Modified abs injection molding process parameter adaptive control system

Abstract

The present application relates to injection molding process control technical field, especially to modified ABS injection molding process parameter adaptive control system, through the physical field space-time acquisition module extracts melt transient electric signal and constructs multi-physical quantity initial matrix, generates space-time gradient tensor flow by using finite difference operation. Riemann orthogonal decoupling module establishes non-Euclidean space model based on nonlinear shear thinning constitutive parameter, and through the orthogonal decomposition of geodesic line, the pure viscous dissipation vector and the pure thermal work vector are stripped out. The phase change cooperative decision module generates shear compensation pulse and temperature adjustment parameter accordingly, which is converted into thrust waveform instruction and steady heat instruction by the instruction pulse issuing module. The injection propulsion servo unit executes mechanical thrust action with high-frequency oscillation, and the barrel matrix heating unit synchronously carries out heat compensation. The present application realizes the deep decoupling of physical parameters and improves the stability of modified ABS molding process.

Description

Technical Field

[0001] This invention relates to the field of injection molding process control technology, and in particular to an adaptive control system for process parameters of modified ABS injection molding. Background Technology

[0002] The adaptive control system is a typical computer-controlled closed-loop control device, primarily used for the dynamic management of process parameters in the injection molding of modified acrylonitrile butadiene styrene copolymer. The modified acrylonitrile butadiene styrene copolymer is formed by polymerizing three monomers to create a base resin, which is then physically blended to become a polymer composite material. This polymer composite material absorbs heat inside the heated barrel, transforming into a flowing melt. Subsequently, under the combined action of mechanical thrust and injection speed, it fills the precision mold cavity. The barrel heating temperature, injection speed, and holding pressure together constitute the core process parameters. The computer control device uses terminal sensors to acquire real-time electrical signals indicating the physical operating status of the processing equipment. The microprocessor compares the received electrical signals with preset target values ​​and calculates the physical deviation. The central microprocessor then issues electrical control commands based on this deviation to autonomously correct the actions of the mold actuator. The adaptive adjustment logic aims to overcome processing state deviations caused by external physical disturbances. The microprocessor's electrical control commands expect to maintain the material molding environment in a constant and ideal state through dynamic compensation. However, the actual modified injection molding manufacturing environment is far more complex than a purely theoretical computer control model.

[0003] Existing high-precision injection molding technology for complex modified ABS components suffers from the following technical challenges: Firstly, the strong coupling between modified ABS process parameters leads to inaccurate decisions in the adaptive control system and fluctuations in product quality. Modified acrylonitrile-butadiene-styrene copolymers involve complex rheological behavior and thermodynamic phase transitions during the plasticizing stage. This results in highly nonlinear correlations between temperature, pressure, and velocity variables when the computer control system performs multivariate closed-loop regulation. Adjusting specific process parameters triggers a chain reaction in related parameters, making it difficult for the control algorithm to accurately decouple multidimensional characteristic variables during the dynamic mold filling stage. Mixed interference signals collected by sensors cause deviations in the underlying calculations, leading to incorrect compensation commands and overshooting of the actuators. For example, when the computer control system increases the barrel heating temperature to improve melt flow, the apparent viscosity of the polymer melt decreases significantly with increasing temperature. Under constant injection conditions, this viscosity decrease directly causes an abnormal increase in pressure inside the mold cavity. Upon receiving this surge signal, the central processing unit immediately issues a command to reduce the injection pressure. This pressure decay then causes surface shrinkage defects far from the gate. The control center is unable to identify the mechanism of the intertwined effects of multiple physical fields in real time, which causes the system commands to oscillate repeatedly between different compensation paths. The conflict of multivariate regulation ultimately causes the processed components to warp and deform. Summary of the Invention

[0004] To address the shortcomings of existing technologies, this invention provides an adaptive control system for the injection molding process parameters of modified ABS. This invention solves the technical problem of inaccurate decision-making and product quality fluctuations caused by the strong coupling between the process parameters of modified ABS.

[0005] To solve the above-mentioned technical problems, the specific contents of the present invention are as follows: The modified ABS injection molding process parameter adaptive control system provided by the present invention includes physical execution equipment and digital control device; The physical execution device includes a barrel matrix heating unit externally mounted on the melt barrel, an injection propulsion servo unit mechanically connected to the propulsion screw, and a mold cavity multi-node sensing array embedded in the inner wall of the molding cavity. The digital control device establishes a communication connection with the physical execution device, including a physical field spatiotemporal acquisition module, a Riemann orthogonal decoupling module, a phase transition collaborative decision-making module, and an instruction pulse sending module; The physical field spatiotemporal acquisition module receives the transient electrical signal of the melt acquired by the multi-node sensing array of the mold cavity, and converts the transient electrical signal of the melt into a spatiotemporal gradient tensor flow; The Riemann orthogonal decoupling module receives the spatiotemporal gradient tensor flow, projects the spatiotemporal gradient tensor flow onto the interior of a preset non-Euclidean space model, and separates the pure viscous dissipation vector and the pure thermal work vector. The phase change collaborative decision module receives the pure viscous dissipation vector and the pure thermodynamic work vector, calculates the topological evolution rate of the pure viscous dissipation vector and generates a shear compensation pulse, calculates the deviation between the pure thermodynamic work vector and the preset temperature reference data and generates a temperature regulation parameter, and encapsulates the shear compensation pulse and the temperature regulation parameter into a collaborative control package. The instruction pulse sending module receives the collaborative control package, converts the collaborative control package into a thrust waveform instruction and a steady-state heat instruction, and sends them to the injection propulsion servo unit and the barrel matrix heating unit, respectively.

[0006] Furthermore, in the modified ABS injection molding process parameter adaptive control system of the present invention, the physical field spatiotemporal acquisition module includes an analog-to-digital conversion unit and a multi-physical quantity matrix construction unit. The analog-to-digital conversion unit receives the transient electrical signal of the melt, performs analog-to-digital conversion on the transient temperature signal, transient pressure signal and transient speed signal included in the transient electrical signal of the melt, and outputs a discrete digital sequence; The multi-physical quantity matrix construction unit receives the discrete digital sequence, extracts the timestamp information and spatial coordinate information included in the discrete digital sequence, and arranges and aligns the discrete digital sequence according to the time dimension and the spatial dimension to construct the initial matrix of multi-physical quantities.

[0007] Furthermore, in the modified ABS injection molding process parameter adaptive control system of the present invention, the physical field spatiotemporal acquisition module further includes a finite difference operation unit. The finite difference operation unit receives the initial matrix of multiple physical quantities, performs finite difference differentiation calculations in both the spatial and temporal dimensions on the initial matrix of multiple physical quantities, generates the spatiotemporal gradient tensor flow, and transmits the spatiotemporal gradient tensor flow to the Riemann orthogonal decoupling module.

