Control method of investment casting automatic wax pattern welding equipment

Abstract

The invention relates to the technical field of investment casting, and particularly provides an investment casting automatic wax pattern welding equipment control method which comprises the following steps: extracting a peak value and duration from heating curve data, and judging a uniformity numerical value of interface temperature distribution through a numerical simulation method; according to the fusion depth change trend and the interface temperature distribution data, the quantized value of the quality index is calculated, and an evaluation result of the welding effect is obtained; if the quality index quantized value is lower than a preset threshold value, historical test data are extracted from a data analysis module, and the adjustment range of the process variable is determined; according to the process variable adjustment range, optimizing the parameter combination of the heating curve and the pressure curve, and generating new control strategy data through an iterative algorithm; new control strategy data is adopted to drive equipment to operate, and dynamic feedback values of the interface temperature and the fusion depth are obtained through a real-time monitoring system; and extracting abnormal fluctuation data from the dynamic feedback value, and judging the influence degree of the solidification rate change on the quality index in combination with a data analysis result.

Description

Technical Field

[0001] The present invention relates to the technical field of investment casting, and in particular to a control method for automated wax pattern welding equipment for investment casting. Background Art

[0002] As one of the core technologies in precision manufacturing, investment casting, with its enhanced automation and intelligence, is crucial for advancing high-end equipment manufacturing. Within this process, intelligent control of wax pattern processing equipment is crucial for ensuring casting quality and production efficiency. Wax pattern welding, a crucial step in wax pattern formation, directly impacts the precision and performance of the final casting. Traditional wax pattern welding methods are no longer able to meet the urgent demands of modern production for consistency and efficiency, necessitating the exploration of new approaches to intelligent control.

[0003] Currently, wax pattern welding relies primarily on manual operation and empirical judgment, lacking systematic parameter control. This results in significant fluctuations in welding quality and makes it difficult to achieve process stability and repeatability. This approach, when applied to complex wax patterns or for large-scale production, exhibits significant drawbacks such as low efficiency and difficulty ensuring quality. The arbitrariness of parameter settings and the lack of process monitoring make welding results difficult to predict and optimize, limiting the further development of investment casting automation.

[0004] In the research on intelligent control of wax pattern welding, the core challenge lies in how to build a comprehensive and controllable parameter system to cope with the technical difficulties brought about by multivariable coupling. Specifically, the coordinated optimization of three major technical factors has not yet been achieved: the uncertainty of basic process conditions such as wax type and ambient temperature, the dynamic adjustment of process control variables such as heating curves and pressure curves, and the real-time feedback of quality evaluation indicators such as interface temperature distribution and fusion degree. Because the interaction mechanism between these factors has not been fully clarified, the parameter adjustment lacks a scientific basis, resulting in the frequent occurrence of unique problems such as wax pattern deformation and insufficient fusion during the welding process.

[0005] Therefore, how to establish a multi-level parameter composition system to systematically analyze the intrinsic relationship between basic process conditions, process control variables and quality evaluation indicators, and thereby achieve precise control of wax pattern welding, has become a key issue in improving the intelligent control level of investment casting automation equipment. Summary of the Invention

[0006] The present invention provides a control method for automated wax pattern welding equipment for investment casting, which mainly includes: Obtain wax type and ambient temperature data, collect wax physical properties and environmental conditions in real time through sensors, and determine the initial solidification rate benchmark value; According to the solidification rate reference value, the initial parameters of the heating curve are calculated, and the temperature change curve data over time is generated using the preset heat conduction model; The peak value and duration are extracted from the heating curve data, and the uniformity of the interface temperature distribution is determined by numerical simulation method; Obtain pressure curve parameters, combine them with interface temperature distribution values, and use finite element analysis technology to determine the trend of wax pattern deformation and fusion depth; The quantitative value of the quality index is calculated through the fusion depth change trend and interface temperature distribution data to obtain the evaluation results of the welding effect; If the quantitative value of the quality indicator is lower than the preset threshold, historical test data is extracted from the data analysis module to determine the adjustment range of the process variables; According to the range of process variables, the parameter combination of heating curve and pressure curve is optimized, and new control strategy data is generated through iterative algorithm; Adopting new control strategy data to drive equipment operation, the dynamic feedback values ​​of interface temperature and fusion depth are obtained through real-time monitoring system; Extract abnormal fluctuation data from dynamic feedback values ​​and combine data analysis results to determine the impact of solidification rate changes on quality indicators; The solution revolves around the goal of intelligent control of wax pattern welding. After obtaining the wax type and ambient temperature data, the solidification rate baseline value is determined, the heating curve is calculated and generated through the heat conduction model, the interface temperature distribution is extracted from the heating curve and the uniformity is judged, the pressure curve is combined with the interface temperature to determine the fusion depth, and the quality index is calculated from the fusion depth and interface temperature. If the index does not meet the standard, the data analysis extracts the adjustment range, the process variables are optimized to generate a control strategy, the equipment operates according to the strategy, and the dynamic feedback value determines the impact of the change in solidification rate, and finally achieves closed-loop control of the parameter system. The algorithms used include heat conduction model, finite element analysis and iterative algorithm. The attributes are logically closely related, and irrelevant attributes such as "control strategy-wax type" are eliminated. The output of each step is connected in sequence to form a complete thinking chain.

[0007] Optionally, the step of obtaining the wax type and ambient temperature data, collecting the physical properties of the wax and the ambient conditions in real time through sensors, and determining the initial solidification rate reference value includes: Sensors are used to collect physical property data such as wax temperature, viscosity, and density. Based on the collected physical property data, the current state index of the wax is calculated. A clustering algorithm is used to classify the wax state index, identify the process stage of the wax, obtain condition data such as ambient temperature, humidity, and air pressure, and establish an environmental parameter model. Combining the wax state index and the environmental parameter model, regression analysis is used to determine the influencing factors of the solidification rate. Through a neural network algorithm, the initial solidification rate baseline value is predicted according to the influencing factors. Real-time correction is performed on the predicted baseline value to dynamically adjust the solidification rate control parameters.

