Continuous fiber additive manufacturing process parameter optimization control system

Abstract

The invention relates to the technical field of additive manufacturing, and particularly discloses a continuous fiber additive manufacturing process parameter optimization control system, which is characterized in that a curvature and a curvature change rate are extracted along a printing path, and a continuous micro-dispatchable function of a printing speed, a fiber tension and an extrusion multiplying power about an arc length is constructed; taking fiber strain energy abrupt change and resin stacking volume deviation minimization in the transition area as targets, and obtaining an optimal speed, tension and extrusion rate curve through offline nesting optimization; discretizing the curve into increment instructions according to a speed change section, and embedding the increment instructions into printing codes; actual tension and stacking width are collected in real time during execution, and a step length factor is dynamically adjusted based on the main deviation for compensation and correction; according to the method, continuous and smooth parameter scheduling and closed-loop compensation in the arc length direction are achieved, and the defects of fiber wrinkles and resin knots in the curve transition area are effectively overcome.

Description

Technical Field

[0001] This invention relates to the field of additive manufacturing technology, and more specifically to a parameter optimization and control system for continuous fiber additive manufacturing process. Background Technology

[0002] Continuous fiber additive manufacturing is an advanced composite material molding technology that simultaneously extrudes and deposits continuous fiber reinforcements with a thermoplastic resin matrix. It boasts advantages such as excellent mechanical properties and strong structural design flexibility, and is widely used in aerospace, automotive lightweighting, and drone structural components. During the printing process, the matching of process parameters such as printing speed, fiber tension, and extrusion ratio directly affects the fiber spread morphology, resin wetting quality, and interlayer bonding strength.

[0003] In the continuous fiber additive manufacturing process, how to eliminate the coupling interference between the sudden change in fiber strain energy and the resin accumulation volume deviation caused by the continuous change in path curvature in the transition zone of the curve, and to achieve continuous and differentiable synchronous scheduling of printing speed, fiber tension and extrusion ratio along the arc length direction, so as to suppress the fiber wrinkles and resin nodules defects that cannot be avoided under the traditional regional parameter switching method. Summary of the Invention

[0004] The purpose of this invention is to provide a parameter optimization and control system for continuous fiber additive manufacturing process to solve the problems mentioned above.

[0005] The objective of this invention can be achieved through the following technical solutions: A parameter optimization and control system for continuous fiber additive manufacturing process includes: The path curvature and rate of change extraction module discretizes the sampling along the arc length direction of the path to be printed, calculates the path curvature and its first derivative at each sampling point, and obtains the curvature curve and the rate of change curve of curvature. The continuous differentiable scheduling function construction module uses the curvature curve and the curvature change rate curve as independent variables to construct continuous differentiable scheduling functions for printing speed, fiber tension, and extrusion ratio with respect to arc length. This allows the scheduling function to automatically return to the baseline value in the straight section when the curvature approaches zero, and smoothly transition to the set value in the curved section when the curvature increases. It also realizes pre-deceleration and pre-tension increase before the curve based on the positive or negative curvature change rate. The offline multi-objective constraint optimization module uses the scheduling function as the decision variable and minimizes the fiber strain energy mutation and resin packing volume deviation in the transition zone as the joint optimization objectives. It solves the constraint optimization problem offline within the feasible domain of printing speed, fiber tension and extrusion ratio to obtain the optimal speed curve, tension curve and extrusion ratio curve. The discretization instruction generation and embedding module discretizes the speed curve, tension curve, and extrusion ratio curve according to a preset sampling period, and converts them into continuous parameter control instructions distributed along the arc length, which are then embedded into the printing motion code. The real-time deviation compensation and correction module collects the actual fiber tension and resin buildup width in real time during the execution of the printing motion code. It compares these values ​​with the expected values ​​of the speed curve, tension curve, and extrusion ratio curve one by one. When the deviation exceeds the threshold, it performs a small-amplitude compensation and correction on the current control command and uses the corrected command as the control reference for the subsequent arc length position.

[0006] As a further aspect of the present invention: the calculation of the path curvature and its first derivative at each sampling point to obtain the curvature curve and the rate of change of curvature curve specifically includes: The path to be printed is divided into equal arc length segments and adaptive encryption segments based on the curvature change rate threshold. Fixed large step size sampling is used in equal arc length segments, while variable step size sampling is used in adaptive encryption segments, which decreases as the curvature change rate increases. For each sampling point, take the two adjacent sampling points before and after it to form a three-point circle, and calculate the reciprocal of the radius of the three-point circle as the path curvature of the corresponding sampling point; The obtained curvature sequence is subjected to a seven-point quadratic smoothing filter to obtain the curvature curve, and the curvature curve is subjected to central difference to obtain the curvature change rate curve.

[0007] As a further aspect of the present invention: the construction process of the continuously differentiable schedulable function is as follows: The total curvature change is obtained by numerically integrating the curvature change rate curve. The total curvature change is then mapped to the negative one and positive one intervals via the arctangent function to obtain the pre-adjustment direction factor. The pre-adjustment amplitude factor is obtained by power-calculating the ratio of the curvature value at each point on the curvature curve to the preset curvature threshold. The printing speed is set as a weighted sum of the baseline value in the straight section and the set value in the curved section, where the weight is the product of the pre-adjustment direction factor and the pre-adjustment amplitude factor. The fiber tension and extrusion ratio are weighted and summed according to the same weighting rule, so that when the curvature change rate is positive, pre-deceleration and pre-tension increase are achieved, and when the curvature change rate is negative, slow acceleration and slow tension decrease are achieved.

