Intelligent control method and system for shell laser welding

Abstract

The invention relates to the technical field of intelligent control, and particularly discloses an intelligent control method and system for shell laser welding. The method comprises the steps that heat flow distribution characteristics on a welding path are collected to construct a heat saturation matrix; generating a transient thermal resistance factor adaptive to the current heat dissipation capability based on the thermal saturation matrix; carrying out advanced rolling prediction by utilizing the transient thermal resistance factor, and extracting a heat flow inflection point coordinate of which the rheological property of the molten pool is mutated; constructing a virtual control boundary independent of the physical inflection point; and when the welding displacement triggers the virtual control boundary, a pre-judgment type negative gain damping strategy is generated. The system comprises a heat flow collection module, a thermal resistance analysis module, a sudden change analysis module, a boundary construction module and a welding regulation and control module. According to the method, by constructing the advanced control strategy containing thermal hysteresis compensation, molten pool instability caused by heat accumulation can be restrained, and heat balance regulation and control in the welding process are achieved.

Description

Technical Field

[0001] This invention relates to the field of intelligent control technology, specifically to an intelligent control method and system for laser welding of housings. Background Technology

[0002] Laser welding is widely used in the sealing welding of various precision housings (such as power battery housings, sensor packaging housings, and aerospace instrument housings) due to its advantages such as high energy density, fast welding speed, and narrow heat-affected zone.

[0003] However, in actual welding processes, the workpiece is not in an ideal thermal conduction steady state as laser energy is continuously injected. For welding of enclosed shells or long paths, there is a heat accumulation effect. Specifically, the heat in the initial stage of welding can be rapidly dissipated through the cold base material; however, as welding progresses, the overall temperature of the base material rises, and the internal heat capacity of the material gradually approaches saturation, resulting in a nonlinear decay of the heat dissipation capacity in subsequent moments.

[0004] Therefore, the present invention provides an intelligent control method and system for laser welding of housings. Summary of the Invention

[0005] The purpose of this invention is to provide an intelligent control method and system for laser welding of housings to solve the aforementioned background problems.

[0006] The objective of this invention can be achieved through the following technical solutions: A smart control method for laser welding of housings includes the following steps: The thermal saturation matrix is ​​obtained and spatial differentiation is performed to obtain the thermal response deviation between the theoretical thermal conductivity and the actual thermal diffusion. The thermal response deviation is combined with the inverse mapping analysis of the material's thermal properties to generate a transient thermal resistance factor that adapts to the current heat dissipation capacity. Based on the transient thermal resistance factor, the thermal field evolution logic is predicted in advance to generate the expected thermal accumulation rate change curve; based on the thermal accumulation rate change curve, the coordinates of the thermal flux inflection point where the rheological properties of the molten pool change abruptly are extracted. Based on the transient thermal resistance factor, the thermal response hysteresis of the heat flow inflection point coordinates is analyzed to obtain the thermal hysteresis time constant between energy input and temperature response; based on the thermal hysteresis time constant, a backtracking mapping process is performed to construct a virtual control boundary existing outside the physical inflection point; If the welding displacement triggers the virtual control boundary, the total amount of excess enthalpy accumulated inside the molten pool is calculated based on the heat saturation matrix; combined with the total amount of excess enthalpy, reverse energy cancellation processing is performed to generate a predictive negative gain damping strategy.

[0007] Furthermore, the method for obtaining the heat saturation matrix is ​​as follows: Energy injection parameters and boundary temperature data along the welding path are collected and feature calculations are performed to obtain heat flux distribution characteristics. Based on the heat flux distribution characteristics, energy superposition processing is performed to establish a heat saturation matrix.

[0008] Furthermore, the method for performing reverse mapping analysis is as follows: Obtain the basic thermal resistance value of the workpiece material and set the impedance sensitivity coefficient to characterize the temperature sensitivity of the material's thermal properties; Based on the impedance sensitivity coefficient, a linear mapping relationship is established between the thermal response deviation and the thermal resistance increment, and the thermal resistance increment is output. The transient thermal resistance factor at the current moment is obtained by calculating the sum of the base thermal resistance value and the thermal resistance increment.

[0009] Furthermore, the linear mapping relationship is established as follows: The impedance sensitivity coefficient is used as the constant of the rate of change of thermal resistance caused by a unit thermal response deviation. The value of the thermal response deviation is multiplied by the impedance sensitivity coefficient, and the result of the multiplication is assigned as the thermal resistance increment.

[0010] Furthermore, the method for extracting the coordinates of the heat flux inflection point is as follows: By differentiating the curve of the rate of heat accumulation with respect to time again, we obtain the second derivative curve characterizing the acceleration of heat accumulation. Scan the second derivative curve to identify the critical peak point where the value changes from its positive maximum value to a decrease. The critical peak point is defined as the moment when the heat accumulation effect exceeds the material's heat dissipation limit and causes instability. Obtain the relative time position of the instability moment within the virtual estimated time window, multiply the relative time position by the current welding speed, and obtain the corresponding spatial warning distance. The spatial warning distance is superimposed on the current actual welding coordinates along the welding path direction to obtain the spatial location, which is the coordinate of the heat flow inflection point where the rheological properties of the molten pool change abruptly.

