Wax mold extrusion control method and system based on direct writing forming technology

Abstract

The invention discloses a wax mold extrusion control method and system based on a direct writing forming technology. The wax mold extrusion control method comprises the following steps that a pressure sensor collects an extrusion pressure signal in real time; a tracking error is calculated according to the collected extrusion pressure signal and a set target pressure value, and BP-PID fusion control network parameters are adjusted, so that wax mold extrusion is tracked and controlled through a BP-PID fusion control algorithm; when it is detected that the wax pattern printing task is completed or a termination instruction is received, a safe closing process is automatically executed, and a standby state is entered; a back propagation neural network and an incremental PID are adopted to form a composite controller, self-adaptive optimization of control parameters is achieved, and therefore a pressure closed-loop control system is constructed, and the extrusion pressure is stabilized at a set value in combination with lead screw transmission. According to the technical scheme, the technical problems that a traditional extrusion system is large in pressure fluctuation, lagged in response and the like can be effectively solved, and the method is suitable for rapid preparation of the precision casting wax mold.

Description

Technical Field

[0001] The present invention relates to the technical field of wax pattern rapid prototyping for investment casting, and in particular to a wax pattern extrusion control method and system based on direct writing molding technology. Background Art

[0002] Investment casting is a high-precision casting process developed on the basis of traditional investment casting. Its core feature is the ability to produce complex metal components with excellent dimensional accuracy and surface quality. The process first prepares a precision wax mold, and then prepares the shell through key processes such as multi-layer slurry coating, refractory sanding, shell drying and hardening, and high-temperature roasting. Finally, liquid metal is poured to obtain a formed casting. However, this process has significant limitations. The casting manufacturing cycle is long, and the development cost of precision molds is high. In response to the above technical limitations, researchers have developed a new rapid investment casting process. This technology combines rapid prototyping technology with traditional investment casting technology. Rapid prototyping technology is used to directly prepare wax molds for casting. It can be produced flexibly in small batches, significantly shortening the production cycle and reducing production costs.

[0003] Direct Ink Writing (DIW) is a pressure-based additive manufacturing technology that creates three-dimensional structures through the precisely controlled extrusion deposition of functional materials. Its core principle is to deposit rheologically optimized materials such as gels, slurries, or composites layer by layer in the form of continuous fibers, ultimately forming complex geometries. Its main advantages include simple equipment configuration, low overall cost, wide material compatibility, and high printing accuracy.

[0004] The wax direct writing molding process based on rapid prototyping technology has significant advantages such as low cost, short cycle, environmental friendliness and recyclable materials. At present, the research on the precision optimization of this process mainly focuses on two aspects: one is to optimize the molding quality by systematically controlling the printing parameters (such as nozzle diameter, extrusion pressure, printing speed, etc.); the other is a dynamic control strategy based on real-time monitoring, which realizes closed-loop control of printing accuracy by online detection of feedback signals such as extrusion force, material rheological properties and structural deformation. For example, the gas pressure control system in the gas pressure control system, 3D printing device and material conveying system (CN202322393644.9) provides gas with target pressure through a load module connected to the piston and the movement of the piston in the cylinder. The disadvantage of this gas pressure control system is that it relies on mechanical load to achieve pressure control, resulting in low pressure regulation accuracy, slow dynamic response and inability to adapt to disturbances in real time. It lacks closed-loop feedback and intelligent optimization capabilities. It is only applicable to simple scenarios with low precision requirements and is difficult to meet the needs of high-precision manufacturing or complex working conditions. Summary of the Invention

[0005] In order to overcome the shortcomings and deficiencies of the existing technology, the present invention proposes a wax mold extrusion control method and system based on direct writing molding technology. The method is a wax extrusion force stabilization control method based on back propagation (BP) neural network PID, which is an intersection technology of investment casting, additive manufacturing and adaptive control. By constructing a wax direct writing molding device and an extrusion force closed-loop control system, the extrusion force stability and adaptive adjustment ability of the extrusion process are improved to solve the shortcomings of the existing equipment in extrusion stability and achieve improved precision and size in the printing process.

