller shaft. The inner mast is then reassembled into the outer mast, and the process is checked during the full stroke of the inner mast. If it encounters resistance and cannot be pulled into place, the inner mast needs to be removed and the shim thickness reduced. If the inner mast wobbles beyond tolerance at a height of 500mm above the outer mast, the inner mast needs to be removed and the shim thickness increased. This process often requires repeated disassembly, trial and error, and further disassembly and reassembly, severely impacting mast assembly efficiency.

[0004] The shortcomings of traditional methods are summarized as follows:

[0005] First, the selection of measurement points is highly arbitrary and easily affected by factors such as the flatness of the gantry surface and the feel of manual operation.

[0006] Second, the available shim thicknesses are usually discrete specification levels (such as every 0.05mm or 0.1mm), while the compensation clearance required for actual gantry assembly is a continuously changing value. Therefore, one can only rely on experience to "fit together" a close shim specification, which lacks precision.

[0007] Third, the data from the current gantry calibration measurement, painting transfer, and assembly shim selection processes are independent of each other. The measurement data cannot be effectively transferred to the assembly process to guide shim selection, resulting in a large amount of valuable dimensional data being wasted.

[0008] Automatic shim matching methods in other fields are mainly for specific small precision components. Their measurement objects, accuracy requirements, and shim specification systems are fundamentally different from those of large welded structural components of forklift masts, and cannot be directly applied. In addition, related solutions in the forklift mast field are mostly focused on improving the mast structure itself or simple shim adjustment structures, and have not yet involved systematic intelligent shim matching. That is to say, at present, there is still a lack of shim matching solutions specifically for forklift mast assembly scenarios and based on full-stroke statistical data analysis. Summary of the Invention

[0009] This invention addresses the shortcomings of existing forklift mast assembly shim selection methods, which rely on manual experience, lack systematic mathematical decision-making methods, and are disconnected from the assembly process. It provides an intelligent shim selection method for mast assembly, aiming to obtain statistically significant mast dimensions through high-frequency data acquisition, data cleaning, and linear fitting of the mast's full-stroke opening dimensions. Then, through shim matching logic, the continuous calculation results are intelligently mapped to a discrete shim specification library to achieve one-time accurate shim selection.

[0010] The technical solution adopted in this invention is as follows:

[0011] This invention provides a method for intelligent selection and matching of gantry assembly gaskets, including:

[0012] Step S1: At the gantry calibration and measurement station, use photoelectric sensors to collect dense data along the roller guide groove direction (i.e., the entire stroke direction) of the outer and inner gantry. The collection frequency can be set according to actual needs, up to 10 times per second, covering the entire stroke range of the gantry. The key points for collection are the opening dimensions at the guide roller positions of the inner and outer gantry.

[0013] Each piece of raw data collected is transmitted to the backend system in real time and bound to the gantry material number information. Specifically, during gantry calibration, codes are affixed to both the outer and inner gantry, and a barcode scanning mechanism is used to associate the gantry opening dimensions with the gantry material number information. Considering that the codes need to be removed during the painting process, when the gantry is put into the painting line, the code information is transferred to the code carrier on the painting line hook by scanning. When the gantry is removed from the painting line, the data information on the code carrier is transferred to a QR code by a barcode scanner and affixed to the gantry, thus achieving complete data transfer from the calibration station to the painting line and then to the assembly station.

[0014] Step S2: The collected raw data needs to be filtered and screened to ensure that the samples involved in the calculation can truly reflect the overall shape of the gantry, rather than being misled by some surface processing defects or occasional sensor noise.

[0015] In at least one possible implementation, the raw data is first filtered according to pre-set upper and lower control limits, and data points exceeding or falling below the acceptable range are automatically removed, with an alarm displayed on the interface. For example, after removing 5 abnormal data points with large fluctuations from 100 collected data points, the remaining 95 valid data points are used for subsequent fitting calculations.

[0016] Step S3: After the first round of screening, a set of data within the acceptable range is obtained. Each data point is a sampling point (x... i , y i ), where x is the sequence number of the measurement (0, 1, 2, …), and y is the sensor measurement value corresponding to the sequence number.