[0008] As a further aspect of the present invention: obtaining the optimal speed curve, tension curve, and extrusion ratio curve specifically includes: The transition zone is discretized into multiple sub-intervals. Within each sub-interval, the values ​​of the scheduling function at both ends are used as design variables to calculate the fiber strain energy mutation value and the resin packing volume deviation value respectively. The inner and outer nested optimization method is adopted. The inner layer uses the interval elimination method with halved step size to successively narrow the feasible domain of printing speed, while the outer layer uses the golden section method to simultaneously search for the optimal combination of fiber tension and extrusion ratio. Substitute the printing speed, fiber tension, and extrusion ratio obtained from each optimization into the joint optimization objective. When the relative change of the objective function value between two consecutive iterations is less than the preset convergence threshold, stop the iteration and output the current speed curve, tension curve, and extrusion ratio curve.

[0009] As a further aspect of the present invention: the outer layer simultaneously searches for the optimal combination of fiber tension and extrusion ratio using the golden ratio method, specifically including: The initial search ranges for fiber tension and extrusion ratio are set separately, and two golden section points are taken in each of the two ranges to generate four two-dimensional candidate points. Calculate the joint optimization objective value for each candidate point, eliminate the sub-interval containing the candidate point with the largest objective value, and form a new two-dimensional search interval by combining the regions containing the remaining three candidate points. Repeat steps one and two until the length of the fiber tension interval and the length of the extrusion ratio interval are both less than their respective preset accuracy thresholds. Take the midpoint of the final interval as the optimal fiber tension and the optimal extrusion ratio.

[0010] As a further aspect of the present invention: the sequential conversion into continuous parameter control commands distributed along the arc length specifically includes: The arc length to be printed is divided into multiple variable speed sections based on the curvature change rate curve. The first sampling period is used in the section with a large absolute value of curvature change rate, and the second sampling period is used in the section with a small absolute value of curvature change rate. The first sampling period is shorter than the second sampling period. At each sampling moment, the speed value, tension value, and extrusion ratio value of the current sampling point are subtracted from the corresponding values ​​of the previous sampling point to obtain the speed increment, tension increment, and extrusion ratio increment. After arranging the speed increment, tension increment, and extrusion ratio increment in a fixed order, they are sequentially written into the parameter extension field following the current line of motion instructions in the printed motion code.

[0011] As a further aspect of the present invention: the compensation and correction process is as follows: The difference between the real-time collected fiber tension and the expected tension is divided by the expected tension to obtain the relative tension deviation, and the difference between the actual value of the resin packing width and the expected width is divided by the expected width to obtain the relative width deviation. The absolute values ​​of the tension relative deviation and the width relative deviation are compared respectively. If the former is greater than or equal to the latter, the tension relative deviation is taken as the main deviation; otherwise, the width relative deviation is taken as the main deviation. When the absolute value of the main deviation exceeds the preset threshold, the compensation direction is determined according to the positive or negative sign of the main deviation, and the absolute value of the main deviation is multiplied by a fixed step size factor to obtain the compensation step size. The current printing speed command, fiber tension command, and extrusion ratio command are adjusted synchronously according to the compensation direction and compensation step size. The adjusted command set is used as the control reference for the subsequent arc length position. At the same time, the main deviation value is accumulated to the deviation history record to adjust the fixed step size factor for the next step.

[0012] As a further aspect of the present invention: the calculation process of the fixed step size factor is as follows: The five most recent major deviation values ​​are stored in memory to form a deviation sliding window. Each time a new major deviation value is obtained, the earliest deviation value in the window is moved out and the new value is moved in. Calculate the sum of the absolute values ​​of all major deviations within the sliding window, and then divide by five to obtain the average deviation amplitude; The initial adjustment value is obtained by dividing the preset baseline step size factor by the average deviation amplitude. If the initial adjustment value is less than the minimum step size limit, the minimum step size limit is taken. If it is greater than the maximum step size limit, the maximum step size limit is taken. The final value is used as the fixed step size factor for the next compensation correction.

[0013] The beneficial effects of this invention are: (1) By constructing a continuously differentiable scheduling function based on path curvature and its rate of change, and combining offline multi-objective optimization and online closed-loop compensation, the fiber wrinkles and resin accumulation defects in the curve transition zone are effectively suppressed, and the molding accuracy and interlayer bonding quality of continuous fiber reinforced composite components are improved.

[0014] (2) An adaptive variable step size sampling and variable speed section discretization strategy is adopted, combined with a compensation step size factor based on dynamic adjustment of sliding window. This reduces the consumption of computing resources while improving the response sensitivity of parameter control to complex path geometric features, thereby enhancing the robustness and engineering applicability of the process optimization method. Attached Figure Description

[0015] The invention will now be further described with reference to the accompanying drawings.