[0006] The present invention is achieved through at least one of the following technical solutions.

[0007] A wax mold extrusion control method based on direct writing molding technology comprises the following steps:

[0008] 1) Extrusion pressure signal collected in real time by pressure sensor;

[0009] 2) Calculate the tracking error based on the collected extrusion pressure signal and the set target pressure value, and adjust the BP-PID fusion control network parameters to implement BP-PID fusion control algorithm to track and control the wax mold extrusion;

[0010] 3) When the wax model printing task is detected to be completed or a termination instruction is received, the system will automatically execute the safety shutdown process and enter the standby state.

[0011] Furthermore, the BP-PID fusion control algorithm includes a BP neural network and an incremental PID controller. The BP neural network adopts a three-layer feedforward structure, and the network architecture consists of an input layer, a hidden layer, and an output layer.

[0012] Furthermore, tracking control of wax pattern extrusion is performed based on the BP-PID fusion control algorithm, including the following steps:

[0013] (1) Calculate the error value of the current control cycle k based on the data collected by the pressure sensor: e(k) = target pressure value - actual pressure measurement value;

[0014] (2) Based on the current extrusion pressure signal and network weight parameters, the BP neural network derives the optimal parameter combination of the PID controller in real time through the nonlinear mapping relationship between the hidden layer and the output layer [K p , K i , K d ], where K p , K i , K d Represent the proportional coefficient, integral coefficient, and differential coefficient respectively;

[0015] (3) The incremental PID controller calculates the control increment Δu(k) according to the incremental PID formula:

[0016] Δu(k)=K p ·[e(k)-e(k-1)]+K i ·e(k)+K d [e(k)-2e(k-1)+e(k-2)]

[0017] Update control output:

[0018] u(k)=u(k-1)+Δu(k)

[0019] Where e(k) represents the tracking error of the previous control cycle k; Δe(k) represents the error change in the current control cycle k, that is, the difference between the tracking error of the current cycle and the tracking error of the previous cycle; k represents the control cycle; e(k-1) and e(k-2) are the historical error values ​​of the previous two control cycles;

[0020] (4) The control quantity u(k) is converted into a control signal through linear mapping and written into the timer CCR register of the main control chip to drive the stepper motor to perform pressure regulation;

[0021] (5) The gradient descent algorithm is combined with the error back propagation mechanism to dynamically adjust the BP neural network weight matrix. Specifically, by calculating the partial derivative of the control tracking error with respect to the BP neural network weight, the weight parameters are iteratively updated along the negative gradient direction, thereby achieving the PID control parameter [K p , K i , K d ]’s online adaptive optimization;

[0022] (6) After completing all operations in the current control cycle k, the time stamp is updated to k=k+1, and the tracking error is recalculated to start a new round of control cycle.

[0023] A system for implementing the wax mold extrusion control method based on direct writing molding technology includes an extrusion nozzle device, a main control chip, a motor driver, and a signal amplifier;

[0024] The PWM signal input end of the motor driver is connected to the output end of the main control chip, and the main control chip processes the pressure sensor data in real time through the BP-PID fusion control algorithm and generates instructions to control the motor driver; the output end of the motor driver is connected to the stepper motor of the extrusion nozzle device; the input end of the signal amplifier is connected to the output end of the pressure sensor of the extrusion nozzle device, and the output end of the signal amplifier is connected to the ADC sampling channel of the main control chip.

[0025] Furthermore, the extrusion nozzle device includes a stepping motor, a moving screw, a pressure sensor fixing part, a pressure sensor, a movable plunger, a hot melt storage tank, and an extrusion nozzle;

[0026] The output shaft of the stepper motor is connected to the upper end of the moving screw through a coupling, driving the moving screw to perform axial feed movement; the pressure sensor fixing piece is connected to the end of the moving screw through a threaded connection to achieve axial positioning; the pressure sensor is fixed on the pressure sensor fixing piece; the upper end face of the movable plunger is in dynamic contact with the force measuring end of the pressure sensor, and axial movement is achieved by the drive of the moving screw, the lower end of the movable plunger extends to the inner cavity of the hot melt storage tank, and the lower end face forms an extrusion working surface for the molten wax material; the bottom of the hot melt storage tank is threadedly connected to the extrusion nozzle, and the extrusion nozzle forms a flow channel with the bottom discharge port of the hot melt storage tank through the thread.