[0017] Using the sequence of measurement times of the outer or inner gantry as the independent variable X and the corresponding sequence of opening size measurement values ​​as the dependent variable Y, the slope m and intercept b of the fitted straight line are calculated using the least squares method, resulting in the fitted straight line equation y = mx + b. This straight line represents the trend of the gantry opening size along the travel direction.

[0018] Then, the mean of the measurement sequence is substituted into the fitted linear equation to obtain the most statistically significant fitted opening size of the gantry under this working condition.

[0019] Step S4: Obtain the fitted opening size Outer of the outer gantry and the fitted opening size Inner of the inner gantry respectively, and calculate the actual assembly gap: Gap = Outer - Inner;

[0020] Then calculate the required pad thickness for compensation: RequiredThickness = TargetGap - Gap;

[0021] TargetGap represents the standard clearance required by the design. If RequiredThickness is positive, it means that shims need to be added to compensate for the clearance. If it is negative, it means that the clearance is too small and shims need to be reduced.

[0022] Step S5: Based on the calculated RequiredThickness, the most suitable gasket specification is automatically selected by comparing it with the currently available discrete gasket specification library; in at least one possible implementation, this invention provides three matching logics, which can be flexibly selected according to actual working conditions:

[0023] (1) Median mode: Automatically calculates the arithmetic mean of two adjacent gasket specifications and uses the median as the threshold for intelligent judgment. For example, there are two gasket specifications, 2.10mm and 2.15mm, and the median of the two is 2.125mm; if the required compensation thickness is greater than 2.125mm, the system will automatically select the 2.15mm gasket; if it is less than 2.125mm, the 2.10mm gasket will be selected.

[0024] (2) Interference mode: Suitable for scenarios with extremely high requirements for gantry stability. The system automatically selects a larger shim size, sacrificing some smoothness of sliding in exchange for higher stability at high position.

[0025] (3) Low-speed mode: Suitable for scenarios where the smoothness of the gantry sliding throughout the entire process is extremely important. The system automatically selects a smaller shim size, sacrificing some high-position stability in exchange for better overall passability.

[0026] Step S6: Finally, the following information can be displayed on the screen: a comparison chart of the actual measurement results and the acceptable range; recommended gasket specifications (including specifications, model, and quantity); alarm prompts for abnormal points, etc. Even better, all raw measurement data, fitting calculation results, gantry numbers, selection results, and other information are stored in the SPC database and timestamped. This allows for searching historical data by time, product model, and other criteria, and data can be exported as documents for convenient quality traceability and statistical analysis.

[0027] Compared with existing technologies, the main design concept of this invention is to eliminate the problem of insufficient representativeness of single-point measurements by using high-frequency acquisition throughout the entire stroke to cover the overall size and shape of the gantry; to use the least squares method to linearly fit the cleaning data, effectively filtering out reading errors caused by sensor noise and local defects, making the calculated dimensions statistically significant; and to use the arithmetic mean of adjacent specifications as the threshold for median decision-making, optimally mapping the continuous compensation amount to the discrete gasket specification library, eliminating the subjectivity of manual matching; in addition, the information transmission through labels, carriers, and QR codes enables the calibration data to be transmitted to the assembly station for guidance; and to provide three modes—median, interference, and low—to flexibly adapt to different working conditions; all data is stored in the SPC database, supporting traceability and statistical analysis, providing support for quality control. Attached Figure Description

[0028] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described below with reference to the accompanying drawings, wherein:

[0029] Figure 1 A flowchart illustrating the intelligent selection method for gantry assembly gaskets provided in an embodiment of the present invention;

[0030] Figure 2 This is a schematic diagram of the data cleaning and fitting fusion process provided in an embodiment of the present invention. Detailed Implementation

[0031] Embodiments of the present invention are described in detail below. Examples of these embodiments are shown in the accompanying drawings, wherein the same or similar reference numerals denote the same or similar elements or elements having the same or similar functions throughout. The embodiments described below with reference to the accompanying drawings are exemplary and are only used to explain the present invention, and should not be construed as limiting the present invention.