[0011] Furthermore, the virtual control boundary is constructed as follows: The thermal hysteresis time constant is obtained and compared to obtain the results of the risk of asynchronous thermal response, and the current real-time welding speed is extracted. The thermal hysteresis time constant is multiplied by the average welding speed to obtain the spatial advance compensation distance used to offset the time delay. Using the coordinates of the heat flow inflection point as the physical reference origin, perform coordinate subtraction along the opposite direction of the welding path to deduct the spatial advance compensation distance. The new set of spatial coordinates obtained after the calculation is defined as the virtual control boundary.

[0012] Furthermore, the thermal hysteresis time constant is obtained as follows: Obtain the lumped heat capacity parameter, call the transient thermal resistance factor, and apply the RC circuit response rule in thermoelectric simulation theory; The transient thermal resistance factor and the lumped thermal capacity parameter are coupled and multiplied, and the product result is used as the thermal hysteresis time constant.

[0013] Furthermore, the reverse energy cancellation process is performed as follows: Obtain the physical distance between the virtual control boundary and the coordinates of the heat flux inflection point, and use it as the damping range; A dynamic thermal balance equation is established, which includes laser input energy, total excess enthalpy, and damping range. The equivalent negative energy density required to make the heat accumulation increment of the molten pool zero after crossing the damping range is solved. Using the equivalent negative energy density as the objective function, a power decay trajectory exhibiting a nonlinear decreasing trend with welding displacement is constructed. The power attenuation trajectory is modulated and mapped with the output command of the laser generator to generate a predictive negative gain damping strategy.

[0014] Furthermore, the method for obtaining the total excess enthalpy is as follows: Call the thermal hysteresis time constant, take the current trigger time as the time end point, and backtrack by the duration of one thermal hysteresis time constant, which is defined as the thermal inertia integral window; Extract the historical residual heat value of the heat saturation matrix corresponding to the welding path coordinates within the thermal inertia integration window; The extracted historical residual calorific value is subjected to a time-dimensional cumulative integral operation, and the sum of the integrals is taken as the total excess enthalpy.

[0015] An intelligent control system for laser welding of housings includes the following modules: Heat flux acquisition module: used to acquire energy injection parameters and boundary temperature data along the welding path, and perform feature calculation to obtain heat flux distribution characteristics; based on the heat flux distribution characteristics, energy superposition processing is performed to establish a heat saturation matrix; Thermal resistance analysis module: used to perform spatial differentiation on the thermal saturation matrix to obtain the thermal response deviation between theoretical thermal conductivity and actual thermal diffusion; combined with the thermal response deviation, it performs inverse mapping analysis of material thermal properties to generate transient thermal resistance factor adapted to the current heat dissipation capacity; The mutation analysis module performs advanced rolling prediction of the thermal field evolution logic based on the transient thermal resistance factor, and generates the expected heat accumulation rate change curve; it extracts the coordinates of the heat flow inflection point where the rheological properties of the molten pool undergo abrupt changes based on the heat accumulation rate change curve. Boundary construction module: Based on the transient thermal resistance factor, the thermal response hysteresis of the heat flow inflection point coordinates is analyzed to obtain the thermal hysteresis time constant of energy input and temperature response; based on the thermal hysteresis time constant, backtracking mapping is performed to construct the virtual control boundary existing outside the physical inflection point; Welding control module: If the welding displacement triggers the virtual control boundary, the total amount of excess enthalpy accumulated inside the molten pool is calculated based on the heat saturation matrix; combined with the total amount of excess enthalpy, reverse energy cancellation processing is performed to generate a predictive negative gain damping strategy.

[0016] The beneficial effects of this invention are as follows: 1. By performing energy superposition processing in the spatiotemporal dimensions, a heat saturation matrix characterizing the overall thermal load state of the workpiece is constructed. By performing rolling integration of the instantaneous linear energy density and the heat retention factor with heat dissipation correction, the historical heating degree of each cell along the welding path is quantified, enabling the control system to perceive the workpiece's heat saturation state. Spatial differentiation and inverse mapping analysis are performed on the heat saturation matrix to generate a transient thermal resistance factor adapted to the current heat dissipation capacity. Utilizing the thermal response deviation between theoretical thermal conductivity and actual heat diffusion, a dynamic correlation between the deviation and material thermal properties is established, which helps identify the physical phenomenon of decreased thermal conductivity of the workpiece under heat saturation.

[0017] 2. By utilizing the transient thermal resistance factor to drive the thermal field evolution logic for advanced rolling simulation, and analyzing the extreme values ​​of the second derivative of the heat accumulation rate change curve, the coordinates of the heat flow inflection point where the molten pool rheological properties undergo abrupt changes are extracted. Simulating the future heat dissipation process within a virtual prediction time window helps identify the critical moment when the heat accumulation effect exceeds the material's heat dissipation limit, and transforms the time-dimensional instability warning into spatial-dimensional physical coordinates, enabling the predictive perception of abrupt changes in the molten pool rheological trend during the welding process.