[0008] Optionally, the step of calculating the initial parameters of the heating curve based on the solidification rate reference value and generating curve data of temperature change over time using a preset heat conduction model includes: The initial parameters are calculated using the solidification rate and the reference value, and the data is processed using a preset heat conduction model to generate a preliminary temperature distribution. Based on the preliminary temperature distribution and combined with the heat conduction model, the characteristic value of temperature change over time is obtained to determine the change trend; If the change trend exceeds the preset model range, the initial parameters are adjusted and the temperature change data are regenerated to obtain a stable distribution; By stabilizing the distribution, the corresponding relationship between time change and temperature change is extracted to generate complete curve data; Use curve data to analyze the effect of solidification rate on temperature distribution and determine data consistency; According to the consistency judgment results, the heat conduction model parameters are optimized to obtain the final temperature distribution; Through the final temperature distribution, the curve data under time change is generated to complete the business processing.

[0009] Optionally, extracting the peak value and duration from the heating curve data and determining the uniformity value of the interface temperature distribution by a numerical simulation method includes: Obtain peak data and duration from the heating curve, determine the effective range using a preset threshold, and obtain a key feature set; For the key feature set, the interface temperature is calculated using numerical simulation methods to obtain temperature distribution data; Calculate the uniformity value and determine the distribution characteristics through temperature distribution data; If the uniformity value exceeds the preset threshold, the optimized temperature distribution data is obtained by adjusting the simulation method parameters; According to the optimized temperature distribution data, the interpolation algorithm is used to process the boundary area to obtain a smooth distribution result; By smoothing the distribution results, the temperature uniformity trend is determined and the final evaluation data is obtained.

[0010] Optionally, obtaining the pressure curve parameters, combining the interface temperature distribution values, and using finite element analysis technology to determine the variation trend of the wax pattern deformation and fusion depth includes: Obtain pressure curve parameter data, record the specific value of pressure changes over time, use a thermal sensor array to measure the temperature of each point on the interface, and construct a temperature distribution numerical matrix. According to the pressure parameters and temperature distribution, establish a three-dimensional geometric model of the wax mold and its material properties. Use finite element analysis software to apply pressure and temperature boundary conditions to the wax mold. Through iterative calculation, obtain the displacement and stress distribution data of each node of the wax mold. If the node displacement exceeds the preset threshold, it is determined that deformation has occurred at that location, record the degree of deformation, calculate the temperature gradient of adjacent material units, and determine the position and depth change of the fusion interface.

[0011] Optionally, the calculation of the quantitative value of the quality index by using the fusion depth variation trend and the interface temperature distribution data to obtain the evaluation result of the welding effect includes: The original information is obtained from the depth variation trend and interface temperature distribution data, and the noise is removed by data preprocessing method to obtain the smoothed fusion information; Features are extracted from the smoothed fusion information, and the contribution rate of depth change and temperature distribution is calculated using the principal component analysis algorithm to obtain quality-related parameters; Quantify the quality-related parameters, calculate the quantitative values ​​of the quality indicators through linear regression algorithm, and determine the preliminary welding effect characteristics; If the quantified value exceeds the preset threshold, the fusion information is analyzed again to obtain trend analysis results and determine the stability of the welding effect; Based on the comparison between the trend analysis results and the quantitative values, the quality indicators were adjusted using the mean calculation method to obtain the corrected effect characteristics; By comparing the corrected effect characteristics with the standard value of the evaluation result, if the deviation is less than the preset threshold, the evaluation result of the welding effect is confirmed; After obtaining the confirmed assessment results, data storage tools are used to save the correlation information of depth change trends and interface temperature distribution to complete business processing.

[0012] Optionally, if the quantified value of the quality indicator is lower than a preset threshold, extracting historical test data from the data analysis module to determine the adjustment range of the process variables includes: If the quantitative value of the quality indicator is lower than the preset threshold, the historical test data is obtained from the data module, and the data integrity is judged to obtain the preliminary screening results; The test data were extracted from the preliminary screening results, and cluster analysis was used to determine the initial grouping of process variables; Obtain variable distribution characteristics based on initial grouping and determine the preliminary boundaries of the process variable adjustment range; If the preliminary boundaries are outside the historical test range, the boundaries are adjusted through linear regression analysis to obtain the optimized range value; Extract relevant quality indicators based on the optimized range values ​​and determine the indicator change trend; Interpolation method is used to process discrete data according to the change trend to determine the final process variable adjustment range; The updated historical test data is obtained by updating the data module through the final adjustment range.

[0013] Optionally, adjusting the range of process variables, optimizing the parameter combination of the heating curve and the pressure curve, and generating new control strategy data through an iterative algorithm may include: By analyzing the process variables, an initial parameter set is obtained from the adjustment range to obtain the variable analysis results; Based on the variable analysis results, the preset threshold is used to determine the preliminary adjustment direction of the heating curve and pressure curve, and the curve adjustment plan is determined; For the curve adjustment scheme, the changing trend of parameter combination is calculated through iterative algorithm to obtain the optimization process data; If the optimization process data exceeds the adjustment range, the parameter combination is updated through variable analysis to generate new control strategy data; Based on the new control strategy data, a machine learning algorithm is used to predict the matching degree of the heating curve and the pressure curve to determine the need for strategy update; Through the strategy update requirements, the improved parameter combination is obtained from the control strategy to obtain the final data generation result.

[0014] Optionally, the new control strategy data is used to drive the equipment operation, and dynamic feedback values ​​of the interface temperature and fusion depth are obtained through a real-time monitoring system, including: Start the equipment operation in a data-driven manner and obtain the initial values ​​of interface temperature and fusion depth in the real-time monitoring system; Use real-time monitoring technology to process the initial value and obtain dynamic feedback values ​​of temperature change and depth adjustment; If the temperature change exceeds the preset threshold, the equipment operating status is adjusted through the control strategy to determine the new operating parameters; According to the adjusted operating parameters, obtain the updated value of the fusion depth to determine whether the depth adjustment meets expectations; Analyze temperature change trends through dynamic feedback values ​​to obtain stability indicators of equipment operating status; Use support vector machine algorithm to classify stability indicators and determine whether the operating status needs further optimization; If the optimization conditions are triggered, the control strategy is adjusted through the gradient descent algorithm to obtain the final operating parameters.