[0027] Furthermore, the stepper motor is fixed to the top of the aluminum alloy connecting frame, and the hot melt storage tank is fixed to the bottom of the aluminum alloy connecting frame.

[0028] Furthermore, the inner cavity of the hot melt storage tank is a tubular inner cavity, which facilitates the movable plunger to achieve stable axial movement through the tubular inner cavity of the hot melt storage tank.

[0029] Furthermore, the upper end of the movable plunger and the force measuring end of the pressure sensor are coupled with a ball head to avoid lateral force interference caused by installation deviation.

[0030] Furthermore, it also includes a host computer, a printer control module, a printer spatial motion mechanism, and a printing substrate; the host computer is connected to the main control chip, and the printer control module is connected to the host computer; the printer control module is used to adjust the temperature of the printing substrate and the XYZ three-axis positioning of the spatial motion mechanism, and the hot melt storage tank of the extrusion nozzle device is fixed on the crossbeam pulley to perform horizontal reciprocating motion in the Y-axis direction; during the printing process, the control module drives the printer spatial motion mechanism to move in coordination with the extrusion nozzle device according to a preset trajectory, and accurately manufactures three-dimensional components through a layer-by-layer stacking molding process.

[0031] Furthermore, the hot melt storage tank has a built-in heating module that supports temperature setting and real-time display functions to achieve precise melting of the wax material.

[0032] Compared with the prior art, the present invention has the following advantages and beneficial effects:

[0033] (1) The device uses direct writing technology to achieve precise extrusion molding of molten wax. The nozzle system has been optimized as a whole to ensure molding accuracy and stability. At the same time, the embedded software development system based on STM32 combined with the pressure closed-loop control algorithm realizes high-precision dynamic adjustment of extrusion pressure, enhancing the stability and controllability of the molding process.

[0034] (2) The extrusion pressure control process breaks through the limitations of traditional PID control and integrates the BP neural network algorithm to achieve adaptive adjustment of control parameters, thereby improving the stability and response speed of pressure control. This method effectively solves the control problem of nonlinear and time-varying characteristics in the extrusion process. BRIEF DESCRIPTION OF THE DRAWINGS

[0035] Figure 1 This is a schematic diagram of the operation principle of a wax mold extrusion system based on direct writing molding technology provided by the present invention;

[0036] Figure 2 This is a schematic structural diagram of a print head device of a wax mold direct writing extrusion molding system provided by the present invention;

[0037] Figure 3 This is a principle block diagram of the BP neural network PID controller provided by the present invention;

[0038] Figure 4 This is a schematic diagram of the BP neural network topology provided by the present invention;

[0039] Figure 5 This is a diagram of a wax mold extrusion system control method based on direct writing molding technology provided by the present invention;

[0040] Figure 6 This is a Simulink simulation architecture diagram of the BP neural network PID controller provided by the present invention and the traditional PID controller;

[0041] Figure 7 This is a dynamic performance comparison curve diagram of the BP neural network PID controller provided by the present invention and the traditional PID controller in the Simulink environment;

[0042] Description of reference numerals:

[0043] 1- Host computer, 2- Printer control module, 3- Printer spatial motion mechanism, 4- Printing substrate, 5- Extrusion nozzle device, 501- Stepper motor, 502- Motion screw, 503- Pressure sensor, 504- Movable plunger, 505- Hot melt storage tank, 506- M4 fixing screw, 507- M3 fixing screw, 508- Pressure sensor fixing part, 509- Connecting frame, 510- M5 fixing screw, 511- Extrusion nozzle, 6- Main control chip, 7- Motor driver, 8- Signal amplifier. DETAILED DESCRIPTION

[0044] In order to make the purpose, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below in combination with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments.

[0045] like Figure 1 As shown, a wax mold extrusion system based on direct writing molding technology includes an extrusion nozzle device 5, an STM32F407 main control chip 6, a motor driver 7, and a signal amplifier 8.