[0016] Figure 1 This is a system block diagram of the present invention. Detailed Implementation

[0017] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0018] Please see Figure 1 As shown, the present invention is a parameter optimization and control system for continuous fiber additive manufacturing process, comprising: The path curvature and rate of change extraction module discretizes the sampling along the arc length direction of the path to be printed, calculates the path curvature and its first derivative at each sampling point, and obtains the curvature curve and the rate of change curve of curvature. The continuous differentiable scheduling function construction module uses the curvature curve and the curvature change rate curve as independent variables to construct continuous differentiable scheduling functions for printing speed, fiber tension, and extrusion ratio with respect to arc length. This allows the scheduling function to automatically return to the baseline value in the straight section when the curvature approaches zero, and smoothly transition to the set value in the curved section when the curvature increases. It also realizes pre-deceleration and pre-tension increase before the curve based on the positive or negative curvature change rate. The offline multi-objective constraint optimization module uses the scheduling function as the decision variable and minimizes the fiber strain energy mutation and resin packing volume deviation in the transition zone as the joint optimization objectives. It solves the constraint optimization problem offline within the feasible domain of printing speed, fiber tension and extrusion ratio to obtain the optimal speed curve, tension curve and extrusion ratio curve. The discretization instruction generation and embedding module discretizes the speed curve, tension curve, and extrusion ratio curve according to a preset sampling period, and converts them into continuous parameter control instructions distributed along the arc length, which are then embedded into the printing motion code. The real-time deviation compensation and correction module collects the actual fiber tension and resin buildup width in real time during the execution of the printing motion code. It compares these values ​​with the expected values ​​of the speed curve, tension curve, and extrusion ratio curve one by one. When the deviation exceeds the threshold, it performs a small-amplitude compensation and correction on the current control command and uses the corrected command as the control reference for the subsequent arc length position.

[0019] In the path curvature and rate of change extraction module, discrete sampling is performed along the arc length direction of the path to be printed, and the path curvature and its first derivative at each sampling point are calculated to obtain the curvature curve and the rate of change curve of curvature, specifically including: First, the path to be printed is divided into equal-arc-length segments and adaptive encryption segments based on a preset curvature change rate threshold. The curvature change rate threshold is a pre-set value based on printing accuracy requirements, for example, set to 0.05 radians per millimeter of arc length. On the path to be printed, the curvature change rate at each point of the current arc length is calculated. If this value is less than or equal to the threshold, the path segment is classified as an equal-arc-length segment, and sampling is performed within this segment using a fixed large step size, set to 1 millimeter. If the curvature change rate is greater than the threshold, the path segment is classified as an adaptive encryption segment, and sampling is performed within this segment using a variable step size. The step size decreases as the curvature change rate increases; specifically, one-tenth of the fixed large step size is used as the minimum step size, and the actual step size is equal to the fixed large step size divided by the ratio of the curvature change rate to the threshold, then compared with the minimum step size, and the larger value is taken.

[0020] Secondly, for each sampling point, a three-point circle is formed by taking the two adjacent sampling points before and after it. The reciprocal of the radius of this three-point circle is used as the path curvature of that sampling point. Specifically, for the current sampling point and its preceding and following sampling points, the spatial coordinates of the three points are extracted. First, the first distance between the preceding and current sampling points, the second distance between the current and following sampling points, and the third distance between the preceding and following sampling points are calculated. Then, the half-perimeter is calculated, which is half the sum of the first, second, and third distances. Next, the area of ​​the triangle is calculated, which is the difference between the half-perimeter multiplied by the half-perimeter and the first, second, and third distances respectively, and then the square root of the product of these four numbers is taken. Finally, the radius of the three-point circle is equal to the product of the first, second, and third distances divided by four times the area of ​​the triangle; the curvature is the reciprocal of this radius. For the path start and end points, since it is not possible to obtain the adjacent points before and after simultaneously, the curvature approximation is obtained by taking the two adjacent points on one side of the path to form a circle or by using the curvature approximation of the line connecting the two adjacent points.

[0021] Then, a seven-point quadratic smoothing filter is applied to the obtained curvature sequence to obtain the curvature curve. The specific steps of this smoothing filter are as follows: For each curvature point to be smoothed in the sequence, take the point itself, its three preceding points, and its three following points, for a total of seven consecutive points (if there are fewer than seven points due to proximity to the end of the sequence, the endpoint values ​​are symmetrically repeated to make up the seven points). A quadratic polynomial is used to perform least-squares fitting between the position indices of these seven points (with the current point index as 0, the indices of the preceding three points as -1, -2, and -3 respectively, and the indices of the following three points as +1, +2, and +3 respectively) and the corresponding curvature values ​​to obtain the three coefficients of the quadratic polynomial. Then, the index 0 is substituted into the quadratic polynomial, and the calculated value is the smoothed curvature. Each curvature point is processed sequentially in the above manner, and all the smoothed curvature values ​​arranged in their original order constitute the curvature curve.

[0022] Finally, the curvature curve is centrally differencing to obtain the curvature change rate curve. The specific calculation method is as follows: For each internal point on the curvature curve, the curvature value at the next sampling point is subtracted from the curvature value at the previous sampling point, and then divided by the arc length interval between the previous and next sampling points (i.e., the arc length step between the previous and current sampling points plus the arc length step between the current and next sampling points). For the starting point on the curvature curve, the curvature change rate is the difference in curvature between the starting point and the next point divided by the arc length step between the two points; for the ending point, the curvature change rate is the difference in curvature between the ending point and the previous point divided by the arc length step between the two points. All calculated curvature change rate values ​​arranged in arc length order constitute the curvature change rate curve.