[0015] Optionally, extracting abnormal fluctuation data from the dynamic feedback value and judging the degree of influence of the solidification rate change on the quality index in combination with the data analysis result includes: Obtain feedback values ​​from dynamic feedback, extract abnormal fluctuation data through time series analysis, and obtain the fluctuation range; Based on the fluctuation range, cluster analysis was used to determine the correlation between the fluctuation data and the solidification rate, and to determine the rate change trend; According to the rate change trend, obtain the historical data of quality indicators and determine the distribution law of indicators with rate changes; Based on the distribution law, regression analysis is used to calculate the influence weight of solidification rate change on quality indicators and obtain the influence degree value; If the impact value exceeds the preset threshold, the characteristic data of abnormal fluctuations are extracted from the feedback value to determine the key fluctuation points; Targeting key fluctuation points, obtain real-time data on solidification rate and quality indicators during the same period to determine the direct connection between abnormal fluctuations and the impact of indicators; Through real-time data, the specific impact of changes in solidification rate at key fluctuation points on quality indicators is calculated to obtain the final judgment result.

[0016] Optionally, the solution is developed around the goal of intelligent control of wax pattern welding; after obtaining the wax type and ambient temperature data, a solidification rate baseline value is determined; the heating curve is calculated and generated using a heat conduction model, and the interface temperature distribution is extracted from the heating curve and the uniformity is judged; the pressure curve is combined with the interface temperature to determine the fusion depth, and the quality index is calculated from the fusion depth and interface temperature; if the index does not meet the standard, data analysis is performed to extract the adjustment range, and the process variables are optimized to generate a control strategy; the equipment operates according to the strategy, and the dynamic feedback value determines the impact of the change in solidification rate, ultimately achieving closed-loop control of the parameter system; the algorithms used include heat conduction models, finite element analysis, and iterative algorithms, with close logical correlations between attributes, and irrelevant attributes such as "control strategy-wax type" are eliminated; the outputs of each step are connected in sequence to form a complete thinking chain, including: Obtain wax type and ambient temperature data, and calculate the solidification rate benchmark value through the heat conduction model; Finite element analysis was used to generate heating curves from solidification rate baseline values ​​and extract interface temperature distribution data; The uniformity value is determined by the interface temperature distribution data. If the uniformity value is lower than the preset threshold, the interface temperature distribution is analyzed to extract the adjustment range. The fusion depth is calculated by combining the adjustment range and the pressure curve data to obtain the fusion depth distribution; Calculate the quality index through the fusion depth distribution and interface temperature distribution to determine whether the quality index reaches the preset threshold; Generate control strategies based on quality indicators, and the equipment adjusts operating parameters according to the control strategies to obtain feedback values; The solidification rate change trend is determined by feedback value and iterative algorithm, and the parameter system is updated to form a closed-loop control.

[0017] The technical solution provided by the embodiment of the present invention may have the following beneficial effects: The present invention discloses a control method for automated wax pattern welding equipment for investment casting, specifically an intelligent control method for wax pattern welding. The method determines the initial solidification rate reference value by real-time collection of wax material physical properties and environmental condition data, and calculates the heating curve parameters based on this. The temperature change curve is generated using a heat conduction model, and the deformation and fusion depth change trends of the wax pattern are determined in combination with pressure curve parameters and finite element analysis technology. Quality indicators are calculated based on the fusion depth and interface temperature distribution to evaluate the welding effect. When the quality does not meet the standards, the present invention extracts the process variable adjustment range from historical data and optimizes the control strategy. Dynamic feedback is obtained through real-time monitoring to determine the impact of solidification rate changes on quality and achieve closed-loop control of parameters. The present invention integrates algorithms such as heat conduction models and finite element analysis, effectively improving the intelligence level and quality stability of wax pattern welding. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] Figure 1 The present invention is a flow chart of a control method for automated wax pattern welding equipment for investment casting.

[0019] Figure 2 The figure is a schematic diagram of a control method for automated wax pattern welding equipment for investment casting according to the present invention.

[0020] Figure 3 This is another schematic diagram of a control method for automated wax pattern welding equipment for investment casting according to the present invention. DETAILED DESCRIPTION

[0021] The following will describe the technical solutions in the embodiments of the present invention in detail with reference to the accompanying drawings. The described embodiments are only a part of the embodiments of the present invention.

[0022] like Figure 1-3 In this embodiment, a control method for automated wax pattern welding equipment for investment casting may specifically include: Step S101 , obtaining wax material type and ambient temperature data, collecting wax material physical properties and ambient condition values ​​in real time through sensors, and determining an initial solidification rate reference value.

[0023] The temperature, viscosity, density and other physical property data of the wax are collected through sensors. The current state index of the wax is calculated based on the collected physical property data. The clustering algorithm is used to classify the wax state index and identify the process stage of the wax. The ambient temperature, humidity, air pressure and other condition data are obtained to establish an environmental parameter model. Combining the wax state index and the environmental parameter model, regression analysis is used to determine the influencing factors of the solidification rate. The neural network algorithm is used to predict the initial solidification rate benchmark value based on the influencing factors. The predicted benchmark value is corrected in real time to dynamically adjust the solidification rate control parameters. Specifically, the sensor collects the wax type and ambient temperature data in real time. Assume that the current wax is paraffin and the ambient temperature is 25 degrees Celsius. The physical properties of the wax obtained by the sensor include viscosity, density and thermal conductivity, which are measured to be 85Pa·s, 900kg / m³ and 15W / (m·K) respectively. The environmental condition values ​​include a relative humidity of 60% and a wind speed of 5m / s. Based on these data, the heat conduction equation and Newton's law of cooling are used to calculate the initial solidification rate reference value. First, according to the thermal conductivity of the wax and the ambient temperature, the heat flux density is calculated using Fourier's law, and the heat flux density is 35W / m². Then, combined with the wax density and viscosity, the solidification rate formula R=α·ΔT / ρ·L is used, where α is the thermal diffusion coefficient, ΔT is the temperature difference, ρ is the density, L is the latent heat, and α is assumed to be 2×10- 7 m² / s, ΔT of 5K, and L of 200kJ / kg, the calculated initial solidification rate baseline is 0.03 m / s. Further analysis shows that the solidification rate can be optimized by adjusting the ambient temperature and wind speed. For example, lowering the ambient temperature to 20 degrees Celsius and increasing the wind speed to 1 m / s results in a recalculated solidification rate of 0.04 m / s. These data and analysis provide a scientific basis for optimizing wax processing technology.

[0024] Step S102 : calculating the initial parameters of the heating curve according to the solidification rate reference value, and generating curve data of temperature change over time using a preset heat conduction model.