[0046] The main control chip 6 processes the pressure sensor data in real time and generates control instructions; the PWM signal input end of the motor driver 7 is connected to the output end of the main control chip 6, and the output end of the motor driver 7 is connected to the stepper motor 501 of the extrusion nozzle device 5. The motor driver 7 outputs a PWM speed regulation signal to drive the stepper motor 501 of the extrusion nozzle device 5; the signal amplifier 8 performs gain adjustment on the pressure signal, and the output end of the signal amplifier 8 is connected to the main control chip 6 through the filtering circuit to form a complete closed-loop control system.

[0047] The wax pattern extrusion system of this embodiment also includes other external devices, including a host computer 1, a printer control module 2, a printer spatial motion mechanism 3, and a printing substrate 4. The host computer 1 is connected to a main control chip 6, and the printer control module 2 is connected to the host computer 1. The printing substrate 4 and the printer spatial motion mechanism 3 are connected to the printer control module 2. The printer control module 2 precisely controls the temperature of the printing substrate 4 and the three-dimensional positioning of the spatial motion mechanism 3. The host computer 1 performs program burning and pressure signal transmission.

[0048] The printer control module 2, printing substrate 4, and spatial motion mechanism 3 are derived from the existing Ender-3 printing platform. Before printing, the 3D model is layered using Creality Slicer professional slicing software, generating a G-code instruction set containing parameters such as printing speed, substrate temperature, and path trajectory. The printer control module 2 then analyzes the G-code instructions in real time to precisely control the heating temperature of the printing substrate 4 and the X, Y, and Z axis positioning of the spatial motion mechanism 3. The hot melt storage tank 505 in the extrusion nozzle assembly 5 is fixed to a crossbeam pulley, performing lateral reciprocating motion in the Y-axis direction, replacing the existing Ender-3 nozzle. During printing, the control module 2 coordinates with the spatial motion mechanism 3 in the Y-axis direction, thereby driving the hot melt storage tank 505 to move synchronously along a preset trajectory, completing the production of the 3D component through a layer-by-layer stacking process.

[0049] Figure 2 This is a schematic diagram of the structure of the print head device, which includes, from top to bottom, a stepper motor 501, a moving screw 502, a pressure sensor fixing part 508, a pressure sensor 503, a movable plunger 504, a hot melt storage tank 505, an extrusion nozzle 511, a connecting frame 509 for the hot melt storage tank, and three fixing screws (506, 507, 510).

[0050] The output shaft of the stepper motor 501 is connected to the upper end of the moving screw 502 through a coupling, driving the moving screw 502 to perform axial feed motion. The pressure sensor fixing member 508 is threadedly connected to the end of the moving screw 502 to achieve axial positioning; the pressure sensor 503 is rigidly fixed to the pressure sensor fixing member 508 by three M3 fixing screws 507;

[0051] The lower end of the movable plunger 504 extends into the interior of the hot melt storage tank 505. The side surface of the lower end of the movable plunger 504 cooperates with the inner wall of the hot melt storage tank 505 to ensure that it moves only in the axial direction and avoid radial deviation. The lower end surface of the movable plunger 504 directly contacts the molten wax, forming an extrusion working surface for the molten wax. The axial displacement changes the volume of the storage tank cavity, achieving quantitative extrusion of the wax. The upper end of the movable plunger 504 maintains dynamic contact with the force-measuring end of the pressure sensor 503, allowing the movable plunger 504 to be freely fine-tuned during pressure testing while simultaneously transmitting the extrusion force to the sensor 503.

[0052] As an embodiment, the upper end of the movable plunger 504 and the force measuring end of the pressure sensor 503 are coupled with a ball head to avoid lateral force interference caused by installation deviation.