[0018] 3. Thermal response hysteresis is analyzed based on the coordinates of the heat flow inflection point, constructing a virtual control boundary independent of the physical inflection point. By combining the RC circuit response law to calculate the thermal hysteresis time constant, the degree of asynchrony between energy input and temperature response is quantified, transforming the time-level hysteresis effect into a spatial-level lead compensation distance. By establishing a virtual control boundary, a buffer space to counteract thermal inertia is reserved in the control logic, which helps to reduce the problem of control intervention timing lag caused by thermal response delay.

[0019] 4. When the welding displacement triggers the virtual control boundary, the total excess enthalpy is calculated based on the thermal saturation matrix, and a reverse energy cancellation processing logic is constructed to generate a predictive negative gain damping strategy. By solving for the equivalent negative energy density, the accumulated heat value is quantified into a specific power attenuation amplitude, and a power trajectory exhibiting a nonlinear decreasing trend with displacement is constructed. This is beneficial for actively reducing energy input to offset thermal inertia pressure before the physical inflection point is reached, maintaining the morphological stability of the molten pool when traversing the heat accumulation region. Attached Figure Description

[0020] The invention will now be further described with reference to the accompanying drawings.

[0021] Figure 1 This is a flowchart of an intelligent control method for laser welding of a shell according to the present invention; Figure 2 This is a flowchart for determining whether there is asynchronous thermal response; Figure 3 This is a functional block diagram of an intelligent control system for laser welding of housings according to the present invention. Detailed Implementation

[0022] To make the technical means, creative features, objectives and effects of this invention easier to understand, the invention will be further described below in conjunction with specific embodiments.

[0023] Example 1 like Figure 1 As shown, an intelligent control method for laser welding of a housing includes the following steps: S1. Collect energy injection parameters and boundary temperature data along the welding path, and perform feature calculation to obtain heat flux distribution characteristics; perform energy superposition processing based on heat flux distribution characteristics to establish a heat saturation matrix; The method for obtaining the instantaneous heat flux distribution characteristics of the workpiece surface by collecting energy injection parameters and boundary temperature data along the welding path and performing feature calculation processing is as follows: Preferably, the laser power and welding speed at the current sampling moment are read in real time through the laser welding communication interface; Meanwhile, by using an infrared sensor array installed on the side of the welding head, the discrete temperatures of multiple sampling points covering the edge of the molten pool and the heat-affected zone at each sampling time are collected to construct a discrete temperature set; The timestamps of laser power, welding speed and discrete temperature were obtained respectively, and the three data parameters were time-aligned according to the timestamps. Obtain the light absorptivity constant of the current welding material, perform energy density analysis on the time-aligned laser power and welding speed to obtain the linear energy density at the current sampling moment, and use the linear energy density as the heat flux distribution characteristic; The preferred method for performing energy density analysis is as follows: The linear energy density at the current sampling moment is obtained by multiplying the light absorptivity constant with the laser power and then comparing the product with the welding speed. The method for establishing the heat saturation matrix based on energy superposition processing according to heat flow distribution characteristics is as follows: Obtain the geometric dimension model of the shell to be welded, and establish a virtual two-dimensional discretized mesh based on the geometric dimension model; Preferably, the method for establishing the two-dimensional discretized mesh is as follows: the spatial resolution of the mesh is set based on the laser spot diameter, and the shell surface is discretized into a grid composed of... A matrix composed of cells is constructed, and a mapping relationship is established between cell indices and the physical coordinates of the shell. Obtain the welding path and coordinates of the laser at the current sampling time; The current welding path coordinates are converted into target cell indices in the grid through a mapping relationship, and the calculated linear energy density is used as the heat increment value to fill the target cells. Identify the edge sampling point that is farthest from the center of the laser spot from the discrete temperature set, and mark the temperature value of the edge sampling point at the current sampling time as the reference temperature of the neighborhood environment. The relative temperature rise of each sampling point is obtained by calculating the difference between the temperature value of each sampling point in the discrete temperature set and the reference temperature of the surrounding environment. Obtain the physical distance between adjacent sampling points in the discrete temperature set, calculate the ratio of the difference in relative temperature rise between adjacent sampling points to the physical distance, and define the ratio as the boundary temperature gradient. The boundary temperature gradient is corrected for heat dissipation, and the retained component is obtained as the residual heat value. Preferably, the method for performing heat dissipation correction and obtaining the retained component as residual calorific value is as follows: establish a positive correlation function between heat dissipation rate and boundary temperature gradient, substitute the calculated boundary temperature gradient into the positive correlation function, and output... The heat dissipation attenuation coefficient is obtained by normalization. Preferably, through formula Construct a positive correlation function; in For heat dissipation rate; A is the thermal conductivity of the material (unit: W / (m·K)); A is the physical surface area of ​​a single cell in the two-dimensional discretized mesh (unit: ... ); The boundary temperature gradient (unit: K / m); The calculated heat dissipation rate The ratio is calculated with the preset theoretical maximum heat dissipation rate, and the normalized heat dissipation attenuation coefficient is output. Substitute the calculated boundary temperature gradient into the linear equation to output the normalized heat dissipation attenuation coefficient. The heat dissipation attenuation coefficient is reverse-mapped to generate a heat retention factor that characterizes the proportion of energy retained per unit time interval. It should be noted that the reverse mapping process involves subtracting the heat dissipation attenuation coefficient from 1. For a cell in a two-dimensional discretized grid, perform an update operation that combines the current input overlay with historical remnants. The update operation of input superposition and historical residual inheritance is as follows: extract the energy value stored in each cell of the grid at the previous sampling time, multiply the energy value with the heat retention factor, and obtain the residual heat value after natural heat loss. For the target cell mapped by the current welding path coordinates, the linear energy density filled in the target cell is used as the heat increment input, and numerical superposition is performed with the residual heat value calculated by the target cell to output the updated heat saturation value of the target cell at the current sampling time. For other non-target cells in the grid that are not covered by the current welding path coordinates, the residual heat value is directly updated to the heat saturation value at the current sampling time; By integrating the update results of all cells at N1 sampling times, a heat saturation matrix is ​​established to characterize the historical degree of heating. Preferably, N1=50.