[0025] Initial parameters are calculated using the solidification rate and baseline value, and the data is processed using a preset heat conduction model to generate a preliminary temperature distribution. Based on this preliminary temperature distribution and in combination with the heat conduction model, characteristic values ​​of temperature changes over time are obtained to determine the trend of change. If the trend of change exceeds the preset model range, the initial parameters are adjusted and the temperature change data is regenerated to obtain a stable distribution. Using the stable distribution, the correspondence between time changes and temperature changes is extracted to generate complete curve data. Using the curve data, the impact of the solidification rate on the temperature distribution is analyzed to determine data consistency. Based on the consistency judgment results, the heat conduction model parameters are optimized to obtain the final temperature distribution. Using the final temperature distribution, curve data under time changes is generated to complete the business processing.

[0026] Specifically, based on the solidification rate benchmark value, assuming it is 5 mm per second, when calculating the initial parameters of the heating curve, the thermal conductivity of the material is first determined to be 50 watts per meter Kelvin, the specific heat capacity is 500 joules per kilogram Kelvin, and the density is 7800 kilograms per cubic meter. Using the preset heat conduction model, numerical calculations are performed using the finite difference method, and the material is divided into 100 grid units, each with a length of 0.1 meter. The initial temperature is set to 20 degrees Celsius, the heating source temperature is 1000 degrees Celsius, and the heating time is 100 seconds. During the calculation process, an explicit difference format is used, and the time step is 1 second to ensure numerical stability. Through iterative calculations, curve data of temperature changes over time are generated, and the temperature distribution is recorded once for each time step.

[0027]

[0028] Where T(t) represents the temperature at any given moment, T_0 represents the initial temperature, A_i represents the amplitude of the i-th temperature component, λ_i represents the attenuation coefficient of the i-th temperature component, t represents time, and n represents the number of temperature components. The analysis results show that during the heating process, the surface temperature of the material rises rapidly, while the internal temperature gradually increases, eventually reaching a steady state. By comparing the temperature distribution at different time points, the accuracy of the heat conduction model can be verified, and heating parameters can be further optimized to improve heating efficiency.

[0029] Step S103 , extracting the peak value and duration from the heating curve data, and determining the uniformity value of the interface temperature distribution by numerical simulation method.

[0030] Peak data and duration are obtained from the heating curve. The effective range is determined using a preset threshold to obtain a key feature set. Numerical simulation methods are used to calculate the interface temperature for this key feature set, generating temperature distribution data. The uniformity value is calculated from this temperature distribution data to determine the distribution characteristics. If the uniformity value exceeds the preset threshold, the simulation method parameters are adjusted to obtain optimized temperature distribution data. Based on this optimized temperature distribution data, an interpolation algorithm is used to process the boundary regions to obtain a smoothed distribution. This smoothed distribution is then used to determine the temperature uniformity trend and obtain the final evaluation data.

[0031] Specifically, the peak values ​​and durations were extracted from the heating curve data. First, the raw data were analyzed in the frequency domain by fast Fourier transform (FFT) to identify the main frequency components.

[0032] For example, during a heating process, the peak temperature is 120°C and the duration is 60 seconds. Next, a moving average filtering algorithm is used to smooth the data, remove noise interference, and ensure the accuracy of the peak value and duration. A heat conduction model is established through numerical simulation methods, and finite element analysis (FEA) is used to simulate the interface temperature distribution. Assuming the thermal conductivity of the material is 50 W / m·K and the initial temperature is 25°C, the simulation results show that the interface temperature is relatively uniform during the heating process, with a maximum temperature difference of 5°C. To further verify the uniformity, the standard deviation of the temperature distribution is calculated. If the standard deviation is less than 2°C, the interface temperature distribution is considered uniform. This analysis method can ensure the uniformity of the interface temperature during the heating process, thereby improving product quality and process stability.

[0033] Step S104 , obtaining pressure curve parameters, combining them with interface temperature distribution values, and using finite element analysis technology to determine the variation trends of the wax pattern deformation and fusion depth.

[0034] Obtain pressure curve parameter data and record the specific value of pressure change over time. Use a thermal sensor array to measure the temperature of each point on the interface and construct a temperature distribution numerical matrix. Based on the pressure parameters and temperature distribution, establish a three-dimensional geometric model of the wax mold and its material properties. Use finite element analysis software to apply pressure and temperature boundary conditions to the wax mold. Through iterative calculation, obtain the displacement and stress distribution data of each node of the wax mold. If the node displacement exceeds the preset threshold, it is determined that deformation has occurred at that location, and the degree of deformation is recorded. Calculate the temperature gradient of adjacent material units to determine the position and depth change of the fusion interface. Specifically, first, the pressure curve parameters of the wax mold under specific process conditions are obtained through experimental equipment. For example, under the conditions of an injection pressure of 5 MPa and a holding time of 10 seconds, the curve of pressure change with time shows a trend of first rising rapidly and then slowly falling. Then, an infrared thermal imager is used to measure the temperature distribution at the interface of the wax mold to obtain temperature field data. For example, the temperature in the center area of ​​the wax mold surface is 80 degrees Celsius and the temperature in the edge area is 65 degrees Celsius. Next, finite element analysis software such as ANSYS is used to use the pressure curve parameters and interface temperature distribution values ​​as input conditions to establish a finite element model of the wax mold. In the model, the material property parameters are set, such as the elastic modulus of the wax mold is 1.2 GPa, the Poisson's ratio is 0.3, and the thermal expansion coefficient is 8.5×10- 5per degree Celsius. Through thermal-structural coupling analysis, the stress and displacement fields of the wax pattern are solved, and the deformation of the wax pattern at different time points is determined. For example, at the end of the pressure holding stage, the deformation of the wax pattern is 0.15 mm. At the same time, the contact between the wax pattern and the mold is analyzed, and the changing trend of the fusion depth is calculated. For example, when the contact pressure is 3 MPa, the fusion depth gradually increases from the initial 0.02 mm to 0.08 mm. By adjusting the process parameters, such as increasing the injection pressure to 6 MPa, it can be observed that the deformation of the wax pattern is reduced to 0.12 mm, while the fusion depth is increased to 0.1 mm, thereby optimizing the molding quality of the wax pattern.

[0035] Step S105 , calculating the quantitative value of the quality index through the fusion depth variation trend and the interface temperature distribution data, and obtaining the evaluation result of the welding effect.