[0053] The moving screw 502 transmits thrust through the pressure sensor fixing part 508. When the screw is rotated and pressed down, the fixing part 508 at the end of the screw and the pressure sensor 503 transmit the force to the upper end of the movable plunger 504. The axial feed amount of the moving screw 502 is controlled by the stepper motor 501, which directly determines the displacement stroke of the plunger, thereby accurately controlling the extrusion pressure. The movable plunger 504 achieves stable axial movement through the tubular inner cavity of the hot melt storage tank 505, and its driving force is directly transmitted by the pressure sensor 503 fixed at the end of the moving screw 502, forming a force chain of "stepper motor → screw → sensor → plunger", ensuring that the extrusion force is measurable and controllable.

[0054] The hot melt storage tank 505 integrates the heating module and temperature controller of the NBW-030F hot melt adhesive dispensing machine. The heating module realizes constant temperature melting of the wax material, supports temperature setting and real-time display functions, and can achieve precise melting of the wax material.

[0055] The bottom of the hot melt storage tank 505 is threadedly connected to the extrusion nozzle 511, and the extrusion nozzle 511 forms a flow channel with the bottom discharge port of the hot melt storage tank 505 through the high-temperature sealing thread; the axial pressure generated by the movable plunger 504 driven by the screw 502 directly acts on the molten wax, forcing the molten wax to pass through the above-mentioned high-pressure flow channel; the conical flow channel design of the extrusion nozzle 511 can realize pressure gradient conversion, converting the mechanical pressure of the plunger into the kinetic energy of the wax flow, thereby realizing wax extrusion.

[0056] The stepper motor 501 is arranged in a rectangular array on the top reference plane of the aluminum alloy connecting frame 509 via four M4 fixing screws 506, and the screws are tightened to ensure the coaxial positioning of the two. The hot melt storage tank 505 is symmetrically locked to the mounting holes on both sides of the aluminum alloy connecting frame 509 via two M5 fixing screws 510.

[0057] The PWM signal input end of the motor driver 7 is connected to the timer output pin of the main control chip 6 through a shielded twisted pair cable, and its power output end adopts a common cathode structure to form a closed loop with the two-phase winding of the stepper motor 501; the differential input end of the signal amplifier 8 is connected to the output end of the pressure sensor 503, and its amplified signal output end is connected to the ADC sampling channel of the main control chip 6 through a filtering circuit; the host computer 1 is connected to the pin of the main control chip 6 through ST-Link to realize program burning and debugging, and at the same time establishes a real-time transmission channel for pressure data through the serial port.

[0058] The process preparation process before printing of a wax model extrusion system based on direct writing technology is as follows:

[0059] Model preprocessing stage: Build a parametric model of the part based on the 3D design platform, perform layered processing using Creality Slicer professional slicing software, and generate a G-code instruction set containing core parameters such as printing speed;

[0060] Data interactive transmission stage: G code instruction set data is transmitted to the printer control module 2;

[0061] Material preparation stage: solid wax material is loaded into the hot melt storage tank 505, and the heating temperature is set by adjusting the heating module in the hot melt storage tank 505 so that the material reaches an optimal rheological state.

[0062] The system runs as follows:

[0063] The target extrusion pressure value is set on the control interface of the host computer 1, and then the executable file generated by the host computer 1 based on the MDK-ARM development environment is compiled and burned into the main control chip 6;

[0064] The extrusion pressure signal is collected in real time by the pressure sensor 503 and transmitted to the main control chip 6 via the ADC module. The BP-PID fusion control algorithm compiled by MDK is run on the main control chip 6. The closed-loop motor driver 7 precisely controls the speed and direction of the stepper motor 501, thereby driving the moving screw 502 to drive the movable plunger 504 for axial displacement. This control system constructs a complete closed-loop regulation circuit of pressure sensing-algorithm processing-motor drive-mechanical execution, and achieves pressure control accuracy through neural network self-tuning of PID parameters. When the system pressure reaches the critical flow resistance threshold of the extrusion nozzle 511 calibrated by rheological properties, it enters the steady-state extrusion stage;

[0065] Synchronously, the printer control module 2 drives the printer spatial motion mechanism 3 to move in coordination with the extrusion nozzle device 5 according to a preset trajectory, and accurately manufactures three-dimensional components through a layer-by-layer stacking molding process.