[0024] S2. Perform spatial differentiation on the thermal saturation matrix to obtain the thermal response deviation between the theoretical thermal conductivity and the actual thermal diffusion; combine the thermal response deviation to perform inverse mapping analysis of the material's thermal properties to generate a transient thermal resistance factor that adapts to the current heat dissipation capacity. The method for obtaining the thermal response deviation between theoretical thermal conductivity and actual thermal diffusion by performing spatial differentiation on the thermal saturation matrix is ​​as follows: Perform spatial differentiation on the heat saturation matrix along the welding path to calculate the thermal potential energy gradient of the target cell relative to its adjacent cells. It should be noted that if the target cell is at the edge of the grid and there are no adjacent cells, the reference temperature of the neighborhood environment will be used as the value of the virtual adjacent cells, or the thermal potential gradient of the target cell will be directly assigned to zero. Obtain the standard thermal conductivity of the workpiece material under standard conditions at room temperature, and multiply the standard thermal conductivity by the thermal potential energy gradient to obtain the theoretical heat flux value under theoretical conditions. The boundary temperature gradient is then multiplied by a preset conversion coefficient to obtain the actual heat flux value, which has the same dimensions as the theoretical heat flux value. Calculate the absolute value of the difference between the theoretical heat flux value and the actual heat flux value, and define the absolute value of the difference as the thermal response deviation, which characterizes the degree of decay of heat conduction efficiency. The method for generating a transient thermal resistance factor that adapts to the current heat dissipation capacity is as follows: This involves combining thermal response deviation with inverse mapping analysis of material thermal properties. Obtain the basic thermal resistance value of the workpiece material and set the impedance sensitivity coefficient to characterize the temperature sensitivity of the material's thermal properties; Preferably, the method for obtaining the impedance sensitivity coefficient is as follows: Select a standard test plate of the same material and thickness, and conduct a set of gradient variable power welding experiments in advance; specifically, set the laser power to start at 80% of the rated power, and increase it to 120% in increments of 5%, while keeping the welding speed constant; record the data trend of the thermal response deviation increasing over time under different power levels; select the thermal response deviation as the abscissa and the corresponding material thermal resistance change rate as the ordinate, perform linear regression analysis, and obtain the slope of the regression line as the impedance sensitivity coefficient; It should be noted that the standard thermal conductivity and basic thermal resistance are obtained as follows: confirm the alloy grade of the shell material through the design drawings of the shell to be welded; query the physical property database of the alloy grade under standard room temperature environment, and extract the standard thermal conductivity of the material as the standard thermal conductivity. The reciprocal of the standard thermal conductivity is used to define the basic thermal resistance value, which reflects the initial thermal resistance of the material in a state without heat accumulation. Based on the impedance sensitivity coefficient, a linear mapping relationship is established between the thermal response deviation and the thermal resistance increment, and the thermal resistance increment is output. Preferably, the method for establishing a linear mapping relationship between thermal response deviation and thermal resistance increment is as follows: The impedance sensitivity coefficient is defined as the constant rate of change of thermal resistance caused by a unit thermal response deviation. The thermal response deviation is multiplied by the impedance sensitivity coefficient, and the result of the multiplication is assigned as the thermal resistance increment, thus realizing the construction of a positively correlated linear proportional relationship between the thermal response deviation and the thermal resistance increment. Calculate the sum of the base thermal resistance value and the thermal resistance increment to obtain the transient thermal resistance factor at the current moment; It is understandable that the transient thermal resistance factor is used to characterize the ability of the workpiece material to prevent heat from spreading to the surroundings at the current moment, and the transient thermal resistance factor increases with the increase of thermal response deviation; The function of generating a transient thermal resistance factor that adapts to the current heat dissipation capacity is as follows: Function 1: Corrects the static error in the traditional heat conduction model that assumes the thermal conductivity to be constant, and quantifies the dynamic decay of the workpiece's thermal conductivity caused by the thermal saturation effect into a real-time impedance variable that can participate in subsequent calculations. Secondly, it provides underlying parameters containing environmental nonlinear characteristics for advanced rolling simulation, which helps the simulation model to perceive the heat dissipation bottleneck that gradually increases as welding progresses, thereby improving the timeliness and accuracy of heat flow inflection point prediction.