[0036] Raw information is obtained from the depth variation trend and interface temperature distribution data. Data preprocessing methods are used to remove noise, resulting in smoothed fused information. Features are extracted from the smoothed fused information, and the contribution rates of the depth variation and temperature distribution are calculated using a principal component analysis algorithm to obtain quality-related parameters. Quality-related parameters are quantified, and the quantified values ​​of quality indicators are calculated using a linear regression algorithm to determine preliminary welding effect characteristics. If the quantified values ​​exceed the preset threshold, the fused information is subjected to a secondary analysis to obtain trend analysis results and determine the stability of the welding effect. Based on the comparison of the trend analysis results with the quantified values, the quality indicators are adjusted using a mean calculation method to obtain a revised effect characteristic. The revised effect characteristic is compared with the standard value of the evaluation result. If the deviation is less than the preset threshold, the welding effect evaluation result is confirmed. After obtaining the confirmed evaluation results, the correlation information of the depth variation trend and interface temperature distribution is saved using a data storage tool to complete the business processing.

[0037] Specifically, during the welding process, real-time data on the fusion depth trend and interface temperature distribution can be collected to quantitatively assess weld quality. First, a sensor is used to collect fusion depth data. Assume that during the welding process, the fusion depth gradually increases from an initial value of 2 mm to 5 mm, with a linear growth trend and a slope of approximately 5 mm per second. Simultaneously, interface temperature distribution data is acquired using an infrared thermal imager. Assume that the temperature in the weld area rapidly increases from room temperature (25°C) to a peak value of 1200°C and then gradually cools to 300°C after welding. Based on this data, a weighted average algorithm is used to calculate the quality index, with a weight of 6 for the fusion depth trend and a weight of 4 for the interface temperature distribution. The specific calculation process is as follows: First, the fusion depth data is normalized to a normalized value of 8; then, the interface temperature distribution data is normalized to a normalized value of 7. Finally, the weighted sum of the two is used to obtain a quality index of 8 × 6 + 7 × 4 = 76. According to the preset quality assessment criteria, a quality index greater than 7 indicates good welding quality. Therefore, the above analysis indicates that the welding quality is good. In addition, to further optimize the welding process, historical data can be combined with machine learning algorithms to predict quality indicators under different welding parameters, thereby realizing intelligent control of the welding process.

[0038] Step S106: If the quantified value of the quality indicator is lower than the preset threshold, historical test data is extracted from the data analysis module to determine the adjustment range of the process variables.

[0039] If the quantitative value of the quality indicator is lower than the preset threshold, historical test data is obtained from the data module, and the data integrity is judged to obtain preliminary screening results. Test data is extracted based on the preliminary screening results, and cluster analysis is used to determine the initial grouping of process variables. The variable distribution characteristics are obtained based on the initial grouping, and the preliminary boundaries of the process variable adjustment range are determined. If the preliminary boundaries exceed the historical test range, the boundaries are adjusted through linear regression analysis to obtain the optimized range values. Relevant quality indicators are extracted for the optimized range values, and the trend of indicator changes is judged. According to the change trend, the interpolation method is used to process discrete data to determine the final process variable adjustment range. The data module is updated through the final adjustment range to obtain the updated historical test data.

[0040] Specifically, when the quantitative value of a quality indicator falls below a preset threshold—for example, if the hardness value of a particular batch of products falls below the standard value of 60 HRC—the system automatically triggers the data analysis module to extract relevant data from the historical test database. This data includes production records for the past 100 batches, covering key process variables such as heating temperature (ranging from 800°C to 1000°C), holding time (ranging from 30 minutes to 60 minutes), and cooling rate (ranging from 5°C / s to 20°C / s). Using a multivariate linear regression algorithm, the system analyzed the correlation between process variables and hardness values ​​in the historical data and determined that the coefficients of influence of heating temperature, holding time, and cooling rate on hardness are 35, 28, and 42, respectively. Based on these coefficients, the system further optimized the adjustment range of process variables using a gradient descent method, calculating that the heating temperature should be increased to 950°C to 980°C, the holding time should be extended to 45 minutes to 55 minutes, and the cooling rate should be adjusted to 12°C / s to 15°C / s to ensure that the hardness value reaches the target range. At the same time, the system combines process constraints, such as equipment temperature limits and energy consumption limits, to make boundary corrections to the adjustment range to ensure the feasibility and economy of process adjustments.

[0041] Step S107 , optimizing the parameter combination of the heating curve and the pressure curve according to the process variable adjustment range, and generating new control strategy data through an iterative algorithm.

[0042] By analyzing process variables, an initial parameter set is obtained from the adjustment range, generating variable analysis results. Based on the variable analysis results, a preset threshold is used to determine the initial adjustment direction for the heating and pressure curves, and a curve adjustment plan is determined. Based on the curve adjustment plan, an iterative algorithm is used to calculate the changing trend of the parameter combination to generate optimized process data. If the optimized process data exceeds the adjustment range, the parameter combination is updated through variable analysis to generate new control strategy data. Based on the new control strategy data, a machine learning algorithm is used to predict the matching degree of the heating and pressure curves and determine the need for a strategy update. Based on the strategy update requirements, an improved parameter combination is obtained from the control strategy, resulting in the final data generation results.

[0043] Specifically, within the adjustment range of process variables, historical data for heating and pressure curves was first acquired through the data acquisition system. Based on current process requirements, the heating temperature range was set to 200°C to 300°C, and the pressure range to 5 MPa to 5 MPa. Based on these parameters, a genetic algorithm was used for optimization, with a population size of 50 and 100 iterations. The process performance of each parameter combination was calculated using a fitness function that comprehensively considers heating efficiency, energy consumption, and product quality, with weights of 4, 3, and 3, respectively. In each iteration, new parameter combinations were generated through selection, crossover, and mutation. For example, in one iteration, two parameter combinations were selected from the parent generation (heating temperature 250°C, pressure 8 MPa and heating temperature 280°C, pressure 2 MPa). A crossover operation was used to generate the parameters of the child generation (heating temperature 265°C, pressure 0 MPa). The heating temperature was then randomly adjusted by ±5°C in a mutation operation, resulting in a new parameter combination (heating temperature 260°C, pressure 0 MPa). Through multiple iterations, the algorithm gradually converged to the optimal parameter combination (heating temperature 270°C, pressure 1 MPa). At this point, the fitness function reached its highest value, indicating that this combination achieved the best performance in terms of heating efficiency, energy consumption, and product quality. Finally, this optimized parameter combination was input into the control system to generate new control strategy data, allowing for real-time adjustments to the heating and pressure curves to ensure process stability and efficiency.