[0008] Furthermore, in the modified ABS injection molding process parameter adaptive control system of the present invention, the Riemann orthogonal decoupling module includes a state space modeling unit and a nonlinear projection unit. The state-space modeling unit receives the spatiotemporal gradient tensor flow, reads the pre-stored nonlinear shear-thinning constitutive parameters of the modified acrylonitrile butadiene styrene copolymer, calculates the metric tensor matrix corresponding to the nonlinear shear-thinning constitutive parameters, and constructs the non-Euclidean space model based on the metric tensor matrix. The nonlinear projection unit extracts the Euclidean coordinate data included in the spatiotemporal gradient tensor flow, nonlinearly maps the Euclidean coordinate data to the interior of the non-Euclidean space model, and outputs a set of manifold surface feature points.

[0009] Furthermore, the modified ABS injection molding process parameter adaptive control system of the present invention further includes a geodesic stripping unit in the Riemann orthogonal decoupling module. The geodesic stripping unit receives the set of feature points of the manifold surface, calculates the geodesic equation inside the non-Euclidean space model, and performs orthogonal decomposition calculation on the set of feature points of the manifold surface along the tangent direction and normal direction set by the geodesic equation to generate orthogonal decomposition values. The geodesic stripping unit separates the pure viscous dissipation vector and the pure thermal work vector based on the orthogonal decomposition values.

[0010] Furthermore, in the modified ABS injection molding process parameter adaptive control system of the present invention, the phase change collaborative decision module includes a micro pulse generation unit and a macro temperature comparison unit. The micro-pulse generation unit receives the pure viscous dissipation vector, calculates the topological evolution rate of the pure viscous dissipation vector, converts the topological evolution rate into the corresponding pulse frequency and pulse amplitude values, and generates the shear compensation pulse based on the pulse frequency and pulse amplitude values. The macroscopic temperature comparison unit receives the pure thermal work vector, extracts the feature value of the pure thermal work vector, calculates the deviation of the feature value from the preset temperature reference data, and outputs the deviation as the temperature adjustment parameter.

[0011] Furthermore, in the modified ABS injection molding process parameter adaptive control system of the present invention, the phase change collaborative decision-making module further includes a collaborative data packaging unit. The collaborative data packaging unit receives the shear compensation pulse and the temperature adjustment parameter, splices the shear compensation pulse and the temperature adjustment parameter into a data frame of the collaborative control packet, and sends the collaborative control packet to the command pulse sending module.

[0012] Furthermore, in the modified ABS injection molding process parameter adaptive control system of the present invention, the instruction pulse sending module includes a protocol decomposition and parsing unit and a hardware digital-to-analog conversion unit. The protocol decomposition and parsing unit receives the cooperative control packet, parses the data frame of the cooperative control packet, separates the shear compensation pulse and the temperature regulation parameter, and transmits them to the hardware digital-to-analog conversion unit. The hardware analog-to-digital converter receives the shear compensation pulse and the temperature regulation parameter, converts the temperature regulation parameter into the steady-state heat command, and converts the shear compensation pulse into a high-frequency oscillating waveform analog electrical signal.

[0013] Furthermore, in the modified ABS injection molding process parameter adaptive control system of the present invention, the instruction pulse sending module further includes an instruction high-frequency superposition output unit. The high-frequency superposition output unit receives the steady-state heat command and the high-frequency oscillation waveform analog electrical signal; The high-frequency superposition output unit extracts the preset basic thrust operation command, superimposes and merges the high-frequency oscillation waveform analog electrical signal into the basic thrust operation command, synthesizes the thrust waveform command, and sends the thrust waveform command to the injection propulsion servo unit.

[0014] Furthermore, in the modified ABS injection molding process parameter adaptive control system of the present invention, the high-frequency superposition output unit of the instruction independently sends the steady-state heat instruction to the barrel matrix heating unit. The barrel matrix heating unit receives the steady-state heat command and outputs heat to the inside of the melt barrel according to the steady-state heat command. The injection propulsion servo unit receives the thrust waveform command and drives the propulsion screw to output mechanical thrust to the inside of the molding cavity according to the thrust waveform command.

[0015] Beneficial effects of this invention: The modified ABS injection molding process parameter adaptive control system provided by this invention utilizes a physical field spatiotemporal acquisition module to receive transient electrical signals from the melt and convert them into a spatiotemporal gradient tensor flow. A Riemann orthogonal decoupling module projects the spatiotemporal gradient tensor flow onto a preset non-Euclidean space model, thereby separating mutually independent pure viscous dissipation vectors and pure thermal work vectors. This non-Euclidean space projection decoupling operation overcomes the limitations of conventional linear matrix dimensionality reduction methods, severing the nonlinear strong coupling chain between temperature and pressure parameters from the underlying physical calculation. A phase change collaborative decision-making module performs independent evaluation calculations on the separated pure viscous dissipation vectors and pure thermal work vectors, generating shear compensation pulses and temperature regulation parameters respectively. Independent data evaluation logic avoids erroneous compensation commands output by the computational center due to mixed interference signals, preventing frequent oscillations in the system feedback loop between different compensation paths. A command pulse issuing module parses and converts the collaborative control package into thrust waveform commands and steady-state heat commands. The injection propulsion servo unit applies a mechanical thrust with high-frequency oscillation to the polymer melt inside the molding cavity based on the thrust waveform command. This high-frequency oscillation generates transient mechanical shear force, directly intervening in the microscopic phase transition process of polymer chain entanglement and unwinding. This transient mechanical-physical intervention mechanism instantly offsets the risk of a surge in internal pressure caused by a local decrease in apparent viscosity, without altering the barrel heating state. The barrel matrix heating unit provides heat compensation based solely on steady-state heat commands, maintaining an absolutely constant macroscopic molding thermal environment. This physical heterogeneous synergy between microscopic mechanical thrust fluctuations and macroscopic steady-state heat control solves the technical challenge of decision-making inaccuracies caused by multivariate regulation conflicts. It eliminates surface shrinkage and warpage defects in injection-molded components from the underlying physical execution level, significantly improving the high-precision molding quality stability of modified acrylonitrile butadiene styrene copolymer products. Attached Figure Description

[0016] To more clearly illustrate the technical solution of the present invention, the drawings used in the embodiments will be briefly introduced below. Obviously, for those skilled in the art, other drawings can be obtained based on the drawings without creative effort.

[0017] Figure 1 This is a system architecture diagram of the adaptive control system for the modified ABS injection molding process parameters of the present invention. Detailed Implementation

[0018] To make the technical solution of the present invention clearer, the present invention will be clearly and completely described below with reference to specific embodiments and corresponding drawings. Obviously, the described embodiments are only a part of the embodiments of the present invention, and not all of them. 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. The present invention provided by various embodiments will be described in detail below with reference to the accompanying drawings. To better understand the purpose of the present invention, the present invention will be described in further detail below.