[0025] Example 2 Please see Figure 1 As shown, an intelligent control method for laser welding of a housing includes the following steps: S3. Based on the transient thermal resistance factor, perform advanced rolling prediction of the thermal field evolution logic to generate the expected thermal accumulation rate change curve; extract the coordinates of the thermal flux inflection point where the rheological properties of the molten pool change abruptly occur based on the thermal accumulation rate change curve. The method for generating the expected heat accumulation rate change curve by performing forward rolling prediction of the thermal field evolution logic based on the transient thermal resistance factor is as follows: Establish a virtual prediction time window extending along the welding direction, assuming that the laser power and welding speed remain unchanged within the virtual prediction time window; The current linear energy density and transient thermal resistance factor are used as the basic parameters for the deduction; Within the virtual prediction time window, the basic parameters of the simulation are subjected to recursive iterative calculations including heat dissipation loss deduction in units of time step to calculate the virtual heat saturation at the next M1 time points and establish the heat saturation evolution sequence. Based on the solution of the first derivative of time in the heat saturation evolution sequence, a curve representing the rate of change of heat accumulation rate per unit time is established. Preferably, the recursive iterative calculation including heat dissipation loss deduction is performed as follows: extract the cell value corresponding to the current welding path coordinate in the heat saturation matrix, define it as the initial virtual heat saturation, and mark it as the current virtual heat saturation at the first time node; Divide the current virtual thermal saturation at the i-th time node by the transient thermal resistance factor to obtain the theoretical heat dissipation loss based on the current thermal potential energy. The current virtual heat saturation at the i-th time node is superimposed with the current linear energy density, and the theoretical heat dissipation loss is subtracted. The result is then assigned as the virtual heat saturation at the (i+1)-th time node. Starting from the initial virtual heat saturation, the process is executed M1 times in a single-step recursive logic until the entire virtual estimated time window is filled. The output contains the virtual heat saturation for the next M1 time nodes, forming a heat saturation evolution sequence. Preferably, M1=100; The method for extracting the coordinates of the heat flux inflection point where the rheological properties of the molten pool change abruptly based on the heat accumulation rate change curve is as follows: By differentiating the curve of the rate of heat accumulation with respect to time again, we obtain the second derivative curve characterizing the acceleration of heat accumulation. Scan the second derivative curve to identify the critical peak point where the value changes from its positive maximum value to a decrease. The critical peak point is defined as the moment when the heat accumulation effect exceeds the material's heat dissipation limit and causes instability. Obtain the relative time position of the instability moment within the virtual estimated time window, and multiply the relative time position with the current average welding speed to obtain the corresponding spatial warning distance; The average welding speed is calculated by taking the average welding speed of the three most recent sampling times. The spatial warning distance is superimposed on the current actual welding coordinates along the welding path direction to obtain the spatial location, which is the coordinate of the heat flow inflection point where the rheological properties of the molten pool change abruptly.