[0044] Step S108: Use the new control strategy data to drive the equipment to operate, and obtain dynamic feedback values ​​of the interface temperature and fusion depth through the real-time monitoring system.

[0045] The equipment is started using a data-driven approach to obtain initial values ​​for the interface temperature and fusion depth from the real-time monitoring system. Real-time monitoring technology is used to process these initial values, generating dynamic feedback on temperature changes and depth adjustments. If the temperature change exceeds a preset threshold, the control strategy adjusts the equipment's operating state and determines new operating parameters. Based on the adjusted operating parameters, an updated fusion depth value is obtained to determine whether the depth adjustment meets expectations. The dynamic feedback values ​​are used to analyze temperature change trends and determine stability indicators for the equipment's operating state. A support vector machine algorithm is used to classify these stability indicators and determine whether the operating state requires further optimization. If the optimization condition is triggered, the control strategy is adjusted using a gradient descent algorithm to obtain the final operating parameters.

[0046] Specifically, a new control strategy uses data-driven equipment operation, obtaining dynamic feedback values ​​of interface temperature and fusion depth through a real-time monitoring system. First, the system collects interface temperature data through a high-precision sensor and transmits it to the central processing unit for real-time analysis.

[0047] For example, when the interface temperature reaches the target value of 1200 degrees Celsius, the system triggers the temperature control algorithm and adjusts the heating power through the PID controller to ensure that the temperature is stable within a range of ±10 degrees Celsius. At the same time, the fusion depth is monitored in real time by a laser scanner. If the fusion depth deviates from the target value by 5 mm, the system automatically adjusts the laser power and scanning speed, using a fuzzy logic algorithm for dynamic compensation to keep the fusion depth within a precise range of 8 mm to 2 mm. In addition, the system combines historical data for trend analysis and uses machine learning algorithms to predict possible temperature fluctuations and fusion deviations, allowing for parameter optimization in advance.

[0048] For example, if the system predicts that the interface temperature may rise to 1250 degrees Celsius within the next 10 seconds, it automatically reduces heating power and optimizes control parameters through a neural network algorithm to ensure stable and accurate operation. The entire process requires no human intervention; it is fully automated by information technology, ensuring efficient and precise operation.

[0049] Step S109 , extracting abnormal fluctuation data from the dynamic feedback value, and combining the data analysis results to determine the degree of influence of the solidification rate change on the quality index.

[0050] Feedback values ​​are obtained from dynamic feedback, and abnormal fluctuation data is extracted through time series analysis to obtain the fluctuation range. For the fluctuation range, cluster analysis is used to determine the correlation between the fluctuation data and the solidification rate, and the rate change trend is determined. Based on the rate change trend, historical data of quality indicators are obtained to determine the distribution pattern of the indicators with rate changes. Based on the distribution pattern, regression analysis is used to calculate the impact weight of the solidification rate change on the quality indicator to obtain the impact degree value. If the impact degree value exceeds the preset threshold, the characteristic data of the abnormal fluctuation is extracted from the feedback value to determine the key fluctuation point. For the key fluctuation point, real-time data of the solidification rate and quality indicators in the same period are obtained to determine the direct connection between the abnormal fluctuation and the impact of the indicator. Through real-time data, the specific impact value of the solidification rate change on the quality indicator at the key fluctuation point is calculated to obtain the final judgment result.

[0051] Specifically, when extracting abnormal fluctuation data from dynamic feedback values, the abnormal points are first identified by setting a threshold.

[0052] For example, the normal range of solidification rate is set at 5 to 5 mm per minute. When the rate is detected to be lower than 3 mm or higher than 8 mm, the system automatically marks it as abnormal. Then, a sliding window algorithm is used to smooth the continuous time series data with a window size of 10 data points. The average and standard deviation of the data within the window are calculated to further confirm abnormal fluctuations.

[0053] For example, if the standard deviation of the coagulation rate exceeds 2 mm over a certain period, the system will identify it as a significant fluctuation. Combined with the data analysis results, a linear regression model is used to assess the impact of changes in coagulation rate on quality indicators. Assuming a regression coefficient of 8, this indicates that for every 1 mm increase in coagulation rate, the quality indicator decreases by 8 units. The calculated coefficient of determination (R²) is 85, indicating that changes in coagulation rate explain 85% of the variation in quality indicators. Finally, principal component analysis (PCA) is used to reduce the dimensionality of the multidimensional data and extract the main influencing factors, further verifying the dominant role of coagulation rate on quality indicators.

[0054] For example, the contribution rate of the first principal component is 70%, among which the loading coefficient of solidification rate is 9, indicating that it plays an important role in the change of quality indicators.

[0055] In step S1010, the solution revolves around the goal of intelligent control of wax pattern welding. After obtaining the wax type and ambient temperature data, the solidification rate baseline value is determined. The heating curve is calculated and generated through the heat conduction model, and the interface temperature distribution is extracted from the heating curve and the uniformity is judged. The pressure curve is combined with the interface temperature to determine the fusion depth, and the quality index is calculated from the fusion depth and the interface temperature. If the index does not meet the standard, the data analysis extracts the adjustment range, and the process variables are optimized to generate a control strategy. The equipment operates according to the strategy, and the dynamic feedback value determines the impact of the change in solidification rate, and finally realizes closed-loop control of the parameter system. The algorithms used include heat conduction model, finite element analysis and iterative algorithm. The attributes are closely logically related, and irrelevant attributes such as "control strategy-wax type" are eliminated. The output of each step is connected in sequence to form a complete thinking chain.

[0056] Obtain wax type and ambient temperature data, and calculate a solidification rate baseline using a heat conduction model. Finite element analysis is used to generate a heating curve from the solidification rate baseline, extracting interface temperature distribution data. Uniformity is determined using the interface temperature distribution data. If the uniformity is below a preset threshold, the interface temperature distribution is analyzed to extract the adjustment range. The fusion depth is calculated using the adjustment range and pressure curve data to obtain a fusion depth distribution. Quality indicators are calculated using the fusion depth distribution and interface temperature distribution to determine whether the quality indicators meet the preset thresholds. A control strategy is generated based on the quality indicators, and the equipment adjusts operating parameters accordingly, obtaining feedback values. The feedback value and iterative algorithm are used to determine the trend of solidification rate changes, and the parameter system is updated to form a closed-loop control system.