[0019] Please see Figure 1 The modified ABS injection molding process parameter adaptive control system provided by the present invention includes physical execution equipment and digital control device; The physical execution device includes a barrel matrix heating unit externally mounted on the melt barrel, an injection propulsion servo unit mechanically connected to the propulsion screw, and a mold cavity multi-node sensing array embedded in the inner wall of the molding cavity. The digital control device establishes a communication connection with the physical execution device, including a physical field spatiotemporal acquisition module, a Riemann orthogonal decoupling module, a phase transition collaborative decision-making module, and an instruction pulse sending module; The physical field spatiotemporal acquisition module receives the transient electrical signal of the melt acquired by the multi-node sensing array of the mold cavity, and converts the transient electrical signal of the melt into a spatiotemporal gradient tensor flow; The Riemann orthogonal decoupling module receives the spatiotemporal gradient tensor flow, projects the spatiotemporal gradient tensor flow onto the interior of a preset non-Euclidean space model, and separates the pure viscous dissipation vector and the pure thermal work vector. The phase change collaborative decision module receives the pure viscous dissipation vector and the pure thermodynamic work vector, calculates the topological evolution rate of the pure viscous dissipation vector and generates a shear compensation pulse, calculates the deviation between the pure thermodynamic work vector and the preset temperature reference data and generates a temperature regulation parameter, and encapsulates the shear compensation pulse and the temperature regulation parameter into a collaborative control package. The instruction pulse sending module receives the collaborative control package, converts the collaborative control package into a thrust waveform instruction and a steady-state heat instruction, and sends them to the injection propulsion servo unit and the barrel matrix heating unit, respectively.

[0020] The adaptive control system for modified ABS injection molding process parameters includes physical actuators and a digital control unit. The physical actuators include a barrel matrix heating unit external to the melt barrel, an injection propulsion servo unit mechanically connected to the propulsion screw, and a multi-node sensing array embedded in the inner wall of the molding cavity. The barrel matrix heating unit is an array structure composed of multiple independently temperature-controlled heating rings, providing independent steady-state heat output for different heating sections of the melt barrel. The injection propulsion servo unit directly drives the mechanical transmission mechanism via a servo motor, thereby driving the propulsion screw located inside the melt barrel to move axially and apply mechanical thrust. The multi-node sensing array integrates miniature temperature sensors, miniature pressure sensors, and photoelectric speed sensors. The digital control unit establishes a communication connection with the physical actuators via an industrial fieldbus and internally includes a physical field spatiotemporal acquisition module, a Riemann orthogonal decoupling module, a phase change collaborative decision-making module, and a command pulse transmission module. The physical field spatiotemporal acquisition module receives the transient electrical signals of the melt collected by the multi-node sensing array of the mold cavity, converts the transient electrical signals of the melt into a spatiotemporal gradient tensor flow, and transmits it to the Riemann orthogonal decoupling module. The Riemann orthogonal decoupling module projects the spatiotemporal gradient tensor flow onto the interior of a preset non-Euclidean space model, extracts and separates the pure viscous dissipation vector and the pure thermal work vector from the interior of the non-Euclidean space model. The phase transition collaborative decision module receives the pure viscous dissipation vector and the pure thermal work vector, calculates the topological evolution rate of the pure viscous dissipation vector and generates a shear compensation pulse, and simultaneously calculates the deviation of the pure thermal work vector from the preset temperature reference data and generates temperature regulation parameters. Then, the shear compensation pulse and the temperature regulation parameters are encapsulated into a collaborative control package. The command pulse issuing module receives the collaborative control package, parses and converts it into a thrust waveform command and a steady-state heat command, respectively sending the thrust waveform command to the injection propulsion servo unit and the steady-state heat command to the barrel matrix heating unit.

[0021] The analog-to-digital conversion unit (ADC) of the physical field spatiotemporal acquisition module receives transient electrical signals from the melt collected by micro-sensors. These transient electrical signals specifically include transient temperature signals characterizing local temperature changes in the mold cavity, transient pressure signals characterizing the wavefront pressure of the melt flow, and transient velocity signals characterizing the polymer filling flow rate. The ADC unit uses a high-frequency sampling circuit to perform continuous analog-to-digital conversion operations on the transient temperature, pressure, and velocity signals, converting analog waveforms into discrete digital sequences. The multi-physical quantity matrix construction unit receives the discrete digital sequence and extracts the timestamp and spatial coordinate information included in the sequence. The timestamp records the absolute time node of the sensor sampling, and the spatial coordinate information identifies the three-dimensional coordinate point of the signal origin within the molding cavity. The multi-physical quantity matrix construction unit arranges and aligns the discrete digital sequence according to the time and spatial dimensions, constructing an initial multi-physical quantity matrix in memory that includes a four-dimensional coordinate system with time and three-dimensional space. The finite difference operation unit receives an initial matrix of multiple physical quantities and performs spatial and temporal finite difference derivative calculations on data points at adjacent time and spatial nodes in the initial matrix to obtain the rate of change of the physical quantities during the passage of time and spatial propagation. Based on this rate of change, the finite difference operation unit generates a spatiotemporal gradient tensor flow and transmits it to the Riemann orthogonal decoupling module.

[0022] The spatiotemporal gradient tensor flow is the fundamental input data of the algorithm model. Its data structure is defined as a four-dimensional tensor matrix including temperature, pressure, and velocity partial derivative components. As a core physical parameter characterizing the transient rheological state of the melt, the spatiotemporal gradient tensor flow is obtained through partial differential operations on discretized multiphysics signals at continuous time slices and discrete spatial nodes. The pure viscous dissipation vector and the pure thermal work vector constitute the orthogonal basis decomposition form of the spatiotemporal gradient tensor flow on a non-Euclidean manifold. The pure viscous dissipation vector represents the mechanical energy component consumed by the melt to overcome flow resistance, while the pure thermal work vector represents the thermal energy component exchanged between the melt and the mold cavity. The topological evolution rate is defined as the dynamic evolution speed of the pure viscous dissipation vector on the Riemannian space geometry, reflecting the degree of change in the microscopic topological structure of long-chain polymer molecules under shear force. These custom physical terms establish a deterministic mapping relationship between the microscopic rheological mathematical model and the macroscopic injection molding process control variables.