[0026] S4. Based on the transient thermal resistance factor, perform thermal response hysteresis analysis on the coordinates of the heat flow inflection point to obtain the thermal hysteresis time constant between energy input and temperature response; perform backtracking mapping based on the thermal hysteresis time constant to construct the virtual control boundary existing outside the physical inflection point; Among them, the method for obtaining the thermal hysteresis time constant of energy input and temperature response by analyzing the thermal response hysteresis of the heat flow inflection point coordinates based on the transient thermal resistance factor is as follows: The specific heat capacity constant and density constant of the workpiece material under constant pressure are obtained, and the effective thermal action volume characterizing the heat absorption capacity of the molten pool region is constructed based on the laser spot diameter and the preset depth of the molten pool. Preferably, the effective thermal effect volume is constructed as follows: the laser spot diameter and the preset depth of the molten pool are obtained, and the standard geometric volume of the molten pool is calculated; a heat-affected zone correction coefficient (with a value range of 1.2-1.5, preferably 1.3) is introduced, and the geometric volume of the molten pool is multiplied by the heat-affected zone correction coefficient to obtain the effective thermal effect volume characterizing the surrounding heat-affected zone; By performing a ternary product operation on the isobaric specific heat capacity constant, density constant and effective heat action volume, the lumped heat capacity parameter characterizing the upper limit of heat that the molten pool can hold is obtained; The transient thermal resistance factor is invoked, and based on the RC circuit response rule in thermoelectric simulation theory, the transient thermal resistance factor and the lumped thermal capacity parameter are coupled and multiplied. The product result is used as the thermal hysteresis time constant. It should be noted that the thermal hysteresis time constant is used to quantify the time delay required for the temperature field to reach a steady-state response after laser energy is injected into the workpiece. Among them, the method for constructing the virtual control boundary outside the physical inflection point by performing backtracking mapping based on the thermal hysteresis time constant is as follows: like Figure 2 As shown, a system response limit threshold is set, and the calculated thermal hysteresis time constant is compared with the system response limit threshold. If the thermal hysteresis time constant is less than the system response limit threshold, then the thermal inertia can be ignored, and the coordinates of the direct heat flow inflection point are defined as the virtual control boundary. If the thermal hysteresis time constant is greater than or equal to the system response limit threshold, it is determined that there is a risk of thermal response asynchrony; then the current real-time welding speed is extracted, and the thermal hysteresis time constant is multiplied by the average welding speed to obtain the spatial advance compensation distance used to offset the time delay. It is understandable that the sampling period of a single frame of the infrared sensor array is obtained; Obtain the algorithm processing cycle of the intelligent control system; Obtain the response delay time of the laser generator from receiving a command to the change in output power; The three time parameters (single frame sampling period, algorithm processing period response delay time) are summed, and a safety redundancy coefficient (preferably 1.1) is introduced. The calculated weighted sum is used as the system response limit threshold. The weights are calculated by consulting the product manuals of the infrared sensor, control system and laser generator, respectively, and the ratio of the maximum response time to the typical response time is used as the weight of the corresponding time parameter. Using the coordinates of the heat flow inflection point as the physical reference origin, perform coordinate subtraction along the opposite direction of the welding path to deduct the spatial advance compensation distance. The new set of spatial coordinates obtained after the calculation is defined as the virtual control boundary; The purpose of constructing a virtual control boundary that exists independently of physical inflection points is as follows: Objective 1: To reduce the time-domain asynchrony between energy input and temperature response, the unavoidable thermal response lag time caused by the thermal resistance-thermal capacity coupling effect is converted into a physical space advance compensation distance through welding speed mapping. Objective 2: To artificially construct a tactical intervention point in the control logic that is different from the physical instability point (heat flow inflection point), forcing the control system to intervene in advance before the thermal inertia is fully released, thus reserving a buffer space sufficient to offset the system hardware delay and heat conduction delay.

[0027] S5. If the welding displacement triggers the virtual control boundary, calculate the total amount of excess enthalpy accumulated inside the molten pool based on the heat saturation matrix; combine the total amount of excess enthalpy with reverse energy cancellation processing to generate a predictive negative gain damping strategy. If the welding displacement triggers the virtual control boundary, the method for calculating the total excess enthalpy accumulated inside the molten pool based on the heat saturation matrix is ​​as follows: It should be noted that the welding displacement triggering virtual control boundary means that the physical coordinates of the welded head monitored in real time fall into the unidirectional damping buffer zone formed by the virtual control boundary constructed by S4 as the starting point and the heat flow inflection point coordinates extracted by S3 as the ending point. Call the thermal hysteresis time constant, take the current trigger time as the time end point, and backtrack by the duration of one thermal hysteresis time constant, which is defined as the thermal inertia integral window; Extract the historical residual heat value of the heat saturation matrix corresponding to the welding path coordinates within the thermal inertia integration window; Perform a time-dimensional cumulative integral operation on the extracted historical residual calorific value, and the sum of the integrals is taken as the total excess enthalpy; Obtain the physical distance between the virtual control boundary and the coordinates of the heat flux inflection point, and use it as the damping range; A dynamic thermal balance equation is established, which includes laser input energy, total excess enthalpy, and damping range. The equivalent negative energy density required to make the heat accumulation increment of the molten pool zero after crossing the damping range is solved. Preferably, the method for constructing the dynamic thermal balance equation and solving for the equivalent negative energy density is as follows: setting a maximum allowable enthalpy value to maintain the molten pool from undergoing rheological abrupt changes. ; Based on the formula: Establish a dynamic thermal equilibrium equation and obtain the equivalent negative energy density. ; in, The total excess enthalpy, The length of the damping range; It should be noted that the equivalent negative linear energy density characterizes the magnitude by which the control system needs to reduce the linear energy density in the spatial dimension in order to offset the excess enthalpy. Among them, if If it equals zero, then directly set the equivalent negative linear energy density. Zero; Using the equivalent negative energy density as the objective function, a power decay trajectory exhibiting a nonlinear decreasing trend with welding displacement is constructed. The power attenuation trajectory is modulated and mapped with the output command of the laser generator to generate a predictive negative gain damping strategy. Preferably, the method for constructing the power attenuation trajectory and generating the predictive negative gain damping strategy is as follows: Extract the current real-time welding speed, multiply the equivalent negative linear energy density by the real-time welding speed, and obtain the power amplitude that needs to be reduced; The target power value is obtained by subtracting the power amplitude to be reduced from the current laser power value. A smooth transition curve between the reference value and the target power value is generated using an exponential decay function; Discretize the smooth transition curve into a power control instruction set as a predictive negative gain damping strategy. It should be noted that when the welding process employs a negative gain damping strategy, the laser power exhibits a damping characteristic of first decreasing sharply and then leveling off, which is used to release thermal inertial pressure in advance before the physical inflection point is reached. The purpose of the predictive negative gain damping strategy based on the total excess enthalpy is as follows: Objective 1: To separate the latent heat (excess enthalpy) accumulated inside the molten pool within the lag time window, and to reverse-calculate the energy fraction that must be forcibly removed, so as to prevent this part of the heat from continuing to heat the molten pool due to inertia after the light is turned off, which would cause the pool to collapse. Objective 2: By creating an artificial cold source effect through a nonlinear power decay trajectory, the evolution path of the heat accumulation acceleration is physically blocked, which helps to suppress the thermodynamic state of the molten pool below the critical threshold of abrupt changes in rheological properties when it passes through high-risk areas.