[0023] In the continuously differentiable schedulable function construction module, using the curvature curve and the rate of change of curvature curve as independent variables, continuously differentiable schedulable functions are constructed for printing speed, fiber tension, and extrusion ratio, each with respect to arc length. This allows the schedulable function to automatically revert to the baseline value in the straight section when the curvature approaches zero, and to smoothly transition to the set value in the curved section when the curvature increases. Furthermore, it implements pre-deceleration and pre-tension increase before the curve based on the sign of the rate of change of curvature. Specifically, this includes: First, the pre-adjustment direction factor is calculated based on the curvature change rate curve. Numerical integration is performed on the curvature change rate curve to obtain the total curvature change from the printing start point to the current arc length position. The numerical integration uses the trapezoidal rule, which multiplies the average curvature change rate of two adjacent sampling points on the curvature change rate curve by the arc length step between the two points, and then sums it to the previous integration result. The obtained total curvature change is used as the independent variable of the arctangent function, whose output range is between -1 / 1000π and +1 / 1000π. The output value is then divided by half of π to ensure the mapping result falls within the negative to positive range. This mapping result is the pre-adjustment direction factor. When the total curvature change is positive, the pre-adjustment direction factor is positive, indicating that pre-deceleration and pre-tension increase are required; when the total curvature change is negative, the pre-adjustment direction factor is negative, indicating that gradual acceleration and gradual tension decrease are required.

[0024] Secondly, the pre-adjustment amplitude factor is calculated based on the curvature curve. A curvature threshold is preset, which is 0.2 radians per millimeter of arc length, representing the critical curvature between straight and curved sections. For each sampling point on the curvature curve, the curvature value at that point is divided by the curvature threshold to obtain a ratio. This ratio is then raised to the power of three (multiplied by itself) to obtain the pre-adjustment amplitude factor. When the curvature value is less than the curvature threshold, the pre-adjustment amplitude factor is between 0 and 1; when the curvature value is greater than the curvature threshold, the pre-adjustment amplitude factor is greater than 1.

[0025] Then, set the baseline values ​​for the straight section and the setting values ​​for the curved section respectively. The baseline values ​​for the straight section include the straight section printing speed, the straight section fiber tension, and the straight section extrusion ratio, where the straight section printing speed is set to 30 mm / s, the straight section fiber tension is set to 5 N / s, and the straight section extrusion ratio is set to 1.0. The setting values ​​for the curved section include the curved section printing speed, the curved section fiber tension, and the curved section extrusion ratio, where the curved section printing speed is set to 10 mm / s, the curved section fiber tension is set to 15 N / s, and the curved section extrusion ratio is set to 1.2.

[0026] Finally, a continuously differentiable scheduling function is constructed. For printing speed, it is set as a weighted sum of the straight-line printing speed and the curved-line printing speed, with the weight being the product of the pre-adjustment direction factor and the pre-adjustment amplitude factor. Specifically, the calculation method is as follows: first, calculate the product of the pre-adjustment direction factor and the pre-adjustment amplitude factor to obtain the combined weight; then, calculate the difference between the curved-line printing speed and the straight-line printing speed, multiply this difference by the combined weight, and add it to the straight-line printing speed to obtain the printing speed at the current arc length position. For fiber tension, the calculation is performed in the same way: first, calculate the difference between the curved-line fiber tension and the straight-line fiber tension, multiply this difference by the combined weight, and add it to the straight-line fiber tension to obtain the fiber tension at the current arc length position. For extrusion ratio, similarly, calculate the difference between the curved-line extrusion ratio and the straight-line extrusion ratio, multiply it by the combined weight, and add it to the straight-line extrusion ratio to obtain the extrusion ratio at the current arc length position. The above calculations ensure that the printing speed, fiber tension, and extrusion ratio all change continuously and differentially with the arc length. When the curvature approaches zero, the combined weight approaches zero, and the three automatically return to the baseline value in the straight section. When the curvature increases, the combined weight increases, and the three smoothly transition to the set value in the curved section.

[0027] In the offline multi-objective constrained optimization module, the scheduling function is used as the decision variable, and the joint optimization objectives are minimizing the abrupt change in fiber strain energy and minimizing the resin packing volume deviation within the transition zone. The constrained optimization problem is solved offline within the feasible region of printing speed, fiber tension, and extrusion ratio to obtain the optimal speed curve, tension curve, and extrusion ratio curve. Specifically, this includes: First, the transition zone is discretized into multiple sub-intervals. The transition zone refers to the complete arc length from the start of the straight section into the curve until the end of the curve and the return to the straight section. This transition zone is uniformly divided into N sub-intervals according to the arc length, where N is 20. The printing speed, fiber tension, and extrusion ratio at the endpoints of each sub-interval have been obtained from the aforementioned scheduling function. Within each sub-interval, the values ​​of the scheduling function at the two endpoints of the sub-interval are used as design variables to calculate the fiber strain energy mutation value and the resin bulk volume deviation value, respectively.

[0028] The fiber strain energy abrupt change value is calculated as follows: For each sub-interval, the fiber tension at the left endpoint is denoted as... The fiber tension at the right end is The printing speed at the left endpoint is The printing speed at the right endpoint is Then the fiber strain energy abrupt change value It is obtained using the following mathematical formula: ; in, and The unit is Newton. and The unit is millimeters per second. The unit is millijoules. This formula characterizes the abrupt change in the strain energy inside the fiber caused by the coupling change of fiber tension and printing speed, and the absolute value calculation ensures that the abrupt change value is non-negative.