[0023] The dimensions of the four-dimensional tensor matrix have a fixed mapping relationship with the physical space inside the molding cavity and the filling time. The first dimension of the four-dimensional tensor matrix maps to the horizontal coordinates of the three-dimensional space inside the molding cavity, the second dimension maps to the vertical coordinates of the three-dimensional space inside the molding cavity, the third dimension maps to the vertical coordinates of the three-dimensional space inside the molding cavity, and the fourth dimension maps to the absolute time node of the processing equipment operation. The four-dimensional tensor matrix completely digitizes the complex physical state of the modified acrylonitrile butadiene styrene copolymer material during the filling stage. The Jacobian matrix transformation operation constitutes the core operation logic of the nonlinear projection unit. The Jacobian matrix transformation operation receives Euclidean coordinate data as the calculation input data. The Euclidean coordinate data includes mutually interfering temperature characteristics and pressure characteristics. The metric tensor matrix serves as a key constraint parameter in the Jacobian matrix transformation operation. Its internal elements are directly related to the consistency coefficient and the non-Newtonian exponent input parameter. The consistency coefficient defines the baseline value of the static flow resistance of the base resin, while the non-Newtonian exponent defines the dynamic nonlinearity of shear thinning in the polymer composite material after physical blending. The Jacobian matrix transformation operation reconstructs the Euclidean coordinate data based on the spatial curvature information provided by the metric tensor matrix. The tensor reconstruction operation outputs a set of manifold surface feature points as the output data of the algorithm model. These manifold surface feature points exhibit a mathematical distribution state where temperature and pressure parameters are independent and orthogonally isolated at the underlying topological dimension. The inherent input-output data mapping logic achieves a precise physical-mathematical conversion from macroscopic multivariate molding process parameters to microscopic constitutive rheological states.

[0024] The state-space modeling unit within the Riemann orthogonal decoupling module receives the spatiotemporal gradient tensor flow and reads the pre-stored nonlinear shear-thinning constitutive parameters of the modified acrylonitrile-butadiene-styrene copolymer from memory. During the construction of the non-Euclidean space model, the preset initial parameter set includes rheological boundary conditions determined by the material properties of the modified acrylonitrile-butadiene-styrene copolymer. The specific form of the non-Euclidean space model is defined by the curvature distribution function of the metric tensor matrix, with each element corresponding to the material shear viscosity response value measured in physical experiments. The preset temperature reference data is stored in the static random access memory of the digital control device, and its numerical range is limited to the upper and lower limits of the melting temperature range of the modified acrylonitrile-butadiene-styrene copolymer. The selection rules for the preset temperature reference data are determined by the technical requirements of the injection molding process for the finished product shrinkage rate, ensuring that the non-Euclidean space model and the preset temperature reference data together provide a physically deterministic decoupling reference system for the phase change collaborative decision-making module.

[0025] The nonlinear shear-thinning constitutive parameters include the consistency coefficient and non-Newtonian exponent, which characterize the change of polymer viscosity with shear rate. The state-space modeling unit calculates the metric tensor matrix describing the fluid deformation behavior based on the consistency coefficient and non-Newtonian exponent, and constructs a non-Euclidean space model based on the curvature characteristics of the metric tensor matrix. The nonlinear projection unit extracts the Euclidean coordinate data from the spatiotemporal gradient tensor flow, and uses the Jacobian matrix transformation algorithm to nonlinearly map the Euclidean coordinate data to the interior of the non-Euclidean space model, eliminating the linearity artifacts of rheological physical quantities and outputting a set of manifold surface feature points that truly reflects the rheological state of the polymer. The geodesic stripping unit receives the set of manifold surface feature points, calculates the geodesic equation for the shortest path between two points within the non-Euclidean space model, and performs orthogonal decomposition calculations on the manifold surface feature point set along the tangent and normal directions defined by the geodesic equation, generating orthogonal decomposition values. The tangent component represents the state change caused by pure flow deformation, and the normal component represents the state change caused by pure environmental heat exchange. Based on the orthogonal decomposition numerical stripping of the mixed physical characteristics, the geodesic stripping element separates the pure viscous dissipation vector that independently characterizes shear heat generation and flow resistance, and the pure thermodynamic work vector that independently characterizes boundary heat conduction.

[0026] The underlying geometric topology construction of the non-Euclidean space model is the core digital foundation for achieving orthogonal mapping decoupling operations of multiple physical quantities. The state-space modeling unit generates a digital virtual multidimensional surface mesh in the computer's memory space based on the curvature characteristic parameters reflected in the calculated and extracted metric tensor matrix. This digital virtual multidimensional surface mesh system constitutes the main data skeleton of the non-Euclidean space model. The element matrix distribution pattern within the metric tensor matrix determines the degree of spatial curvature of the non-Euclidean space model in different physical coordinate directions, allowing the nonlinear rheological physical properties exhibited by the polymer melt under extreme dynamic strong shear to be directly mathematically mapped to the geometric distortion and deformation characteristics within the non-Euclidean space model. The nonlinear projection unit receives and extracts Euclidean coordinate data generated by the initial matrix transformation of multiple physical quantities. This Euclidean coordinate data includes temperature and pressure variables exhibiting highly linear mixed interference. Using the Jacobian matrix transformation operation logic under a specific constraint algorithm, the nonlinear projection unit maps the input Euclidean coordinate data point-to-point to the corresponding node coordinates of the digital virtual multidimensional surface mesh, and assembles and outputs the three-dimensional numerical sequence attached to the new coordinate nodes as a set of manifold surface feature points. Multidimensional projection operations force the linear physical coupling data collected by macroscopic terminal sensors into a nonlinear separation state from a microscopic perspective, enabling the geodesic stripping unit to find the physical state evolution path with the least rheological resistance along a specific geometric surface, thereby accurately extracting independent control components that do not interfere with each other.

[0027] The micro-pulse generation unit in the phase transition collaborative decision-making module receives the pure viscous dissipation vector and calculates the topological evolution rate of the vector over time. This rate reflects the microscopic intensity of entanglement or unwinding of polymer chain segments in a specific region. The micro-pulse generation unit converts the topological evolution rate into corresponding pulse frequency and amplitude values, and generates shear compensation pulses with specific oscillating waveforms based on these values. The conversion from topological evolution rate to pulse frequency and amplitude values ​​follows a linear mapping function rule. The specific implementation of the conversion logic relies on a preset set of conversion constant coefficients in the digital control device, including frequency and amplitude conversion constant coefficients. The pulse frequency value is obtained by multiplying the topological evolution rate by the frequency conversion constant coefficient, and the pulse amplitude value is also obtained by multiplying the topological evolution rate by the amplitude conversion constant coefficient. The values ​​of the frequency and amplitude conversion constant coefficients are determined by the dynamic response bandwidth of the injection propulsion servo unit and the shear thinning critical value of the modified acrylonitrile-butadiene-styrene copolymer.