[0028] Example 3 Please see Figure 3 As shown, an intelligent control system for laser welding of housings includes the following modules: Heat flux acquisition module: used to acquire energy injection parameters and boundary temperature data along the welding path, and perform feature calculation to obtain heat flux distribution characteristics; based on the heat flux distribution characteristics, energy superposition processing is performed to establish a heat saturation matrix; Thermal resistance analysis module: used to perform spatial differentiation on the thermal saturation matrix to obtain the thermal response deviation between theoretical thermal conductivity and actual thermal diffusion; combined with the thermal response deviation, it performs inverse mapping analysis of material thermal properties to generate transient thermal resistance factor adapted to the current heat dissipation capacity; The mutation analysis module performs advanced rolling prediction of the thermal field evolution logic based on the transient thermal resistance factor, and generates the expected heat accumulation rate change curve; it extracts the coordinates of the heat flow inflection point where the rheological properties of the molten pool undergo abrupt changes based on the heat accumulation rate change curve. Boundary construction module: Based on the transient thermal resistance factor, the thermal response hysteresis of the heat flow inflection point coordinates is analyzed to obtain the thermal hysteresis time constant of energy input and temperature response; based on the thermal hysteresis time constant, backtracking mapping is performed to construct the virtual control boundary existing outside the physical inflection point; Welding control module: If the welding displacement triggers the virtual control boundary, the total amount of excess enthalpy accumulated inside the molten pool is calculated based on the heat saturation matrix; combined with the total amount of excess enthalpy, reverse energy cancellation processing is performed to generate a predictive negative gain damping strategy.

[0029] The foregoing has provided a detailed description of one embodiment of the present invention, but this description is merely a preferred embodiment and should not be construed as limiting the scope of the invention. All equivalent variations and modifications made within the scope of the present invention should still fall within the scope of the present invention.

Claims

1. An intelligent control method for laser welding of housings, characterized in that: Includes the following steps: The thermal saturation matrix is ​​obtained and spatial differentiation is performed to obtain the thermal response deviation between the theoretical thermal conductivity and the actual thermal diffusion. The thermal response deviation is combined with the inverse mapping analysis of the material's thermal properties to generate a transient thermal resistance factor that adapts to the current heat dissipation capacity. Based on the transient thermal resistance factor, the thermal field evolution logic is predicted in advance to generate the expected thermal accumulation rate change curve; based on the thermal accumulation rate change curve, the coordinates of the thermal flux inflection point where the rheological properties of the molten pool change abruptly are extracted. Based on the transient thermal resistance factor, the thermal response hysteresis of the heat flow inflection point coordinates is analyzed to obtain the thermal hysteresis time constant of energy input and temperature response. Based on the thermal hysteresis time constant, backtracking mapping is performed to construct a virtual control boundary that exists outside the physical inflection point; If the welding displacement triggers the virtual control boundary, the total amount of excess enthalpy accumulated inside the molten pool is calculated based on the heat saturation matrix. By combining the total excess enthalpy with reverse energy cancellation, a predictive negative gain damping strategy is generated.

2. The intelligent control method for laser welding of a shell according to claim 1, characterized in that: The method for obtaining the heat saturation matrix is ​​as follows: Energy injection parameters and boundary temperature data along the welding path are collected and feature calculations are performed to obtain heat flux distribution characteristics. Based on the heat flux distribution characteristics, energy superposition processing is performed to establish a heat saturation matrix.

3. The intelligent control method for laser welding of a shell according to claim 1, characterized in that: The method for performing reverse mapping analysis is as follows: Obtain the basic thermal resistance value of the workpiece material and set the impedance sensitivity coefficient to characterize the temperature sensitivity of the material's thermal properties; Based on the impedance sensitivity coefficient, a linear mapping relationship is established between the thermal response deviation and the thermal resistance increment, and the thermal resistance increment is output. The transient thermal resistance factor at the current moment is obtained by calculating the sum of the base thermal resistance value and the thermal resistance increment.