[0057] Specifically, during the intelligent control process for wax pattern welding, sensors first determine the wax material type: polyethylene wax, and the ambient temperature: 25°C. Based on a database of material thermophysical properties, a baseline solidification rate of 5 mm / s is determined. Using a heat conduction model and a finite element analysis algorithm, a heating curve is calculated, with a heating power of 500 W and a heating time of 10 seconds. The interface temperature distribution is extracted from the heating curve. A uniformity assessment algorithm determines that the temperature standard deviation is 2°C, meeting uniformity requirements. Based on the interface temperature, a pressure curve calculation algorithm is used to determine the fusion depth of 2 mm. Based on the fusion depth and interface temperature, a quality index calculation algorithm determines a quality index of 85, below the standard value of 9. A data analysis algorithm extracts an adjustment range of ±50 W for heating power and ±2 seconds for heating time. An iterative optimization algorithm is used to generate a control strategy of 550 W for heating power and 11 seconds for heating time. The equipment operates according to this control strategy, generating a dynamic feedback solidification rate change of 52 mm / s. An impact analysis algorithm determines that this has a positive impact on the quality index, ultimately achieving closed-loop control of the parameter system.

[0058] The above embodiments are intended to illustrate the technical solutions of the present invention and are not intended to limit the present invention. The present invention is described in detail with reference to the preferred embodiments only. It should be understood by those skilled in the art that the technical solutions of the present invention may be modified or replaced with equivalents without departing from the spirit and scope of the technical solutions of the present invention, and all such modifications or equivalents should be included within the scope of the claims of the present invention.

Claims

1. A control method for automated wax pattern welding equipment for investment casting, characterized in that: The method comprises: Obtain wax type and ambient temperature data, collect wax physical properties and environmental conditions in real time through sensors, and determine the initial solidification rate benchmark value; According to the solidification rate reference value, the initial parameters of the heating curve are calculated, and the temperature change curve data over time is generated using the preset heat conduction model; The peak value and duration are extracted from the heating curve data, and the uniformity of the interface temperature distribution is determined by numerical simulation method; Obtain pressure curve parameters, combine them with interface temperature distribution values, and use finite element analysis technology to determine the trend of wax pattern deformation and fusion depth; The quantitative value of the quality index is calculated through the fusion depth change trend and interface temperature distribution data to obtain the evaluation results of the welding effect; If the quantitative value of the quality indicator is lower than the preset threshold, historical test data is extracted from the data analysis module to determine the adjustment range of the process variables; According to the range of process variables, the parameter combination of heating curve and pressure curve is optimized, and new control strategy data is generated through iterative algorithm; Adopting new control strategy data to drive equipment operation, the dynamic feedback values ​​of interface temperature and fusion depth are obtained through real-time monitoring system; Extract abnormal fluctuation data from dynamic feedback values ​​and combine data analysis results to determine the impact of solidification rate changes on quality indicators; The solution revolves around the goal of intelligent control of wax pattern welding. After obtaining the wax type and ambient temperature data, the solidification rate baseline value is determined. The heating curve is calculated and generated through the heat conduction model. The interface temperature distribution is extracted from the heating curve and the uniformity is judged. The pressure curve is combined with the interface temperature to determine the fusion depth. The quality index is calculated from the fusion depth and interface temperature. If the index does not meet the standard, the adjustment range is extracted through data analysis, and the process variables are optimized to generate a control strategy. The equipment operates according to the strategy, and the dynamic feedback value determines the impact of the change in solidification rate, ultimately achieving closed-loop control of the parameter system. The algorithms used include heat conduction models, finite element analysis and iterative algorithms. The attributes are logically closely related, and irrelevant attributes such as "control strategy-wax type" are eliminated. The output of each step is connected in sequence to form a complete thinking chain.

2. The control method for automated wax pattern welding equipment for investment casting according to claim 1, characterized in that: The method of obtaining the wax type and ambient temperature data, collecting the physical properties of the wax and the ambient conditions in real time through sensors, and determining the initial solidification rate reference value includes: Sensors are used to collect physical property data such as wax temperature, viscosity, and density. Based on the collected physical property data, the current state index of the wax is calculated. A clustering algorithm is used to classify the wax state index, identify the process stage of the wax, obtain condition data such as ambient temperature, humidity, and air pressure, and establish an environmental parameter model. Combining the wax state index and the environmental parameter model, regression analysis is used to determine the influencing factors of the solidification rate. Through a neural network algorithm, the initial solidification rate baseline value is predicted according to the influencing factors. Real-time correction is performed on the predicted baseline value to dynamically adjust the solidification rate control parameters.

3. The control method for automated wax pattern welding equipment for investment casting according to claim 1, characterized in that: The method of calculating the initial parameters of the heating curve based on the solidification rate reference value and generating the curve data of the temperature changing with time using the preset heat conduction model includes: The initial parameters are calculated using the solidification rate and the reference value, and the data is processed using a preset heat conduction model to generate a preliminary temperature distribution. Based on the preliminary temperature distribution and combined with the heat conduction model, the characteristic value of temperature change over time is obtained to determine the change trend; If the change trend exceeds the preset model range, the initial parameters are adjusted and the temperature change data are regenerated to obtain a stable distribution; By stabilizing the distribution, the corresponding relationship between time change and temperature change is extracted to generate complete curve data; Use curve data to analyze the effect of solidification rate on temperature distribution and determine data consistency; According to the consistency judgment results, the heat conduction model parameters are optimized to obtain the final temperature distribution; Through the final temperature distribution, the curve data under time change is generated to complete the business processing.

4. The control method for automated wax pattern welding equipment for investment casting according to claim 1, characterized in that: The method of extracting the peak value and duration from the heating curve data and determining the uniformity of the interface temperature distribution by a numerical simulation method includes: Obtain peak data and duration from the heating curve, determine the effective range using a preset threshold, and obtain a key feature set; For the key feature set, the interface temperature is calculated using numerical simulation methods to obtain temperature distribution data; Calculate the uniformity value and determine the distribution characteristics through temperature distribution data; If the uniformity value exceeds the preset threshold, the optimized temperature distribution data is obtained by adjusting the simulation method parameters; According to the optimized temperature distribution data, the interpolation algorithm is used to process the boundary area to obtain a smooth distribution result; By smoothing the distribution results, the temperature uniformity trend is determined and the final evaluation data is obtained.