[0028] The macroscopic temperature comparison unit receives the pure thermodynamic work vector, extracts the characteristic values ​​representing the average thermal energy state from the pure thermodynamic work vector, and calculates the difference between the characteristic values ​​and the preset temperature reference data. The difference reflects the degree to which the barrel heating deviates from the ideal molding temperature. The macroscopic temperature comparison unit directly outputs the calculated difference as a temperature adjustment parameter. The collaborative data packaging unit receives the shear compensation pulse and the temperature adjustment parameter, concatenates the waveform digital bit stream included in the shear compensation pulse and the amplitude digital bit stream included in the temperature adjustment parameter according to the industrial communication protocol standard format, generates a collaborative control packet data frame including a complete control information frame, and sends the collaborative control packet to the command pulse sending module.

[0029] The protocol decomposition and parsing unit inside the command pulse issuing module receives the collaborative control packet. It parses the data frame of the collaborative control packet according to the preset frame header and tail identifiers, losslessly separating the shear compensation pulse and temperature adjustment parameters of the independent control logic, and transmitting them to the hardware digital-to-analog converter (DAC). The DAC receives the shear compensation pulse and temperature adjustment parameters, converting the temperature adjustment parameters into a steady-state heat command for adjusting the heating resistor power. Simultaneously, it uses an internally integrated signal generator to convert the shear compensation pulse into a high-frequency oscillating waveform analog electrical signal with high-frequency pulsation characteristics. The command high-frequency superposition output unit receives the steady-state heat command and the high-frequency oscillating waveform analog electrical signal, extracts the basic thrust operation command issued by the main control system to maintain the normal mold filling process, and superimposes and merges the high-frequency oscillating waveform analog electrical signal into the basic thrust operation command to synthesize a thrust waveform command with micro-oscillation characteristics. The command high-frequency superposition output unit sends the thrust waveform command to the injection propulsion servo unit and independently sends a steady-state heat command to the barrel matrix heating unit. The barrel matrix heating unit receives steady-state heat commands and outputs constant heat to the inside of the melt barrel to maintain a stable macroscopic temperature field. The injection propulsion servo unit receives thrust waveform commands and drives the propulsion screw to add high-frequency micro-amplitude vibrations to the basic propulsion action. The propulsion screw outputs mechanical thrust with high-frequency vibrations to the polymer melt inside the molding cavity. The high-frequency micro-amplitude vibrations of the mechanical thrust can generate transient mechanical shear forces, directly intervening in the microscopic process of polymer chain entanglement and uncoupling. The underlying control structure uses high-frequency mechanical thrust to counteract the impact of the surge in apparent viscosity of the melt, while not changing the macroscopic heating state of the barrel. The physical-mechanical intervention mechanism cuts off the physical transmission chain that would inevitably cause an abnormal increase in the overall pressure of the mold cavity when adjusting a single parameter, achieving precise decoupling and adaptive adjustment of multiple physical field variables in the molding process of complex modified ABS components. The thrust waveform commands directly act on the current loop controller of the injection propulsion servo unit, generating a high-frequency shear stress field by changing the axial oscillation displacement of the propulsion screw. High-frequency shear stress fields are used to adjust the apparent viscosity distribution of the modified acrylonitrile-butadiene-styrene copolymer melt in real time, achieving physical-level real-time compensation for pressure fluctuations inside the mold cavity. Steady-state heat commands synchronously drive the power solid-state relays of the barrel matrix heating unit, adjusting the heat flux through the barrel's inner wall. Microscopic kinetic energy compensation generated by thrust waveform commands and macroscopic thermal energy compensation generated by steady-state heat commands form a heterogeneous physical feedback loop inside the melt barrel. This heterogeneous physical feedback loop ensures that the decoupled calculation results of the non-Euclidean space model can be directly converted into precise mechanical movements of the injection molding machine's actuators, eliminating the technical gap between pure algorithm control and the physical execution end.

[0030] The adaptive control system for modified ABS injection molding process parameters sets multi-level system safety execution thresholds in the underlying computational kernel of the phase change collaborative decision-making module, and embeds macroscopic temperature adjustment threshold parameters within the macroscopic temperature comparison unit, limiting the single physical temperature adjustment range to no more than 3.0 degrees Celsius. Limiting a single adjustment to a maximum of 3.0 degrees Celsius effectively prevents the barrel matrix heating unit from receiving extreme overload signals, thereby avoiding irreversible thermal degradation of the modified acrylonitrile-butadiene-styrene copolymer material within the narrow flow channel. The microscopic pulse generation unit's data channel sets both a pulse frequency execution threshold and a pulse amplitude control threshold. The pulse frequency execution threshold limits the upper limit of the digital waveform to 50.0 Hz, and the pulse amplitude control threshold limits the upper limit of the output mechanical waveform to 15.0 MPa. A pulse frequency execution threshold with an extreme value of 50.0 Hz can prevent high-frequency oscillation and metal fatigue damage to the mechanical transmission mechanism and screw bearings inside the injection propulsion servo unit. A pulse amplitude control threshold with an extreme value of 15.0 MPa can prevent excessive instantaneous impact force from the physical propulsion screw onto the closed mold cavity, thus preventing mold expansion and product overflow risks. Before splicing the calculated data to generate the collaborative control package, the phase change collaborative decision module extracts the digital bitstream of the shear compensation pulse and the digital bitstream of the temperature regulation parameters. The extracted digital bitstream values ​​are logically compared with the set execution threshold values. Abnormal data values ​​exceeding the control threshold boundaries are reduced to the system's allowable safe critical value range by the phase change collaborative decision module using a forced filtering algorithm. The physical extreme value threshold limitation logic protects the long-term operating life of the high-intensity physical execution equipment and the overall processing safety.

[0031] The finite difference operation unit receives an initial matrix of multiple physical quantities, including spatial and temporal node data. The finite difference operation unit performs spatial-dimensional finite difference differentiation and temporal-dimensional finite difference differentiation on the initial matrix of multiple physical quantities. Taking the calculation of a transient temperature signal as an example, the temporal-dimensional finite difference differentiation model is defined as follows:

[0032] In the formula, Represents the value of the transient temperature signal. Represents the sign of the partial derivative. Represents the current absolute time point. Represents the next absolute time point. Represents spatial coordinates. This represents the time span between adjacent absolute time points. The spatial finite-difference derivative calculation model is defined as follows:

[0033] In the formula, Represents the value of the transient temperature signal. Represents the sign of the partial derivative. Represents the current absolute time point. Represents the current spatial coordinates. Represents the next adjacent spatial coordinate point. This represents the physical distance between adjacent spatial coordinate points. The finite difference operation unit simultaneously processes transient pressure and velocity signals according to the aforementioned finite difference derivative calculation model, combining the calculated partial derivatives with respect to the time and spatial dimensions into a matrix, and outputting a spatiotemporal gradient tensor flow including the rate of change of multidimensional flow states. The state-space modeling unit reads the pre-stored nonlinear shear-thinning constitutive parameters of the modified acrylonitrile butadiene-styrene copolymer. The nonlinear shear-thinning constitutive parameters specifically include the consistency coefficient and the non-Newtonian exponent. The state-space modeling unit calculates the metric tensor matrix corresponding to the consistency coefficient and the non-Newtonian exponent. The metric tensor matrix calculation model is defined as follows:

[0034] In the formula, Represents the elements within the metric tensor matrix. Represents the consistency coefficient. This represents the local microscopic shear rate extracted from the spatiotemporal gradient tensor flow. Represents non-Newtonian exponents. Representing the Kronecker notation. The state-space modeling unit constructs a non-Euclidean space model based on the metric tensor matrix. The nonlinear projection unit extracts the Euclidean coordinate data included in the spatiotemporal gradient tensor flow and nonlinearly maps the Euclidean coordinate data to the interior of the non-Euclidean space model. The Jacobian matrix nonlinear mapping calculation model is defined as follows:

[0035] In the formula, This represents the set of feature points of the manifold surface generated within the non-Euclidean space model. This represents the Jacobian transformation matrix constructed by taking the partial derivatives of the Euclidean coordinate system with respect to the non-Euclidean coordinate system. The Euclidean coordinate data representing the spatiotemporal gradient tensor flow.

[0036] The geodesic stripping element receives a set of feature points from the manifold surface and calculates the geodesic equations within the non-Euclidean space model. The geodesic equations are defined as follows:

[0037] In the formula, Represents the differential symbol. The coordinates representing the target dimension within the non-Euclidean space model. The parameter representing the arc length on the manifold surface. The Christopher notation is derived from the metric tensor matrix. The coordinates of the first independent variable dimension within the non-Euclidean space model. The coordinates represent the second independent variable dimension within the non-Euclidean space model. The geodesic stripping element performs orthogonal decomposition calculations on the feature point set of the manifold surface along the tangent and normal directions defined by the geodesic equation. The orthogonal decomposition calculation model is defined as follows:

[0038] In the formula, Represents the pure viscous dissipation vector. Represents the set of feature points of the manifold surface. This represents the unit vector along the tangent direction defined by the geodesic equation. Represents the vector of work done by pure heat. This represents the unit vector along the normal direction defined by the geodesic equation. The geodesic stripping unit, based on the calculated orthogonal decomposition values, strips away the multi-physics coupling state, separating the independent pure viscous dissipation vector and the pure thermal work vector. The microscopic pulse generation unit in the phase transition co-decision module receives the pure viscous dissipation vector and calculates its topological evolution rate. The topological evolution rate calculation model is defined as:

[0039] In the formula, Represents the rate of topological evolution. Represents the differential symbol. Represents the pure viscous dissipation vector. This represents the time parameter. The micro-pulse generation unit converts the topological evolution rate into corresponding pulse frequency and amplitude values. The micro-pulse conversion model is defined as follows:

[0040]

[0041] In the formula, Represents the pulse frequency value. This represents the pre-defined frequency conversion constant coefficient. Represents the rate of topological evolution. Represents the pulse amplitude value. This represents the pre-set amplitude conversion constant coefficient. The micro-pulse generation unit generates a shear compensation pulse based on the pulse frequency and pulse amplitude values. The macro-temperature comparison unit receives the pure thermodynamic work vector, extracts the characteristic values ​​of the pure thermodynamic work vector, and calculates the deviation between the characteristic values ​​and the preset temperature reference data. The deviation calculation model is defined as follows:

[0042] In the formula, Represents the deviation amount. The characteristic value representing the vector of work done by pure heat. This represents the preset temperature reference data. The macroscopic temperature comparison unit outputs the deviation as a temperature adjustment parameter.

[0043] To verify the feasibility of the multiphysics decoupling operation logic, the system executed a complete numerical processing embodiment for the precision injection molding process of modified ABS components. During the spatiotemporal acquisition phase of the physics field, the time span between adjacent absolute time nodes set in the initial matrix of the multiphysics quantities was set to... Seconds. The transient temperature signal value at the current absolute time point is acquired as follows: The transient temperature signal value at the next absolute time point is acquired in degrees Celsius. Degrees Celsius. Substituting the system into a finite-difference time-division derivative calculation model, the time partial derivative with respect to the temperature variable is obtained as follows: Degrees Celsius per second. In the state-space modeling phase, the system reads pre-stored material property data of the modified acrylonitrile-butadiene-styrene copolymer, where the consistency coefficient value is... The non-Newtonian exponent value is The numerical value of the microscopic local shear rate extracted in real time from the spatiotemporal gradient tensor flow is: The Kronecker symbol on the main diagonal takes a value per second. Substituting the numerical values ​​of each feature into the metric tensor matrix calculation model, the values ​​of the corresponding main diagonal elements within the metric tensor matrix are calculated as follows: In the geodesic decomposition stage, the set of feature points of the manifold surface generated by the nonlinear mapping is input into the geodesic stripping unit. After orthogonal decomposition calculation, the magnitude of the pure viscous dissipation vector projected onto the unit vector of the geodesic tangent direction is obtained. In the phase transition co-decision stage, the numerical value of the topological evolution rate is obtained through differential operations. The micro-pulse generation unit extracts the frequency conversion constant coefficient value from the memory. The amplitude conversion constant coefficients were extracted as follows: Substituting the values ​​into the microscopic pulse conversion model, the pulse frequencies were calculated as follows: Hertz, pulse amplitude value is Megapascals. The feature values ​​of the pure thermodynamic work vector extracted from the macroscopic temperature comparison unit are: Degrees Celsius, retrieve the preset temperature reference data under ideal process conditions. Substituting Celsius into the deviation calculation model yields a negative deviation value. Degrees Celsius. Based on the calculated data stream, the digital control unit ultimately continuously outputs superimposed values ​​to the injection propulsion servo unit. Hertz frequency same The thrust command of the megapascal amplitude oscillation waveform is synchronously and independently issued to the barrel matrix heating unit for lifting. Steady-state heat command output in degrees Celsius.

[0044] To verify the multi-physics field decoupling operation logic and action coordination effect of the adaptive control system for modified ABS injection molding process parameters, a set of specific operating numerical configuration examples are provided for the high-precision injection molding process of complex modified acrylonitrile butadiene styrene copolymer components. In the physical execution device, the barrel matrix heating unit externally mounted on the melt barrel is configured as an electrothermal control matrix including five independent heating zones. The mold cavity multi-node sensing array embedded in the inner wall of the molding cavity is equipped with miniature thermocouples, miniature piezoelectric sensors, and a laser Doppler velocimetry probe.