[0066] The BP-PID fusion control algorithm includes BP neural network and incremental PID controller. Figure 4 As shown in the figure, the BP neural network adopts a three-layer feedforward structure design. Its network architecture consists of an input layer, a hidden layer, and an output layer. The number of neuron nodes in each layer is determined to be a 3-5-3 configuration after system optimization.

[0067] like Figures 3 to 5 As shown, the control method of the wax mold extrusion control system based on the direct writing molding technology includes the following steps:

[0068] P01, initialization, set the BP neural network structure parameters (number of input layer nodes N, number of hidden layer nodes M), randomly initialize the weight matrix W connecting the input layer and the hidden layer ji (2) and the weight matrix W connecting the hidden layer and the output layer il (3) , configure the learning rate η and momentum factor α;

[0069] P02, Data Acquisition and Error Calculation: The system collects the pressure setpoint r(k) and the actual output measurement value y(k) at the current moment k in real time. By calculating the difference between the two, the tracking error e(k) = r(k) - y(k) is obtained. The error signal e(k) will be used in subsequent neural network training and parameter adjustment processes to achieve precise tracking and control of the system output.

[0070] P03, forward calculation, BP neural network is based on the current input signal and network weight parameters, through the nonlinear mapping relationship between the hidden layer and the output layer, to derive the optimal parameter combination of PID controller in real time [K p , K i , K d ];

[0071] P04, control quantity calculation, the system calculates the control output u(k) of the previous control cycle k in real time based on the PID control algorithm. The algorithm combines the three components of current error, error integral and error differential, and calculates the control output u(k) through the proportional coefficient K p , integral coefficient K i and differential coefficient K d The weighted combination of generates the precise control quantity u(k) as follows:

[0072] u(k)=K p ·e(k)+K i ∑e(k)+Kd Δe(k)

[0073] Where e(k) represents the tracking error of the previous control cycle k, Δe(k) represents the error change in the current control cycle k, and k represents the control cycle.

[0074] P05, weight optimization, the system uses the gradient descent algorithm combined with the error back propagation mechanism to dynamically adjust the neural network weight matrix W ji (2) and W il (3) This process calculates the partial derivative of the control error with respect to the network weight and iteratively updates the weight parameter along the negative gradient direction, thereby achieving the PID control parameter [K p , K i , K d ]’s online adaptive optimization;

[0075] P06, Timing Update: After the system completes all operations in the current control cycle k, it updates the time marker to k=k+1 and jumps back to step P02 data acquisition and error calculation to start a new round of control cycle;

[0076] P07, system ends. When it detects that the wax model printing task is completed or receives a termination command, the system will automatically execute the safety shutdown process and enter the standby state.

[0077] like Figure 6 As shown in FIG, this embodiment provides a Simulink simulation architecture diagram of a BP neural network PID controller and a traditional PID controller. The subsystem serves as the core processing unit of the entire control system and is responsible for signal transmission, error calculation, and interaction with the controller.

[0078] Subsystem input signals include:

[0079] rin - the target value set;

[0080] e-real-time control error;

[0081] yout-system output (actual output of the controlled object);

[0082] u-controller output (controlled quantity).

[0083] Subsystem output signals include:

[0084] rin(k) - reference input at the current moment;

[0085] yout(k-1)-the system output at the previous moment;

[0086] yout(k)-system output at the current moment;

[0087] e(k-2)-error between the first two moments;

[0088] e(k-1)-error at the previous moment;

[0089] e(k)-error at the current moment;

[0090] u(k-1)-the control quantity at the previous moment;

[0091] u(k)-the control quantity at the current moment.

[0092] The S function is the implementation module of the control algorithm, which is used to dynamically adjust the PID parameters (K p , K i , K d ), and generates control signals, s_bppid is the source code file name of the S function.

[0093] Transfer function 1 / transfer function 2 (Transfer Fcn1 / Transfer Fcn2) represents a transfer function model of the controlled object and simulates the dynamic characteristics of the actual controlled object.

[0094] out.Kp / out.Ki / out.Kd represent the PID parameters adjusted in real time: proportional gain, integral gain, and differential gain respectively.