5. The control method for automated wax pattern welding equipment for investment casting according to claim 1, characterized in that: The pressure curve parameters are obtained, combined with the interface temperature distribution value, and the finite element analysis technology is used to determine the change trend of the wax pattern deformation and fusion depth, including: Obtain pressure curve parameter data, record the specific value of pressure changes over time, use a thermal sensor array to measure the temperature of each point on the interface, and construct a temperature distribution numerical matrix. According to the pressure parameters and temperature distribution, establish a three-dimensional geometric model of the wax mold and its material properties. Use finite element analysis software to apply pressure and temperature boundary conditions to the wax mold. Through iterative calculation, obtain the displacement and stress distribution data of each node of the wax mold. If the node displacement exceeds the preset threshold, it is determined that deformation has occurred at that location, record the degree of deformation, calculate the temperature gradient of adjacent material units, and determine the position and depth change of the fusion interface.

6. The control method for automated wax pattern welding equipment for investment casting according to claim 1, characterized in that: The quantitative value of the quality index is calculated by the fusion depth variation trend and the interface temperature distribution data to obtain the evaluation result of the welding effect, including: The original information is obtained from the depth variation trend and interface temperature distribution data, and the noise is removed by data preprocessing method to obtain the smoothed fusion information; Features are extracted from the smoothed fusion information, and the contribution rate of depth change and temperature distribution is calculated using the principal component analysis algorithm to obtain quality-related parameters; Quantify the quality-related parameters, calculate the quantitative values ​​of the quality indicators through linear regression algorithm, and determine the preliminary welding effect characteristics; If the quantified value exceeds the preset threshold, the fusion information is analyzed again to obtain trend analysis results and determine the stability of the welding effect; Based on the comparison between the trend analysis results and the quantitative values, the quality indicators were adjusted using the mean calculation method to obtain the corrected effect characteristics; By comparing the corrected effect characteristics with the standard value of the evaluation result, if the deviation is less than the preset threshold, the evaluation result of the welding effect is confirmed; After obtaining the confirmed assessment results, data storage tools are used to save the correlation information of depth change trends and interface temperature distribution to complete business processing.

7. The control method for automated wax pattern welding equipment for investment casting according to claim 1, characterized in that: If the quantified value of the quality indicator is lower than the preset threshold, extracting historical test data from the data analysis module to determine the adjustment range of the process variables includes: If the quantified value of the quality indicator is lower than the preset threshold, obtaining historical test data from the data module, judging the data integrity to obtain a preliminary screening result; The test data were extracted from the preliminary screening results, and cluster analysis was used to determine the initial grouping of process variables; Obtain variable distribution characteristics based on initial grouping and determine the preliminary boundaries of the process variable adjustment range; If the preliminary boundaries are outside the historical test range, the boundaries are adjusted through linear regression analysis to obtain the optimized range value; Extract relevant quality indicators based on the optimized range values ​​and determine the indicator change trend; Interpolation method is used to process discrete data according to the change trend to determine the final process variable adjustment range; The updated historical test data is obtained by updating the data module through the final adjustment range.

8. The control method for automated wax pattern welding equipment for investment casting according to claim 1, characterized in that: The step of optimizing the parameter combination of the heating curve and the pressure curve according to the process variable adjustment range and generating new control strategy data through an iterative algorithm includes: obtaining an initial parameter set from the adjustment range by analyzing the process variables and obtaining a variable analysis result; Based on the variable analysis results, the preset threshold is used to determine the preliminary adjustment direction of the heating curve and pressure curve, and the curve adjustment plan is determined; For the curve adjustment scheme, the changing trend of parameter combination is calculated through iterative algorithm to obtain the optimization process data; If the optimization process data exceeds the adjustment range, the parameter combination is updated through variable analysis to generate new control strategy data; Based on the new control strategy data, a machine learning algorithm is used to predict the matching degree of the heating curve and the pressure curve to determine the need for strategy update; Through the strategy update requirements, the improved parameter combination is obtained from the control strategy to obtain the final data generation result.

9. The control method for automated wax pattern welding equipment for investment casting according to claim 1, characterized in that: The new control strategy data drives the equipment operation, and the dynamic feedback values ​​of the interface temperature and fusion depth are obtained through the real-time monitoring system, including: Start the equipment operation in a data-driven manner and obtain the initial values ​​of interface temperature and fusion depth in the real-time monitoring system; Use real-time monitoring technology to process the initial value and obtain dynamic feedback values ​​of temperature change and depth adjustment; If the temperature change exceeds the preset threshold, the equipment operating status is adjusted through the control strategy to determine the new operating parameters; According to the adjusted operating parameters, obtain the updated value of the fusion depth to determine whether the depth adjustment meets expectations; Analyze temperature change trends through dynamic feedback values ​​to obtain stability indicators of equipment operating status; Use support vector machine algorithm to classify stability indicators and determine whether the operating status needs further optimization; If the optimization conditions are triggered, the control strategy is adjusted through the gradient descent algorithm to obtain the final operating parameters.

10. The control method for automated wax pattern welding equipment for investment casting according to claim 1, characterized in that: The method of extracting abnormal fluctuation data from the dynamic feedback value and combining the data analysis results to determine the degree of influence of the solidification rate change on the quality index includes: Obtain feedback values ​​from dynamic feedback, extract abnormal fluctuation data through time series analysis, and obtain the fluctuation range; Based on the fluctuation range, cluster analysis was used to determine the correlation between the fluctuation data and the solidification rate, and to determine the rate change trend; According to the rate change trend, obtain the historical data of quality indicators and determine the distribution law of indicators with rate changes; Based on the distribution law, regression analysis is used to calculate the influence weight of solidification rate change on quality indicators and obtain the influence degree value; If the impact value exceeds the preset threshold, the characteristic data of abnormal fluctuations are extracted from the feedback value to determine the key fluctuation points; Targeting key fluctuation points, obtain real-time data on solidification rate and quality indicators during the same period to determine the direct connection between abnormal fluctuations and the impact of indicators; Through real-time data, the specific impact of changes in solidification rate at key fluctuation points on quality indicators is calculated to obtain the final judgment result.

Images