[0045] During the dynamic filling stage, the physical field spatiotemporal acquisition module is communicatively connected to the multi-node sensing array of the mold cavity, continuously receiving transient electrical signals of the melt inside the mold cavity. The analog-to-digital conversion unit receives the transient electrical signals of the melt and performs analog-to-digital conversion operations at a sampling frequency of 1000 Hz on the transient temperature, transient pressure, and transient velocity signals included in the transient electrical signals of the melt, outputting discrete digital sequences of the corresponding physical quantity channels. The multi-physical quantity matrix construction unit receives the discrete digital sequences, extracts the timestamp information and three-dimensional spatial coordinate information included in the discrete digital sequences, aligns the discrete digital sequences according to the time and spatial dimensions, and constructs an initial multi-physical quantity matrix in the system memory. The finite difference operation unit receives the initial multi-physical quantity matrix, extracts the time span value between adjacent absolute time nodes set to 0.01 seconds, and extracts the physical distance value between adjacent spatial coordinate points set to 2 millimeters. The finite difference operation unit performs time-difference differentiation by subtracting the data points corresponding to adjacent time nodes in the initial matrix of multi-physical quantities and dividing by the time span value. Simultaneously, it performs spatial-difference differentiation by subtracting the data points corresponding to adjacent spatial nodes in the initial matrix of multi-physical quantities and dividing by the physical distance value. After finite difference differentiation, the finite difference operation unit obtains the multidimensional flow state change rate and generates a spatiotemporal gradient tensor flow, which is then transmitted to the Riemann orthogonal decoupling module.

[0046] The state-space modeling unit in the Riemann orthogonal decoupling module receives the spatiotemporal gradient tensor flow and reads the nonlinear shear-thinning constitutive parameters of the modified acrylonitrile-butadiene-styrene copolymer pre-loaded into the system memory. The state-space modeling unit extracts the consistency coefficient (value 12000) and the non-Newtonian exponent (value 0.38) from the nonlinear shear-thinning constitutive parameters. Combined with the local microscopic shear rate extracted from the spatiotemporal gradient tensor flow, it calculates the metric tensor matrix corresponding to the nonlinear shear-thinning constitutive parameters. The state-space modeling unit constructs a non-Euclidean space model based on the non-Euclidean geometric curvature characteristics reflected by the metric tensor matrix. The nonlinear projection unit extracts the Euclidean coordinate data included in the spatiotemporal gradient tensor flow and uses Jacobian matrix transformation to nonlinearly map the Euclidean coordinate data to the interior of the non-Euclidean space model, eliminating the linearization of physical quantities and outputting a set of manifold surface feature points. The geodesic stripping unit receives the set of manifold surface feature points and calculates the geodesic equations for the shortest distance between connected state points within the non-Euclidean space model based on differential geometry principles. The geodesic stripping unit performs orthogonal decomposition calculations on the feature point set of the manifold surface along the tangent and normal directions defined by the geodesic equation, generating orthogonal decomposition values. The tangent direction represents the pure fluid deformation state change component, and the normal direction represents the pure heat exchange state change component. Based on the orthogonal decomposition values, the geodesic stripping unit strips the mixed parameter features, separating a pure viscous dissipation vector with a modulus of 45.8 and a pure thermal work vector with a modulus of 12.6.

[0047] The microscopic pulse generation unit in the phase transition collaborative decision-making module receives the pure viscous dissipation vector and calculates its topological evolution rate based on the gradient of the pure viscous dissipation vector over time, obtaining a topological evolution rate value of 15.6. The microscopic pulse generation unit extracts a frequency conversion constant coefficient of 2.0 and an amplitude conversion constant coefficient of 0.5 (preset value). Multiplying the topological evolution rate by the frequency conversion constant coefficient yields a pulse frequency value of 31.2 Hz, and multiplying the topological evolution rate by the amplitude conversion constant coefficient yields a pulse amplitude value of 7.8 MPa. The microscopic pulse generation unit generates a shear compensation pulse based on the 31.2 Hz pulse frequency and the 7.8 MPa pulse amplitude. The macroscopic temperature comparison unit receives the pure thermodynamic work vector and extracts a characteristic value representing the average thermal energy state of the pure thermodynamic work vector, which is 228.0 degrees Celsius. The macroscopic temperature comparison unit extracts preset temperature reference data set at 230.0 degrees Celsius under ideal molding conditions, calculates the deviation obtained by subtracting the feature value from the preset temperature reference data, and obtains the deviation value as -2.0 degrees Celsius. The macroscopic temperature comparison unit outputs the deviation of -2.0 degrees Celsius as a temperature adjustment parameter. The collaborative data packaging unit receives the shear compensation pulse and the temperature adjustment parameter, concatenates the digital bit streams included in the shear compensation pulse and the digital bit streams included in the temperature adjustment parameter according to the industrial fieldbus protocol, and sends the collaborative control packet to the command pulse sending module.

[0048] The protocol decomposition and parsing unit in the command pulse transmission module receives the cooperative control packet, parses the data frame of the cooperative control packet according to the frame header and frame tail identifiers, separates the independently encoded shear compensation pulse and temperature adjustment parameters, and transmits them to the hardware digital-to-analog conversion unit. The hardware digital-to-analog conversion unit receives the shear compensation pulse and temperature adjustment parameters, converts the temperature adjustment parameters into a steady-state heat command to improve the power output of the corresponding heating zone, and uses an internal signal generation circuit to convert the shear compensation pulse into a high-frequency oscillation waveform analog electrical signal corresponding to 31.2 Hz and 7.8 MPa. The command high-frequency superposition output unit receives the steady-state heat command and the high-frequency oscillation waveform analog electrical signal, and extracts the basic thrust operation command set to 80.0 MPa from the main control program. The command high-frequency superposition output unit superimposes and merges the high-frequency oscillation waveform analog electrical signal into the basic thrust operation command, synthesizing a thrust waveform command with a center reference of 80.0 MPa and an accompanying 31.2 Hz frequency oscillation. The command high-frequency superposition output unit sends the thrust waveform command to the injection propulsion servo unit, and simultaneously sends the steady-state heat command independently to the barrel matrix heating unit. The barrel matrix heating unit receives steady-state heat commands and, based on these commands, increases the output heat by 2 degrees Celsius to the inside of the melt barrel, adding to the existing heating power. The injection propulsion servo unit receives thrust waveform commands and, based on these commands, drives the propulsion screw to perform a small oscillation motion with a base thrust of 80.0 MPa and an amplitude of 7.8 MPa, outputting a mechanical thrust with high-frequency oscillation to the polymer melt inside the molding cavity.