[0032] This invention proposes an embodiment of a method for intelligent selection and matching of gantry assembly gaskets, specifically, as follows: Figure 1 As shown, it includes:

[0033] Step S1: Full-stroke high-frequency data acquisition;

[0034] In the actual manufacturing process of forklift masts, after the mast structural components are welded, they need to go through a calibration and measurement station for dimensional inspection and calibration. In actual operation, photoelectric sensors (preferably laser displacement sensors) can be deployed at this station to collect full-stroke data on the opening dimensions at the guide roller positions of the outer and inner masts.

[0035] As a large welded structural component, the mast of a forklift does not have a constant, single opening dimension. Instead, it fluctuates to varying degrees along the mast height direction, i.e., along the entire sliding stroke of the inner mast relative to the outer mast. This fluctuation stems from the combined effects of multiple factors, including welding residual stress, channel steel rolling precision, and calibration processes. Traditional single-point measurement methods can only obtain the opening dimension at a specific location, and the selection of this location is often arbitrary. Single-point data cannot represent the true dimensional distribution of the mast throughout its entire stroke. Those skilled in the art, in long-term practice, are accustomed to using feeler gauges for single-point measurements at a typical location, unaware of the fluctuation pattern of the mast opening dimension along its entire stroke, the potential significant deviation between single-point measurements and the actual dimensional distribution throughout the stroke, and even less aware of the need for dense data collection throughout the entire stroke to capture the overall dimensional shape of the mast.

[0036] Specifically, in this step, the data acquisition frequency can be set to 1 to 10 times per second, and the acquisition range is the entire stroke from the lowest to the highest position of the inner mast. Taking a certain type of 3-ton forklift mast as an example, its entire stroke is about 3000mm. With a acquisition frequency of 10 times per second and the inner mast taking about 10 seconds to pass through the entire stroke at a constant speed, about 100 data points can be collected, which is sufficient to cover the dimensional changes of the mast throughout its entire stroke.

[0037] The key data collection points are the opening dimensions at the guide roller positions of the outer and inner gantry, namely the inner opening dimension of the outer gantry and the outer opening dimension of the inner gantry. These two dimensions directly determine the actual gap after the inner and outer gantry are assembled.

[0038] Then, each piece of raw measurement data collected needs to be bound to the gantry material number information to ensure that the data can be correctly retrieved in subsequent processes. The specific implementation method can be found below:

[0039] During mast calibration, barcodes (such as, but not limited to, barcodes or QR codes containing material number information) are affixed to both the outer and inner masts. A scanning mechanism is then used to associate the mast opening dimensions with the material number information. Considering that the barcodes applied during the painting process require removal due to the high-temperature baking process, which is specific to the manufacturing process of the forklift mast, the industry has traditionally treated calibration measurement data and the painted mast as independent processes, without considering establishing a data relay mechanism within the painting process. This embodiment proposes a preferred method: scanning the barcodes upon entering the painting line to transfer the barcode information to a carrier on the painting line hook, such as a rewritable electronic tag. This carrier moves with the hook throughout the painting line, enabling data relay in mid-air. When the painting is completed and the mast is removed from the line, a barcode scanner transfers the data from the carrier to a new QR code, which is then affixed to the mast. This avoids damage to the barcodes caused by the high temperatures of painting and achieves complete data flow from the calibration station to the painting line and then to the assembly station.

[0040] Step S2: Data cleaning and outlier removal;

[0041] During the high-frequency data acquisition process throughout the entire stroke, this invention considers that the sensor may generate occasional abnormal readings due to localized oxide scale, welding slag, oil stains, or other deposits on the gantry surface; the gantry surface may also have localized impacts, scratches, or other processing defects leading to localized dimensional anomalies. If these abnormal data points are directly included in subsequent fitting calculations, they will distort the fitting results, causing the calculated representative dimensions to deviate from the true shape of the gantry. Based on this analysis, this invention proposes establishing systematic data cleaning rules to remove outliers, thereby effectively improving the accuracy of the fitting results.