[0029] The resin packing volume deviation is calculated as follows: For each sub-interval, first calculate the theoretical extrusion volume, which is equal to the extrusion ratio E multiplied by the printing speed v, then multiplied by the fiber cross-sectional area constant (0.05 mm²), and then multiplied by the sub-interval arc length step Δs. Simultaneously, calculate the actual spreading volume, which is equal to the measured resin packing width (preset desired width 1.2 mm) multiplied by the printing layer height (0.2 mm), and then multiplied by the sub-interval arc length step Δs. Subtract the actual spreading volume from the theoretical extrusion volume and take the absolute value to obtain the resin packing volume deviation value ΔV for that sub-interval. Add up the ΔU and ΔV values ​​for all sub-intervals within the transition zone to obtain the total fiber strain energy mutation in the transition zone. Total deviation of resin packing volume The joint optimization objective is defined as follows: and The weighted sum, where The weight is set to 0.6. The weight is set to 0.4.

[0030] Secondly, a nested inner and outer optimization approach is used to find the minimum value of the joint optimization objective. The inner optimization targets the printing speed variable, while the outer optimization targets the fiber tension and extrusion ratio variables. The inner optimization uses a step-by-step interval elimination method to gradually narrow down the feasible region of the printing speed. Specifically, the initial feasible region of the printing speed is set to a lower limit of 5 mm / s and an upper limit of 40 mm / s. Within the current search interval, the interval is divided into left and right halves by taking the midpoint of the interval, and the joint optimization objective value corresponding to the midpoint of the left and right halves is calculated respectively (at this time, the fiber tension and extrusion ratio are fixed to the current candidate values ​​of the outer layer). The two objective values ​​are compared, and the half with the smaller objective value is retained as the new search interval, while the search step size is halved. The above process is repeated until the length of the search interval is less than 0.1 mm / s, and the printing speed at the midpoint of the final interval is taken as the current optimal printing speed.

[0031] The outer layer optimization employs the golden section method to simultaneously search for the optimal combination of fiber tension and extrusion ratio. The initial search interval for fiber tension is set to a lower limit of 2 Newtons and an upper limit of 20 Newtons, while the initial search interval for extrusion ratio is set to a lower limit of 0.8 and an upper limit of 1.5. Within each interval, two dividing points are selected according to the golden ratio of 0.618. For the fiber tension interval, the left golden section point equals the lower limit plus 0.382 multiplied by the interval length, and the right golden section point equals the lower limit plus 0.618 multiplied by the interval length. The left and right golden section points for the extrusion ratio interval are obtained in the same way. The two dividing points for fiber tension and the two dividing points for extrusion ratio are then paired to generate four two-dimensional candidate points: (left point of fiber tension, left point of extrusion ratio), (left point of fiber tension, right point of extrusion ratio), (right point of fiber tension, left point of extrusion ratio), and (right point of fiber tension, right point of extrusion ratio).

[0032] For each candidate point, its fiber tension and extrusion ratio are fixed, and the inner layer optimization is called to obtain the optimal printing speed and the corresponding joint optimization objective value. The objective values ​​corresponding to the four candidate points are compared, and the candidate point with the largest objective value is found. The sub-interval containing this candidate point will be eliminated. The specific elimination rules are as follows: if the fiber tension corresponding to the candidate point with the largest objective value is located at the left dividing point, then the left half of the fiber tension interval is eliminated; otherwise, the right half is eliminated. Similarly, the corresponding half of the extrusion ratio interval is eliminated based on the dividing point position of the extrusion ratio. The regions containing the remaining three candidate points constitute a new two-dimensional search interval, namely the new fiber tension interval and the new extrusion ratio interval.

[0033] Repeat the above process of calculating the golden ratio point, generating candidate points, comparing target values, and eliminating intervals until the length of the fiber tension interval is less than 0.1 Newtons and the length of the extrusion ratio interval is less than 0.01. At this point, take the midpoint of the final fiber tension interval as the optimal fiber tension and the midpoint of the final extrusion ratio interval as the optimal extrusion ratio, and call the inner layer optimization to obtain the corresponding optimal printing speed.

[0034] Finally, the optimal printing speed, optimal fiber tension, and optimal extrusion ratio obtained from each outer layer optimization iteration are substituted into the joint optimization objective function to calculate the objective function value for the current iteration. Iteration stops when the relative change in the objective function value between two consecutive iterations is less than a preset convergence threshold. The preset convergence threshold is 0.001, and the relative change is calculated as: the absolute value of the current objective function value minus the absolute value of the previous objective function value, divided by the previous objective function value. After stopping the iteration, the speed curve, tension curve, and extrusion ratio curve obtained from the current iteration are output; that is, the curves formed by arranging the optimal printing speed, optimal fiber tension, and optimal extrusion ratio at the endpoints of each sub-interval in the entire transition zone in order of arc length.

[0035] In the discretization instruction generation and embedding module, the speed curve, tension curve, and extrusion ratio curve are discretized according to a preset sampling period and sequentially converted into continuous parameter control instructions distributed along the arc length, which are then embedded into the printing motion code. Specifically, this includes: First, the arc length to be printed is divided into multiple variable-speed segments based on the curvature change rate curve. The curvature change rate curve is obtained from the preceding steps, with each arc length position corresponding to a curvature change rate value. A first threshold and a second threshold are preset, where the first threshold is 0.01 radians per millimeter of arc length, and the second threshold is 0.05 radians per millimeter of arc length. The absolute value of the curvature change rate is scanned point-by-point along the arc length direction. If the absolute value is greater than or equal to the second threshold, the arc length segment is marked as a high-speed change segment; if the absolute value is between the first and second thresholds, it is marked as a medium-speed change segment; and if the absolute value is less than or equal to the first threshold, it is marked as a low-speed change segment. A first sampling period of 5 milliseconds is used in the high-speed change segments; a second sampling period of 20 milliseconds is used in both the medium-speed and low-speed change segments. The starting and ending arc length positions of each variable-speed segment are determined by the point where the absolute value of the curvature change rate crosses the threshold boundary.