4. The intelligent control method for laser welding of a shell according to claim 3, characterized in that: The linear mapping relationship is established as follows: The impedance sensitivity coefficient is used as the constant of the rate of change of thermal resistance caused by a unit thermal response deviation. The thermal response deviation is multiplied by the impedance sensitivity coefficient, and the result of the multiplication is assigned as the thermal resistance increment.

5. The intelligent control method for laser welding of a shell according to claim 1, characterized in that: The method for extracting the coordinates of the heat flux inflection point is as follows: By differentiating the curve of the rate of heat accumulation with respect to time again, we obtain the second derivative curve characterizing the acceleration of heat accumulation. Scan the second derivative curve to identify the critical peak point where the value changes from its positive maximum value to a decrease. The critical peak point is defined as the moment when the heat accumulation effect exceeds the material's heat dissipation limit and causes instability. Obtain the relative time position of the instability moment within the virtual estimated time window, multiply the relative time position by the current welding speed, and obtain the corresponding spatial warning distance. The spatial warning distance is superimposed on the current actual welding coordinates along the welding path direction to obtain the spatial location, which is the coordinate of the heat flow inflection point where the rheological properties of the molten pool change abruptly.

6. The intelligent control method for laser welding of a shell according to claim 1, characterized in that: The virtual control boundary is constructed as follows: The thermal hysteresis time constant is obtained and compared to obtain the results of the risk of asynchronous thermal response, and the current real-time welding speed is extracted. The thermal hysteresis time constant is multiplied by the average welding speed to obtain the spatial advance compensation distance used to offset the time delay. Using the coordinates of the heat flow inflection point as the physical reference origin, perform coordinate subtraction along the opposite direction of the welding path to deduct the spatial advance compensation distance. The new set of spatial coordinates obtained after the calculation is defined as the virtual control boundary.

7. The intelligent control method for laser welding of a shell according to claim 6, characterized in that: The method for obtaining the thermal hysteresis time constant is as follows: Obtain the lumped heat capacity parameter, call the transient thermal resistance factor, and apply the RC circuit response rule in thermoelectric simulation theory; The transient thermal resistance factor and the lumped thermal capacity parameter are coupled and multiplied, and the product result is used as the thermal hysteresis time constant.

8. The intelligent control method for laser welding of a shell according to claim 1, characterized in that: The reverse energy cancellation process is performed as follows: Obtain the physical distance between the virtual control boundary and the coordinates of the heat flux inflection point, and use it as the damping range; A dynamic thermal balance equation is established, which includes laser input energy, total excess enthalpy, and damping range. The equivalent negative energy density required to make the heat accumulation increment of the molten pool zero after crossing the damping range is solved. Using the equivalent negative energy density as the objective function, a power decay trajectory exhibiting a nonlinear decreasing trend with welding displacement is constructed. The power attenuation trajectory is modulated and mapped with the output command of the laser generator to generate a predictive negative gain damping strategy.

9. The intelligent control method for laser welding of a shell according to claim 8, characterized in that: The method for obtaining the total excess enthalpy is as follows: Call the thermal hysteresis time constant, take the current triggering time as the time end point, and backtrack by the duration of one thermal hysteresis time constant, which is defined as the thermal inertia integral window; Extract the historical residual heat value of the heat saturation matrix corresponding to the welding path coordinates within the thermal inertia integration window; The extracted historical residual calorific value is subjected to a time-dimensional cumulative integral operation, and the sum of the integrals is taken as the total excess enthalpy.

10. An intelligent control system for laser welding of housings, used to implement the intelligent control method for laser welding of housings as described in any one of claims 1-9, characterized in that: Includes the following modules: Heat flux acquisition module: used to acquire energy injection parameters and boundary temperature data along the welding path, and perform feature calculation to obtain heat flux distribution characteristics; based on the heat flux distribution characteristics, energy superposition processing is performed to establish a heat saturation matrix; Thermal resistance analysis module: used to perform spatial differentiation on the thermal saturation matrix to obtain the thermal response deviation between theoretical thermal conductivity and actual thermal diffusion; By combining thermal response deviation with inverse mapping analysis of material thermal properties, a transient thermal resistance factor adapted to the current heat dissipation capacity is generated. The mutation analysis module performs advanced rolling prediction of the thermal field evolution logic based on the transient thermal resistance factor, and generates the expected heat accumulation rate change curve; it extracts the coordinates of the heat flow inflection point where the rheological properties of the molten pool undergo abrupt changes based on the heat accumulation rate change curve. Boundary construction module: Based on the transient thermal resistance factor, the thermal response hysteresis of the heat flow inflection point coordinates is analyzed to obtain the thermal hysteresis time constant of energy input and temperature response; based on the thermal hysteresis time constant, backtracking mapping is performed to construct the virtual control boundary existing outside the physical inflection point; Welding control module: If the welding displacement triggers the virtual control boundary, the total amount of excess enthalpy accumulated inside the molten pool is calculated based on the heat saturation matrix; By combining the total excess enthalpy with reverse energy cancellation, a predictive negative gain damping strategy is generated.

Images