[0042] To elaborate, refer to Figure 2 As shown, the upper control limit (UCL) and lower control limit (LCL) can be pre-set according to the design tolerance requirements of the mast model. Taking a certain type of 3-ton forklift mast as an example, the design acceptable range of the inner opening dimension of its outer mast is 596.0mm to 602.0mm, so set UCL=602.0mm and LCL=596.0mm.

[0043] The system performs real-time judgment on each piece of raw data collected: if the data falls within the range of [LCL, UCL], it is retained as valid data; if it exceeds the range, it is automatically rejected and an alarm prompt is displayed on the system interface, marking the location number and measurement value of the rejected data point, so that the operator can confirm whether there is a sensor failure or obvious defect in the gantry, which requires manual intervention.

[0044] Based on this concept, it can also be pointed out that, in order to further improve the reliability of data cleaning, in some preferred embodiments of the present invention, a preferred data cleaning process is also provided: when the number of abnormal data points removed accounts for more than a preset threshold (such as 10%) of the total amount of collected data, the system can issue an early warning message indicating abnormal data quality, which is used to prompt the operator to check the sensor status and the condition of the gantry surface, and decide whether to re-collect data as needed, thereby avoiding the situation where a large amount of invalid data is mixed into the sample due to sensor failure or severe contamination of the gantry surface.

[0045] Step S3: Least squares linear fitting;

[0046] After data cleaning, a set of valid data points representing the distribution of mast opening dimensions along the travel direction were obtained. Due to various random factors in the mast manufacturing process, these data points are not strictly distributed on a straight line, but fluctuate around a certain trend line. In current forklift mast assembly scenarios, operators often directly take the average value as the representative value when faced with discrete measurement data, without realizing that the average value cannot reflect the changing trend of the mast opening dimensions along the travel, such as whether the mast has a "trumpet mouth" shape that is larger at the top and smaller at the bottom or vice versa. Furthermore, they did not realize that linear fitting can not only obtain statistically significant representative dimensions, but also determine the changing trend of the mast opening dimensions along the travel by the slope of the fitted line. This is of great value in determining whether the mast needs to be pre-calibrated and in evaluating the actual sliding performance after assembly.

[0047] Based on this, combined Figure 2 As illustrated, this invention proposes: assuming that after cleaning, n valid data points (x...) are obtained... i , y i ), where x i Let y be the measurement number (0, 1, 2, …, n-1). i These are the corresponding opening dimensions.

[0048] Using the sequence of measurement times for the outer or inner gantry as the independent variable X and the corresponding sequence of opening dimension measurements as the dependent variable Y, the slope m and intercept b of the fitted straight line are calculated using the least squares method. The sum S of the X sequences is then calculated. X Sum of squares of X sequence XX The sum of the Y sequences S YThe sum of the products of X and Y, S XY Furthermore, the slope m is calculated based on the least squares principle, with the aim of minimizing the sum of the squares of the perpendicular distances from all data points to the fitted line; the intercept b ensures that the fitted line passes through the centroid of the data points; in summary, the equation of the fitted line is obtained: y = mx + b.

[0049] The slope *m* reflects the rate of change of the gantry opening dimension along the travel direction: *m* > 0 indicates that the gantry opening dimension gradually increases with the travel direction; *m* < 0 indicates that it gradually decreases; the larger the absolute value of *m*, the more drastic the fluctuation of the gantry opening dimension along the travel direction. This slope information cannot be provided by the traditional averaging method, and it is of crucial guiding significance for determining whether there is a systematic deformation trend in the gantry structure and whether targeted correction is needed at the calibration station.

[0050] Continuing from the previous text, the mean X̄ of all X values ​​(i.e., the arithmetic mean of the measurement sequence numbers) is substituted into the fitted linear equation to obtain the most statistically significant fitted span size of the gantry at the midpoint of its travel. It should be noted that the reason for choosing the mean position instead of other positions is that the mean position represents the center point of data collection, and the fitted value at this position is least affected by the extreme values ​​at both ends, thus best representing the overall size level of the gantry.