[0036] Secondly, at each sampling moment, the speed, tension, and extrusion ratio values ​​of the current sampling point are subtracted from the corresponding values ​​of the previous sampling point to obtain the speed increment, tension increment, and extrusion ratio increment. Specifically, the speed increment equals the speed value of the current sampling point minus the speed value of the previous sampling point; the tension increment equals the tension value of the current sampling point minus the tension value of the previous sampling point; and the extrusion ratio increment equals the extrusion ratio value of the current sampling point minus the extrusion ratio value of the previous sampling point. For the first sampling point within each speed-changing section, its preceding sampling point is the adjacent sampling point before the starting point of that section (if it is the starting point of the entire printing path, all values ​​of the preceding sampling point are set to 0). All increment values ​​are retained to three decimal places.

[0037] Finally, the speed increment, tension increment, and extrusion ratio increment are arranged in a fixed order and then sequentially written into the parameter extension field after the current line of motion instruction in the printing motion code. The fixed order is: first write the speed increment, then the tension increment, and finally the extrusion ratio increment. Each increment value is preceded by an instruction prefix: "V" for speed increment, "T" for tension increment, and "E" for extrusion ratio increment. In the printing motion code, each line of motion instruction (e.g., linear interpolation instruction G1) is followed by "V value", "T value", and "E value", separated by spaces. For example, if the speed increment at a sampling point is 0.5 mm / s, the tension increment is 0.2 N / s, and the extrusion ratio increment is 0.01, then the content written is "V0.5 T0.2 E0.01". When the printing controller executes this line of code, it reads each increment value and superimposes it onto the current operating state, thus achieving continuous and smooth changes in speed, tension, and extrusion ratio along the arc length. For increment values ​​of 0, the corresponding instruction is still written, and the value is recorded as 0.

[0038] In the real-time deviation compensation and correction module, during the execution of the printing motion code, the actual fiber tension and resin buildup width are collected in real time and compared with the expected values ​​of the speed curve, tension curve, and extrusion ratio curve one by one. When the deviation exceeds the threshold, the current control command is compensated and corrected by a small amplitude, and the corrected command is used as the control reference for the subsequent arc length position. Specifically, this includes: First, the actual fiber tension and resin buildup width are collected in real time. A tension sensor is installed at the printhead exit, sampling at a frequency of 100 Hz, meaning the actual fiber tension value is read every 10 milliseconds, in Newtons. Simultaneously, a laser displacement sensor is fixed behind the printhead to measure the lateral spread width of the molten resin on the printed layer, also sampling at 100 Hz, with the actual width value read in millimeters. The desired tension and width values ​​are obtained by interpolation based on the current arc length position from the aforementioned optimal speed curve, tension curve, and extrusion ratio curve. The actual fiber tension is subtracted from the desired tension, and the difference is divided by the desired tension to obtain the relative tension deviation. The actual resin buildup width is subtracted from the desired width, and the difference is divided by the desired width to obtain the relative width deviation. In the above division operations, if the desired tension or desired width is zero, the relative deviation is directly set to zero.

[0039] Next, determine the primary deviation and calculate the compensation step size. Compare the absolute values ​​of the tension relative deviation and the width relative deviation. If the absolute value of the tension relative deviation is greater than or equal to the absolute value of the width relative deviation, the tension relative deviation is used as the primary deviation; otherwise, the width relative deviation is used. The preset deviation threshold is 0.05, or five percent. When the absolute value of the primary deviation is less than or equal to this threshold, no compensation correction is performed, and the current instruction continues to be executed. When the absolute value of the primary deviation exceeds this threshold, the compensation direction is determined based on the sign of the primary deviation: if the primary deviation is positive, it indicates that the actual value is greater than the expected value, and the corresponding instruction value needs to be reduced, with a negative compensation direction; if the primary deviation is negative, it indicates that the actual value is less than the expected value, and the corresponding instruction value needs to be increased, with a positive compensation direction. Multiply the absolute value of the primary deviation by the current fixed step size factor to obtain the compensation step size. The initial value of the fixed step size factor is set to 0.1.

[0040] Then, the current printing speed command, fiber tension command, and extrusion ratio command are adjusted synchronously according to the compensation direction and compensation step size. Specifically, the new value of the printing speed command is equal to the current printing speed command value plus the product of the compensation direction and compensation step size, where the compensation direction is either positive or negative 1; the new value of the fiber tension command is equal to the current fiber tension command value plus the product of the same compensation amount and compensation direction; and the new value of the extrusion ratio command is equal to the current extrusion ratio command value plus the product of the same compensation amount and compensation direction. These three adjusted command values ​​are used as a new set of control references for subsequent arc length position printing control. Simultaneously, the main deviation value (with positive and negative signs) calculated in this operation is stored in the deviation history record in memory.