[0095] Figure 6 The simulation architecture sets up a dual-channel controller architecture, uses the S function to implement the neural network forward calculation, and establishes the nozzle extrusion system transfer function in advance. As an embodiment, during the simulation process, the initial parameters of the traditional PID controller are first determined by trial and error, which are the proportional coefficient K p =10, integral coefficient K i =1 and differential coefficient K d = 3. For the BP-PID controller, the simulation parameters were set as follows: learning rate α = 0.2, inertia coefficient η = 0.5, and weight coefficients randomly initialized within the interval [-1, 1]. To verify the performance of the control algorithm, the simulation experiment used a step signal with a pressure change of 1 after a 10-second delay as the input signal, testing both the traditional PID controller and the BP-PID controller.

[0096] like Figure 7As shown, this embodiment provides a dynamic performance comparison curve diagram of the BP neural network PID controller and the traditional PID controller in the Simulink environment. According to the image information, the BP neural network PID control is significantly better than the traditional PID control in terms of dynamic performance. When the set pressure is 1, the BP-PID can quickly converge to the set pressure. Compared with the PID control curve, it takes less time to enter the steady state, the response curve is smoother, and there is no overshoot and fluctuation. Theory shows that the BP-PID control system can more accurately stabilize the extrusion force during the wax extrusion process, and has more superior dynamic performance. This result further verifies the effectiveness of the control method of the present invention in complex time-varying and nonlinear systems.

[0097] The above embodiments are preferred implementation modes of the present invention, but the implementation modes of the present invention are not limited to the embodiments. Any other changes, modifications, substitutions, combinations, and simplifications that do not deviate from the spirit and principles of the present invention should be considered as equivalent replacement methods and are included in the scope of protection of the present invention.

Claims

1. A wax mold extrusion control method based on direct writing technology, characterized in that: The following steps are involved: 1) The pressure sensor collects the extrusion pressure signal in real time; 2) Calculate the tracking error based on the collected extrusion pressure signal and the set target pressure value, and adjust the BP-PID fusion control network parameters to implement BP-PID fusion control algorithm to track and control the wax mold extrusion; 3) When the wax model printing task is detected to be completed or a termination instruction is received, the system will automatically execute the safety shutdown process and enter the standby state.

2. A wax mold extrusion control system based on direct writing molding technology according to claim 1, characterized in that: The BP-PID fusion control algorithm includes BP neural network and incremental PID controller. The BP neural network adopts a three-layer feedforward structure, and the network architecture consists of an input layer, a hidden layer, and an output layer.

3. The wax mold extrusion control method based on direct writing technology according to claim 1, characterized in that: The tracking control of wax pattern extrusion based on BP-PID fusion control algorithm includes the following steps: (1) Calculate the error value of the current control cycle k based on the data collected by the pressure sensor: e(k) = target pressure value - actual pressure measurement value; (2) Based on the current extrusion pressure signal and network weight parameters, the BP neural network derives the optimal parameter combination of the PID controller in real time through the nonlinear mapping relationship between the hidden layer and the output layer [K p , K i , K d ], where K p , K i , K d Represent the proportional coefficient, integral coefficient, and differential coefficient respectively; (3) The incremental PID controller calculates the control increment Δu(k) according to the incremental PID formula: Δu(k)=K p [e(k)-e(k-1)]+K i ·e(k)+K d [e(k)-2e(k-1)+e(k-2)] Update control output: u(k)=u(k-1)+Δu(k) Where e(k) represents the tracking error of the previous control cycle k; Δe(k) represents the error change in the current control cycle k, that is, the difference between the tracking error of the current cycle and the tracking error of the previous cycle; k represents the control cycle; e(k-1) and e(k-2) are the historical error values ​​of the previous two control cycles; (4) The control quantity u(k) is converted into a control signal through linear mapping and written into the timer CCR register of the main control chip to drive the stepper motor to perform pressure regulation; (5) The gradient descent algorithm is combined with the error back propagation mechanism to dynamically adjust the BP neural network weight matrix. Specifically, by calculating the partial derivative of the control tracking error with respect to the BP neural network weight, the weight parameters are iteratively updated along the negative gradient direction, thereby achieving the PID control parameter [K p , K i , K d ]’s online adaptive optimization; (6) After completing all operations in the current control cycle k, the time stamp is updated to k=k+1, and the tracking error is recalculated to start a new round of control cycle.