[0051] To further ensure the reliability of the fitting results, the following optimization process can be referenced:

[0052] When the number of valid data n after cleaning is lower than the preset threshold (e.g., 30 data points), the system issues an alarm such as insufficient valid data, prompting the operator to check the collection process and collect data again, thereby ensuring that the sample size involved in the fitting has sufficient statistical significance.

[0053] When the absolute value of the fitted slope m exceeds the preset slope threshold (e.g., 0.05), the system issues a warning signal, such as an abnormal change in the gantry opening size along the travel direction, indicating that the gantry may have severe "flared" deformation and suggesting that correction be performed before proceeding to the shim selection process. Compared to the past, where jamming or shaking problems could only be detected during the full-stroke pull test after trial assembly, but by then the initial installation of the shims had already been completed and they needed to be removed and redone, the above-mentioned solution of this invention predicts the dimensional uniformity problem of the gantry itself through slope analysis before shim selection, providing the possibility for early intervention.

[0054] Step S4: Calculate the required thickness of the compensation shim;

[0055] As mentioned earlier, in this step, the fitted opening size Outer of the outer gantry and the fitted opening size Inner of the inner gantry can be obtained respectively. The difference between the two is the actual assembly gap Gap of the inner and outer gantry in the actual assembly state. The actual gap is compared with the standard gap TargetGap required by the design. The difference is the thickness that needs to be compensated by shims.

[0056] The formula for calculating the actual assembly gap is: Gap = Outer - Inner, and the formula for calculating the required compensation shim thickness is: RequiredThickness = TargetGap - Gap.

[0057] Where Outer is the fitted opening size of the outer gantry (i.e., the fitted value of the inner opening size of the outer gantry); Inner is the fitted opening size of the inner gantry (i.e., the fitted value of the outer opening size of the inner gantry); TargetGap is the standard clearance required by the design. If RequiredThickness is positive, it means that the actual clearance between the outer and inner gantry is less than the standard clearance, and shims need to be added under the roller shaft to compensate; if RequiredThickness is negative, it means that the actual clearance is greater than the standard clearance, and shims need to be reduced.

[0058] It is important to emphasize that in actual gantry assembly, the thickness of the shims on both sides should be consistent to ensure the alignment accuracy of the inner gantry within the outer gantry. Therefore, the calculated RequiredThickness represents the thickness adjustment required on both sides (half of the total compensation is already reflected in the calculation logic). Assemblers can simply install shims of the same thickness symmetrically on both sides according to the system's recommended specifications.

[0059] Step S5: Intelligent matching of gasket library;

[0060] This is one of the key aspects of this solution. As mentioned earlier, in actual production, the thickness specifications of gaskets are discrete. Standard parts are typically graded in increments of 0.05mm or 0.1mm. However, the RequiredThickness calculated in the above steps is a continuous value, which may fall between any two adjacent specifications, such as 2.127mm falling between 2.10mm and 2.15mm. Mapping the continuous calculation results to a discrete specification library is crucial for achieving accurate one-time matching. Previously, rounding down was the main solution for similar situations, but this approach was unstable because even small fluctuations in measurement or calculation could cause the matching result to "jump" between adjacent specifications, resulting in inconsistent gasket matching results for the same model of gantry in different batches of assembly. Therefore, this invention proposes to adopt median decision logic: the arithmetic mean of the specifications of two adjacent gaskets is used as the threshold for judgment, rather than simply comparing the absolute difference. It should be noted that although this logic of the present invention seems to have the same result as "rounding down" in most cases, there is an essential difference in the underlying principle and applicable scenarios: median decision is more suitable for situations where the measured value itself has statistical fluctuations, and can provide a more stable decision boundary.

[0061] Specifically, first obtain the currently available gasket specification library. Taking a gasket specification interval of 0.05mm as an example, the specification library is: ..., 2.00mm, 2.05mm, 2.10mm, 2.15mm, 2.20mm, ...

[0062] The system automatically determines which two adjacent specifications a calculated RequiredThickness falls between. For example, if RequiredThickness = 2.127mm, it will fall between 2.10mm and 2.15mm.

[0063] Calculate the arithmetic mean of these two adjacent dimensions: (2.10 + 2.15) / 2 = 2.125 mm.