[0041] Finally, the fixed step size factor is dynamically updated for the next compensation correction. A sliding window of fixed length 5 is allocated in memory to store the five most recent principal deviation values. Each time a new principal deviation value is obtained, the earliest stored principal deviation value is removed from the window, and the new value is moved into the window. The sum of the absolute values ​​of all five principal deviation values ​​in the current sliding window is calculated, and this sum is divided by 5 to obtain the average deviation amplitude of the five most recent principal deviations. A baseline step size factor is preset with a value of 0.1; a minimum step size limit of 0.02 and a maximum step size limit of 0.5 are also set. The baseline step size factor is divided by the average deviation amplitude to obtain the preliminary adjustment value. If the preliminary adjustment value is less than the minimum step size limit, the fixed step size factor is set to the minimum step size limit; if the preliminary adjustment value is greater than the maximum step size limit, the fixed step size factor is set to the maximum step size limit; otherwise, the fixed step size factor is set to the preliminary adjustment value. This updated fixed step size factor will be used for the compensation step size calculation when the deviation exceeds the threshold again. If there are fewer than five valid main deviation values ​​within the sliding window (e.g., at the beginning of printing), the fixed step size factor will not be updated and the previous value will still be used.

[0042] The working principle of this invention is as follows: First, discrete sampling is performed along the arc length direction of the path to be printed. Based on the curvature change rate threshold, equal arc length segments and adaptively encrypted segments are divided. The path curvature at each sampling point is calculated using the three-point circle method. A curvature curve is obtained after seven-point quadratic smoothing filtering, and then the curvature change rate curve is obtained through central difference. Second, using the curvature curve and the curvature change rate curve as independent variables, a pre-adjustment direction factor is obtained through numerical integration and arctangent mapping. A pre-adjustment amplitude factor is obtained by power-lawing the ratio of the curvature value to the curvature threshold. The product of the pre-adjustment direction factor and the pre-adjustment amplitude factor is used as a combined weight to perform a weighted summation of the baseline value in the straight section and the set value in the curved section, constructing a continuously differentiable scheduling function of printing speed, fiber tension, and extrusion ratio with respect to arc length. Then, the transition zone is discretized into multiple sub-intervals. With minimizing the fiber strain energy mutation and the resin packing volume deviation as joint optimization objectives, an inner layer step-size halving interval elimination method and an outer layer... The golden ratio method is used for nested optimization, and the optimal speed curve, tension curve, and extrusion ratio curve are obtained offline. Then, the speed-varying segments are divided according to the absolute value of the rate of change of curvature, and different sampling periods are used. The difference between the speed, tension, and extrusion ratio of the current sampling point and the previous sampling point is used as the increment and written into the parameter extension field of the printing motion code in a fixed order to generate continuous parameter control instructions. Finally, during the execution process, the actual fiber tension and resin stacking width are collected in real time, and the relative deviation of tension and width are calculated. The larger absolute value is taken as the main deviation. When the main deviation exceeds the threshold, the compensation direction is determined according to the sign of the main deviation and multiplied by the fixed step size factor to obtain the compensation step size. The printing speed, fiber tension, and extrusion ratio instructions are adjusted synchronously. At the same time, the average deviation amplitude is calculated by saving the sliding window of the last five main deviation values. The baseline step size factor is divided by the average deviation amplitude to limit the amplitude, and the fixed step size factor is dynamically updated to achieve closed-loop compensation correction.

[0043] 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 claims of this invention should still fall within the patent coverage of this invention.

Claims

1. A parameter optimization and control system for continuous fiber additive manufacturing process, characterized in that, include: The path curvature and rate of change extraction module discretizes the sampling along the arc length direction of the path to be printed, calculates the path curvature and its first derivative at each sampling point, and obtains the curvature curve and the rate of change curve of curvature. The continuous differentiable scheduling function construction module uses the curvature curve and the curvature change rate curve as independent variables to construct continuous differentiable scheduling functions for printing speed, fiber tension, and extrusion ratio with respect to arc length. This allows the scheduling function to automatically return to the baseline value in the straight section when the curvature approaches zero, and smoothly transition to the set value in the curved section when the curvature increases. It also realizes pre-deceleration and pre-tension increase before the curve based on the positive or negative curvature change rate. The offline multi-objective constraint optimization module uses the scheduling function as the decision variable and minimizes the fiber strain energy mutation and resin packing volume deviation in the transition zone as the joint optimization objectives. It solves the constraint optimization problem offline within the feasible domain of printing speed, fiber tension and extrusion ratio to obtain the optimal speed curve, tension curve and extrusion ratio curve. The discretization instruction generation and embedding module discretizes the speed curve, tension curve, and extrusion ratio curve according to a preset sampling period, and converts them into continuous parameter control instructions distributed along the arc length, which are then embedded into the printing motion code. The real-time deviation compensation and correction module collects the actual fiber tension and resin buildup width in real time during the execution of the printing motion code. It compares these values ​​with the expected values ​​of the speed curve, tension curve, and extrusion ratio curve one by one. When the deviation exceeds the threshold, it performs a small-amplitude compensation and correction on the current control command and uses the corrected command as the control reference for the subsequent arc length position.

2. The continuous fiber additive manufacturing process parameter optimization control system according to claim 1, characterized in that, The calculation of the path curvature and its first derivative at each sampling point, resulting in the curvature curve and the rate of change of curvature curve, specifically includes: The path to be printed is divided into equal arc length segments and adaptive encryption segments based on the curvature change rate threshold. Fixed large step size sampling is used in equal arc length segments, while variable step size sampling is used in adaptive encryption segments, which decreases as the curvature change rate increases. For each sampling point, take the two adjacent sampling points before and after it to form a three-point circle, and calculate the reciprocal of the radius of the three-point circle as the path curvature of the corresponding sampling point; The obtained curvature sequence is subjected to a seven-point quadratic smoothing filter to obtain the curvature curve, and the curvature curve is subjected to central difference to obtain the curvature change rate curve.