4. A system for implementing the wax mold extrusion control method based on direct writing technology as described in claim 1, characterized in that: It includes an extrusion nozzle device (5), a main control chip (6), a motor driver (7), and a signal amplifier (8); The PWM signal input end of the motor driver (7) is connected to the output end of the main control chip (6), and the main control chip (6) processes the pressure sensor data in real time through the BP-PID fusion control algorithm and generates instructions for controlling the motor driver (7); the output end of the motor driver (7) is connected to the stepper motor (501) of the extrusion nozzle device (5); the input end of the signal amplifier (8) is connected to the output end of the pressure sensor (503) of the extrusion nozzle device (5), and the output end of the signal amplifier (8) is connected to the ADC sampling channel of the main control chip (6).

5. The wax mold extrusion control system based on direct writing technology according to claim 4 is characterized in that: The extrusion nozzle device (5) includes a stepping motor (501), a moving screw (502), a pressure sensor fixing member (508), a pressure sensor (503), a movable plunger (504), a hot melt storage tank (505), and an extrusion nozzle (511); The output shaft of the stepper motor (501) is connected to the upper end of the moving screw (502) through a coupling, driving the moving screw (502) to perform axial feeding movement; the pressure sensor fixing member (508) is connected to the end of the moving screw (502) through a threaded connection to achieve axial positioning; the pressure sensor (503) is fixed on the pressure sensor fixing member (508); the upper end face of the movable plunger (504) is in dynamic contact with the force measuring end of the pressure sensor (503), and axial movement is achieved by the drive of the moving screw (502); the lower end of the movable plunger (504) extends to the inner cavity of the hot melt storage tank (505), and the lower end face forms an extrusion working surface for the molten wax material; the bottom of the hot melt storage tank (505) is threadedly connected to the extrusion nozzle (511), and the extrusion nozzle (511) forms a flow channel with the bottom discharge port of the hot melt storage tank (505) through the thread.

6. The wax mold extrusion control system based on direct writing technology according to claim 5, characterized in that: The stepper motor (501) is fixed to the top of the aluminum alloy connecting frame (509), and the hot melt storage tank (505) is fixed to the bottom of the aluminum alloy connecting frame (509).

7. The wax mold extrusion control system based on direct writing technology according to claim 5, characterized in that: The inner cavity of the hot melt storage tank (505) is a tubular inner cavity, which facilitates the movable plunger (504) to achieve stable axial movement through the tubular inner cavity of the hot melt storage tank (505).

8. The wax mold extrusion control system based on direct writing technology according to claim 5, characterized in that: The upper end of the movable plunger (504) and the force measuring end of the pressure sensor (503) are coupled with a ball head to avoid lateral force interference caused by installation deviation.

9. The wax mold extrusion control system based on direct writing technology according to claim 4, characterized in that: The invention also includes a host computer (1), a printer control module (2), a printer spatial motion mechanism (3), and a printing substrate (4); the host computer (1) is connected to a main control chip (6), and the printer control module (2) is connected to the host computer (1); the printer control module (2) is used to control the temperature of the printing substrate (4) and the XYZ three-axis positioning of the spatial motion mechanism (3); the hot melt storage tank (505) of the extrusion nozzle device (5) is fixed on the crossbeam pulley to perform lateral reciprocating motion in the Y-axis direction; during the printing process, the control module (2) drives the printer spatial motion mechanism (3) to move in coordination with the extrusion nozzle device (5) according to a preset trajectory, and accurately manufactures a three-dimensional component through a layer-by-layer stacking molding process.

10. A wax mold extrusion control system based on direct writing technology according to any one of claims 4 to 9, characterized in that: The hot melt storage tank (505) has a built-in heating module, which supports temperature setting and real-time display functions to achieve accurate melting of the wax material.

Images