[0064] Judgment rules: If RequiredThickness > median, select the larger specification (2.15mm); if RequiredThickness < median, select the smaller specification (2.10mm); if it is exactly equal to the median (extremely low probability), the system will default to selecting the smaller specification and prompt the operator for confirmation.

[0065] Furthermore, based on the above concept, this invention provides three matching modes, which can be flexibly selected according to actual working conditions:

[0066] (1) Median mode: Matching is performed according to the above median decision logic. This mode achieves the optimal mapping from continuous calculated values ​​to discrete specifications while ensuring assembly quality. It is a recommended mode applicable to most conventional working conditions.

[0067] (2) Interference mode: After making a median decision, the system automatically adjusts the result up by one specification level. For example, if the median mode recommends 2.10mm, the interference mode recommends 2.15mm. This mode is suitable for situations where the stability of the gantry at a high position is required. The slightly larger shim thickness makes the gap between the inner and outer gantry smaller, reducing the amplitude of high-position sway, but the sliding resistance will increase slightly throughout the process.

[0068] (3) Low-level mode: After completing the median decision, the system automatically adjusts the result down one specification level. For example, if the median mode recommends 2.10mm, the low-level mode recommends 2.05mm. This mode is suitable for high-frequency lifting operations where the smoothness of the gantry sliding throughout the entire process is extremely important. The slightly smaller shim thickness makes the gap between the inner and outer gantry larger, and the sliding throughout the entire process is smoother, but the high-level swaying amplitude will increase slightly.

[0069] Furthermore, regarding this step, it can be added that in actual production, the gasket specification library may not be strictly evenly distributed (e.g., some specifications may be temporarily out of stock due to supplier reasons). In this case, the present invention provides the following preferred process: Before making a median decision, first check if there is a specification in the gasket specification library that is exactly equal to RequiredThickness; if so, directly select that specification without performing a median judgment. This avoids unnecessary errors caused by rounding when a precisely matching specification exists. However, if the required specification is out of stock, a median decision can be automatically performed again among adjacent available specifications, and the output result can be marked with information such as "This specification is an alternative selection, confirmation is recommended."

[0070] Furthermore, it should be noted that while it is generally believed that finer shim specifications (smaller intervals) result in higher fitting accuracy, excessively fine intervals can actually introduce new problems in actual assembly: the more specifications there are, the more difficult it becomes for workers to identify and select shims, potentially increasing the misassembly rate. To address this, the present invention provides further optimization: the system can automatically recommend the optimal shim specification interval for this type of gantry based on statistical analysis of historical fitting data (e.g., adjusting from 0.05mm to 0.10mm), reducing the number of specifications while ensuring assembly accuracy and lowering the complexity of on-site management.

[0071] Step S6: Result Output and Data Traceability.

[0072] For example, the following information can be displayed to the operator on the screen:

[0073] (a) A comparison diagram of the actual measurement results and the acceptable range, showing the distribution of all valid data points in the form of a scatter plot, and overlaying the fitted line and the boundary of the acceptable range, so that the operator can intuitively understand the distribution of the gantry opening size;

[0074] (b) Recommended gasket specifications, including but not limited to specifications, model, required quantity (one on each side, two in total) and matching pattern;

[0075] (c) Anomaly alarm prompts, indicating the location and value of the removed abnormal data points for quality traceability reference.

[0076] Even better, all raw measurement data, fitting calculation results (including slope, intercept, fitted opening size, actual clearance, and required compensation thickness), gantry number, selection results, operator information, etc., are stored in the SPC database and timestamped. Furthermore, the system supports searching historical data by time range, product model, gantry number, and other criteria, and can export the data to Excel documents for convenient quality traceability and statistical analysis. Thus, through the long-term accumulation of data in the SPC database, process engineers can perform statistical analysis on a large amount of historical selection data for the same model of gantry, identifying systematic deviation trends in the gantry manufacturing process, such as an overall undersized opening size in a certain batch of gantry gantry, thereby guiding adjustments to upstream correction processes.