Claims

1. An adaptive control system for process parameters of modified ABS injection molding, characterized in that, Includes physical execution devices and digital control devices; The physical execution device includes a barrel matrix heating unit externally mounted on the melt barrel, an injection propulsion servo unit mechanically connected to the propulsion screw, and a mold cavity multi-node sensing array embedded in the inner wall of the molding cavity. The digital control device establishes a communication connection with the physical execution device, including a physical field spatiotemporal acquisition module, a Riemann orthogonal decoupling module, a phase transition collaborative decision-making module, and an instruction pulse sending module; The physical field spatiotemporal acquisition module receives the transient electrical signal of the melt acquired by the multi-node sensing array of the mold cavity, and converts the transient electrical signal of the melt into a spatiotemporal gradient tensor flow; The Riemann orthogonal decoupling module receives the spatiotemporal gradient tensor flow, projects the spatiotemporal gradient tensor flow onto the interior of a preset non-Euclidean space model, and separates the pure viscous dissipation vector and the pure thermal work vector. The phase change collaborative decision module receives the pure viscous dissipation vector and the pure thermodynamic work vector, calculates the topological evolution rate of the pure viscous dissipation vector and generates a shear compensation pulse, calculates the deviation between the pure thermodynamic work vector and the preset temperature reference data and generates a temperature regulation parameter, and encapsulates the shear compensation pulse and the temperature regulation parameter into a collaborative control package. The instruction pulse sending module receives the collaborative control package, converts the collaborative control package into a thrust waveform instruction and a steady-state heat instruction, and sends them to the injection propulsion servo unit and the barrel matrix heating unit, respectively.

2. The adaptive control system for modified ABS injection molding process parameters according to claim 1, characterized in that, The physical field spatiotemporal acquisition module includes an analog-to-digital conversion unit and a multi-physical quantity matrix construction unit; The analog-to-digital conversion unit receives the transient electrical signal of the melt, performs analog-to-digital conversion on the transient temperature signal, transient pressure signal and transient speed signal included in the transient electrical signal of the melt, and outputs a discrete digital sequence; The multi-physical quantity matrix construction unit receives the discrete digital sequence, extracts the timestamp information and spatial coordinate information included in the discrete digital sequence, and arranges and aligns the discrete digital sequence according to the time dimension and the spatial dimension to construct the initial matrix of multi-physical quantities.

3. The adaptive control system for modified ABS injection molding process parameters according to claim 2, characterized in that, The physical field spatiotemporal acquisition module also includes a finite difference operation unit; The finite difference operation unit receives the initial matrix of multiple physical quantities, performs finite difference differentiation calculations in both the spatial and temporal dimensions on the initial matrix of multiple physical quantities, generates the spatiotemporal gradient tensor flow, and transmits the spatiotemporal gradient tensor flow to the Riemann orthogonal decoupling module.

4. The adaptive control system for modified ABS injection molding process parameters according to claim 1, characterized in that, The Riemann orthogonal decoupling module includes a state-space modeling unit and a nonlinear projection unit; The state-space modeling unit receives the spatiotemporal gradient tensor flow, reads the pre-stored nonlinear shear-thinning constitutive parameters of the modified acrylonitrile butadiene styrene copolymer, calculates the metric tensor matrix corresponding to the nonlinear shear-thinning constitutive parameters, and constructs the non-Euclidean space model based on the metric tensor matrix. The nonlinear projection unit extracts the Euclidean coordinate data included in the spatiotemporal gradient tensor flow, nonlinearly maps the Euclidean coordinate data to the interior of the non-Euclidean space model, and outputs a set of manifold surface feature points.

5. The adaptive control system for modified ABS injection molding process parameters according to claim 4, characterized in that, The Riemann orthogonal decoupling module also includes a geodesic stripping unit; The geodesic stripping unit receives the set of feature points of the manifold surface, calculates the geodesic equation inside the non-Euclidean space model, and performs orthogonal decomposition calculation on the set of feature points of the manifold surface along the tangent direction and normal direction set by the geodesic equation to generate orthogonal decomposition values. The geodesic stripping unit separates the pure viscous dissipation vector and the pure thermal work vector based on the orthogonal decomposition values.

6. The adaptive control system for modified ABS injection molding process parameters according to claim 1, characterized in that, The phase change collaborative decision-making module includes a micro-pulse generation unit and a macro-temperature comparison unit; The microscopic pulse generation unit receives the pure viscous dissipation vector, calculates the topological evolution rate of the pure viscous dissipation vector, converts the topological evolution rate into the corresponding pulse frequency and pulse amplitude values, and generates the shear compensation pulse based on the pulse frequency and pulse amplitude values. The macroscopic temperature comparison unit receives the pure thermal work vector, extracts the feature value of the pure thermal work vector, calculates the deviation of the feature value from the preset temperature reference data, and outputs the deviation as the temperature adjustment parameter.

7. The adaptive control system for modified ABS injection molding process parameters according to claim 6, characterized in that, The phase change collaborative decision-making module also includes a collaborative data packaging unit; The collaborative data packaging unit receives the shear compensation pulse and the temperature adjustment parameter, splices the shear compensation pulse and the temperature adjustment parameter into a data frame of the collaborative control packet, and sends the collaborative control packet to the command pulse sending module.

8. The adaptive control system for modified ABS injection molding process parameters according to claim 1, characterized in that, The instruction pulse sending module includes a protocol decomposition and parsing unit and a hardware digital-to-analog conversion unit; The protocol decomposition and parsing unit receives the cooperative control packet, parses the data frame of the cooperative control packet, separates the shear compensation pulse and the temperature regulation parameter, and transmits them to the hardware digital-to-analog conversion unit. The hardware analog-to-digital converter receives the shear compensation pulse and the temperature regulation parameter, converts the temperature regulation parameter into the steady-state heat command, and converts the shear compensation pulse into a high-frequency oscillating waveform analog electrical signal.

9. The adaptive control system for modified ABS injection molding process parameters according to claim 8, characterized in that, The instruction pulse sending module also includes an instruction high-frequency superposition output unit; The high-frequency superposition output unit receives the steady-state heat command and the high-frequency oscillation waveform analog electrical signal; The high-frequency superposition output unit extracts the preset basic thrust operation command, superimposes and merges the high-frequency oscillation waveform analog electrical signal into the basic thrust operation command, synthesizes the thrust waveform command, and sends the thrust waveform command to the injection propulsion servo unit.

10. The adaptive control system for modified ABS injection molding process parameters according to claim 9, characterized in that, The high-frequency superposition output unit of the command independently sends the steady-state heat command to the barrel matrix heating unit; The barrel matrix heating unit receives the steady-state heat command and outputs heat to the inside of the melt barrel according to the steady-state heat command. The injection propulsion servo unit receives the thrust waveform command and drives the propulsion screw to output mechanical thrust to the inside of the molding cavity according to the thrust waveform command.

Images