3. The continuous fiber additive manufacturing process parameter optimization control system according to claim 1, characterized in that, The construction process of the continuously differentiable schedulable function is as follows: The total curvature change is obtained by numerically integrating the curvature change rate curve. The total curvature change is then mapped to the negative one and positive one intervals via the arctangent function to obtain the pre-adjustment direction factor. The pre-adjustment amplitude factor is obtained by power-calculating the ratio of the curvature value at each point on the curvature curve to the preset curvature threshold. The printing speed is set as a weighted sum of the baseline value in the straight section and the set value in the curved section, where the weight is the product of the pre-adjustment direction factor and the pre-adjustment amplitude factor. The fiber tension and extrusion ratio are weighted and summed according to the same weighting rule, so that when the curvature change rate is positive, pre-deceleration and pre-tension increase are achieved, and when the curvature change rate is negative, slow acceleration and slow tension decrease are achieved.

4. The parameter optimization and control system for continuous fiber additive manufacturing process according to claim 1, characterized in that, Obtaining the optimal speed curve, tension curve, and extrusion ratio curve specifically includes: The transition zone is discretized into multiple sub-intervals. Within each sub-interval, the values ​​of the scheduling function at both ends are used as design variables to calculate the fiber strain energy mutation value and the resin packing volume deviation value respectively. The inner and outer nested optimization method is adopted. The inner layer uses the interval elimination method with halved step size to successively narrow the feasible domain of printing speed, while the outer layer uses the golden section method to simultaneously search for the optimal combination of fiber tension and extrusion ratio. Substitute the printing speed, fiber tension, and extrusion ratio obtained from each optimization into the joint optimization objective. When the relative change of the objective function value between two consecutive iterations is less than the preset convergence threshold, stop the iteration and output the current speed curve, tension curve, and extrusion ratio curve.

5. The continuous fiber additive manufacturing process parameter optimization control system according to claim 4, characterized in that, The outer layer simultaneously searches for the optimal combination of fiber tension and extrusion ratio using the golden ratio method, specifically including: The initial search ranges for fiber tension and extrusion ratio are set separately, and two golden section points are taken in each of the two ranges to generate four two-dimensional candidate points. Calculate the joint optimization objective value for each candidate point, eliminate the sub-interval containing the candidate point with the largest objective value, and form a new two-dimensional search interval by combining the regions containing the remaining three candidate points. Repeat steps one and two until the length of the fiber tension interval and the length of the extrusion ratio interval are both less than their respective preset accuracy thresholds. Take the midpoint of the final interval as the optimal fiber tension and the optimal extrusion ratio.

6. The parameter optimization control system for continuous fiber additive manufacturing process according to claim 1, characterized in that, The sequential conversion into continuous parameter control commands distributed along the arc length specifically includes: The arc length to be printed is divided into multiple variable speed sections based on the curvature change rate curve. The first sampling period is used in the section with a large absolute value of curvature change rate, and the second sampling period is used in the section with a small absolute value of curvature change rate. The first sampling period is shorter than the second sampling period. At each sampling moment, the speed value, tension value, and extrusion ratio value of the current sampling point are subtracted from the corresponding values ​​of the previous sampling point to obtain the speed increment, tension increment, and extrusion ratio increment. After arranging the speed increment, tension increment, and extrusion ratio increment in a fixed order, they are sequentially written into the parameter extension field following the current line of motion instructions in the printed motion code.

7. The continuous fiber additive manufacturing process parameter optimization control system according to claim 1, characterized in that, The compensation and correction process is as follows: The difference between the real-time collected fiber tension and the expected tension is divided by the expected tension to obtain the relative tension deviation, and the difference between the actual value of the resin packing width and the expected width is divided by the expected width to obtain the relative width deviation. The absolute values ​​of the tension relative deviation and the width relative deviation are compared respectively. If the former is greater than or equal to the latter, the tension relative deviation is taken as the main deviation; otherwise, the width relative deviation is taken as the main deviation. When the absolute value of the main deviation exceeds the preset threshold, the compensation direction is determined according to the positive or negative sign of the main deviation, and the absolute value of the main deviation is multiplied by a fixed step size factor to obtain the compensation step size. The current printing speed command, fiber tension command, and extrusion ratio command are adjusted synchronously according to the compensation direction and compensation step size. The adjusted command set is used as the control reference for the subsequent arc length position. At the same time, the main deviation value is accumulated to the deviation history record to adjust the fixed step size factor for the next step.

8. The continuous fiber additive manufacturing process parameter optimization control system according to claim 7, characterized in that, The calculation process for the fixed step size factor is as follows: The five most recent major deviation values ​​are stored in memory to form a deviation sliding window. Each time a new major deviation value is obtained, the earliest deviation value in the window is moved out and the new value is moved in. Calculate the sum of the absolute values ​​of all major deviations within the sliding window, and then divide by five to obtain the average deviation amplitude; The initial adjustment value is obtained by dividing the preset baseline step size factor by the average deviation amplitude. If the initial adjustment value is less than the minimum step size limit, the minimum step size limit is taken. If it is greater than the maximum step size limit, the maximum step size limit is taken. The final value is used as the fixed step size factor for the next compensation correction.

Images