[0077] Finally, in conjunction with the various embodiments described above, this invention will use the mast assembly of a certain type of 3-ton forklift as an example to illustrate the complete implementation process of the above solution.

[0078] Both the outer and inner gantry are welded structural components. At the gantry calibration and measurement station, photoelectric sensors are used to collect full-stroke data on the guide roller positions of the outer and inner gantry. The collection frequency is set to 10 times per second. It takes about 10 seconds for the inner gantry to pass through the full stroke at a constant speed, and a total of about 100 data points are collected.

[0079] After the 100 raw data points were collected and transmitted to the backend system, they were first filtered for range. Based on the design tolerance requirements of this gantry model, the upper control limit was set to 602.0 mm and the lower control limit to 596.0 mm. The system automatically removed abnormal data points that exceeded these ranges. Assuming that 5 abnormal data points were removed (3 of which were due to occasional high values ​​caused by the sensor passing through the weld slag area, and 2 were due to low values ​​caused by local bumps on the gantry surface), 95 valid data points remained.

[0080] Then, a least-squares linear fitting was performed: with the measurement sequence number as X and the opening dimension measurement value as Y, the fitted straight line equation was calculated as y = 0.0281x + 597.0148. Substituting the mean of the measurement sequence X̄ = 50.684 into the equation, the fitted opening dimension of the inner gantry was obtained as 598.4395 mm.

[0081] Similarly, the measurement data of the outer gantry were collected, cleaned, and fitted in the same way, and the fitted opening size of the outer gantry was obtained as 600.5000mm.

[0082] Calculate the actual assembly clearance: Gap + 600.5000 - 598.4395 + 2.0605 mm.

[0083] Assuming the standard clearance TargetGap required for this model of gantry is 2.10mm, then the required thickness of the compensation shim is RequiredThickness = 2.10 - 2.0605 = 0.0395mm.

[0084] The available gasket sizes in the current library are in 0.05mm increments, namely 2.00mm, 2.05mm, 2.10mm, 2.15mm, 2.20mm... RequiredThickness=0.0395mm is close to 0.05mm, so the system determines that the actual required total gasket thickness should be around 2.10mm. Since the 2.10mm size is available in the gasket library, the system selects it directly.

[0085] Finally, the system outputs a recommended gasket specification of 2.10mm (one on each side) on the screen, and displays a comparison chart of the actual measurement results and the acceptable range, as well as information such as the location of all abnormal points; and all data, calculation results, gantry numbers, and selection results are stored in the SPC database.

[0086] The assembly personnel completed the shim installation in one go according to the shim specifications recommended by the system. Actual verification showed that the inner gantry moved smoothly throughout the entire movement within the outer gantry, and the high-position sway amplitude met design requirements, requiring no secondary adjustments.

[0087] As mentioned above, for the same model of mast, if the forklift is used for high-altitude operations, the user can select "interference mode," and the system will automatically recommend 2.15mm shims to sacrifice a little smoothness in exchange for higher high-position stability; conversely, if used for high-frequency lifting operations, the user can select "low mode," and the system will automatically recommend 2.05mm shims.

[0088] In this invention, when directional terms are mentioned, they are relative concepts based on the embodiments. Furthermore, "at least one" refers to one or more, and "more than one" refers to two or more. "And / or" describes the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent the existence of A alone, A and B simultaneously, or B alone. A and B can be singular or plural. The character " / " generally indicates that the preceding and following related objects have an "or" relationship. "At least one of the following" and similar expressions refer to any combination of these items, including any combination of single or plural items. For example, at least one of a, b, and c can represent: a, b, c, a and b, a and c, b and c, or a and b and c, where a, b, and c can be single or multiple.

[0089] The above description of the structure, features, and effects of the present invention is based on the embodiments shown in the figures. However, the above are only preferred embodiments of the present invention. It should be noted that the technical features involved in the above embodiments and their preferred methods can be reasonably combined and matched by those skilled in the art to form a variety of equivalent solutions without departing from or changing the design concept and technical effects of the present invention. Therefore, the present invention is not limited to the scope of implementation shown in the figures. Any changes made in accordance with the concept of the present invention, or modifications to equivalent embodiments, that do not exceed the spirit covered by the specification and figures, should be within the protection scope of the present invention.

Claims

1. A method for intelligent selection and matching of gantry assembly gaskets, characterized in that, include: At the gantry calibration and measurement station, the opening dimension data is collected along the entire stroke direction of the outer and inner gantry, and the collected data is bound to the gantry material number information; Outliers are removed from the collected raw data according to the preset acceptable range, and valid data that falls within the acceptable range are retained. The effective data after cleaning were linearly fitted using the least squares method to obtain the fitted linear equations of the opening dimensions of the outer and inner gantry along the travel direction. The values ​​of the fitted linear equations of the outer and inner masts at the mean positions of their respective measurement sequences are used as the fitted opening size of the outer mast and the fitted opening size of the inner mast, respectively. The actual assembly gap is calculated, and the required compensation shim thickness is calculated based on the preset standard gap. Select the matching gasket specification from the gasket specification library according to the required compensation gasket thickness; Output the selection results, which should include at least the recommended gasket specifications and quantity.

2. The intelligent selection method for gantry assembly gaskets according to claim 1, characterized in that, The acquisition of the opening size data includes: data acquisition through photoelectric sensors, with an acquisition frequency of 1 to 10 times per second, and the acquisition range covering the entire stroke of the gantry.

3. The intelligent selection method for gantry assembly gaskets according to claim 1, characterized in that, During the data acquisition process, the measurement data is transferred from the calibration station to the painting line and then to the assembly station through the cooperation of the labeling and scanning mechanism, the label carrier, and the QR code: At the calibration station, the opening dimension data is bound to the gantry material number by affixing and scanning codes; When the painting line is in operation, the information of the label is transmitted to the label carrier on the hook of the painting line by scanning the barcode; When the painting is completed and the product comes off the production line, the data information in the carrier is transferred to the QR code by a barcode scanner and then affixed to the gantry.

4. The intelligent selection method for gantry assembly gaskets according to claim 1, characterized in that, In the linear fitting process, the sequence of measurement times of the outer or inner gantry is used as the independent variable X, and the corresponding sequence of opening size measurement values ​​is used as the dependent variable Y. The slope m and intercept b of the fitted line are calculated by the least squares method, and the fitted line equation y = mx + b is obtained.

5. The intelligent selection method for gantry assembly gaskets according to claim 1, characterized in that, When calculating the gap, the actual assembly gap Gap = Outer - Inner, and the required compensation shim thickness RequiredThickness = TargetGap - Gap. Where Outer is the fitted opening size of the outer gantry, Inner is the fitted opening size of the inner gantry, and TargetGap is the preset standard gap; if RequiredThickness is positive, it indicates that shims need to be added; if it is negative, it indicates that shims need to be reduced.

6. The intelligent selection method for gantry assembly gaskets according to claim 1, characterized in that, During the gasket matching process, the median value of two adjacent gasket specifications is used as a threshold for judgment: if the required compensation gasket thickness is greater than the median value, the larger specification gasket is selected; if the required compensation gasket thickness is less than the median value, the smaller specification gasket is selected.

7. The intelligent selection method for gantry assembly gaskets according to claim 1, characterized in that, The gasket matching step offers three matching modes to choose from: median mode, which uses the arithmetic median of adjacent gasket specifications as a threshold for judgment; interference mode, which selects a larger specification gasket; and low-end mode, which selects a smaller specification gasket.

8. The intelligent selection method for gantry assembly gaskets according to any one of claims 1 to 7, characterized in that, While outputting the results, the collected raw data, fitting calculation results, gantry number, and selection results are stored in the database and timestamps are recorded to support historical data retrieval and export by time and / or product model.

9. The intelligent selection method for gantry assembly gaskets according to any one of claims 1 to 7, characterized in that, The selection method further includes: when the number of valid data after cleaning is lower than a preset threshold, issuing a data shortage alarm and prompting for re-collection; when the slope of the fitted straight line equation exceeds a preset slope threshold, issuing a warning signal that the gantry opening size changes abnormally along the travel direction.

Images