22 results about "Process development" patented technology

The invention relates to an intelligent process research and development system and method based on AI model closed-loop iteration, and belongs to the field of intelligent manufacturing and artificial intelligence. In order to solve the problems that in existing process research and development, due to the fact that links of knowledge analysis, model optimization and experimental verification are separated, the efficiency is low, and manual experience is highly relied on, the method comprises the steps that unstructured process knowledge is automatically analyzed into machine executable scripts, and new process parameters are recommended according to experimental data through a process optimization model. The core of the method is that the system automatically feeds back verification results of new parameters, updates the verification results to a data set, and drives a model to carry out next round of iterative training, so that an automatic closed loop from knowledge to the model to an experiment is constructed, and continuous self-evolution and efficient convergence of process research and development are realized.

This application relates to a method and apparatus for determining process parameters in metal compound thin film deposition. The method constructs a two-stage cascaded process prediction model. The first stage, a feature extraction model, performs nonlinear transformations and feature extraction on the process parameters, extracting intermediate feature information that comprehensively characterizes the sputtering process state. The second stage, a probabilistic prediction model, establishes a probabilistic mapping relationship between the intermediate feature information and the thin film response information, decoupling the complex nonlinear coupling relationships between multiple parameters in reactive sputtering. It outputs predicted values ​​and corresponding confidence information. Based on the predicted values ​​and confidence information, process parameters are screened, considering both target performance requirements and prediction reliability requirements. This improves the efficiency and reliability of process parameter screening, reduces the number of invalid experiments, and enables the prediction of key response quantities such as thin film deposition rate and chemical composition even with limited experimental data, shortening the process development cycle.

The invention provides a wheel product heat treatment process parameter optimization method and system, and the method comprises the steps: taking the production parameters of wheels of different furnaces and corresponding equipment health vectors as features, taking the performances of the corresponding wheels as labels, and constructing sample data, the production parameters including a heating parameter, a cooling parameter and a tempering parameter; constructing a double-layer Bayesian network model, wherein the double-layer Bayesian network model comprises a first-layer Bayesian network model and a second Bayesian network model connected with the first-layer Bayesian network model; and inputting the target performance of the wheel and the current equipment feature vector into a double-layer Bayesian network model, wherein the double-layer Bayesian network model outputs multiple groups of prediction production parameters with high confidence. While multiple performance indexes such as hardness, static balance and residual stress are met, the number of physical tests is reduced, the process development period is shortened, and the production stability is guaranteed through uncertainty quantification and online self-learning.

This invention discloses a dynamic simulation device and a method for optimizing the molding quality of overflow forming of OLED display glass, belonging to the field of glass overflow forming process development technology. The device includes a support frame, an overflow pull-down mechanism, a simulated liquid heating furnace, a cooling mechanism, and a traction roller. The overflow pull-down mechanism is equipped with an overflow trough, which is provided with a near-end and far-end dual lifting mechanism, a far-end height adjustment component, and a material flow rate control valve. It can dynamically adjust the trough depth, far-end height, and inlet flow rate without stopping the machine for disassembly and assembly, completely replicating the entire process of overflow, cooling, and traction forming, and preparing a solid sheet. The method establishes a physical model based on the Reynolds number as a similarity criterion, and adjusts the inlet flow rate, trough depth, and far-end height in a closed loop according to the sheet thickness deviation to quickly correct the thickness uniformity. The parameters can be proportionally converted to the industrial prototype, improving process development efficiency, reducing costs, and compensating for the deviation between numerical simulation and actual production, providing reliable support for the optimization of OLED display glass overflow forming process.

The invention discloses a software engineering process development method based on big data, and relates to the technical field of software engineering development, the method comprises the following steps: constructing a unified data acquisition layer to capture multi-source heterogeneous development data in real time, and injecting the multi-source heterogeneous development data into a data lake; constructing a software engineering knowledge graph based on the domain ontology, and realizing semantic association of six types of core entities; calculating a process health degree index fusing the four-dimensional index through a sliding window; training a defect high-incidence area prediction model by using an LSTM neural network; and when the health degree is lower than a threshold value or the risk exceeds a limit, triggering an automatic decision engine to generate a regulation and control instruction and reversely injecting a tool chain to realize closed-loop optimization. According to the method, depth observability, quality risk early warning and flow adaptive regulation and control in the development process are realized, and the delivery punctuality rate, the defect prevention capability and the resource utilization efficiency are remarkably improved.

This application provides a method, apparatus, system, and medium for optimizing standard cell layouts. Based on the actual process flow and structural parameters of an initial standard cell layout, electrical simulations are performed. The simulation performance indicators of the initial standard cell layout are calculated based on the simulation results. When the simulation performance indicators match the preset performance indicators, a multinomial model of the layout dependency effect is established based on the fitting factor of the layout dependency effect model of the initial standard cell layout. The structural parameters of the initial standard cell layout are then corrected to obtain the corrected device structural parameters. A compact model is established, and a greedy algorithm is used to optimize the initial standard cell layout. By utilizing semiconductor process and device simulation tools to model based on the layout dependency effect, the cost of test fabrication is reduced. By generating a compact model and combining it with a greedy algorithm, optimized standard cell layouts are quickly generated, improving the efficiency of standard cell layout optimization, guiding the improvement of standard cell layout performance in the early stages of process development, and shortening the R&D cycle.

The invention discloses a tightening process development method, device and system suitable for a bolt group, and the method comprises the steps: calculating the pre-tightening force range of a bolt based on the material performances of the bolt and a connected part; taking the constraint condition that the final pre-tightening force obtained by each bolt in the bolt group under the action of the tightening torque corresponding to the initial pre-tightening force is within the pre-tightening force range, and continuously iterating the elastic interaction matrix according to the relationship among the target final pre-tightening force matrix, the elastic interaction matrix and the initial pre-tightening force matrix to obtain a final pre-tightening force matrix; and correcting the initial pre-tightening force of each bolt in the initial pre-tightening force matrix to obtain the final initial pre-tightening force of each bolt. According to the invention, the consistency of the connection pretightening force of each bolt in the bolt group can be ensured, and the bolt connection quality is improved.

The invention relates to a metal casting process optimization method, and relates to the technical field of metal casting process optimization, 3D design software is used for constructing a casting three-dimensional model and a mold model, MAGMASOFT software is used for setting process parameters, the process parameters comprise temperature parameters, time parameters and cooling conditions, corresponding thermophysical parameters of a built-in material library of the MAGMASOFT software are called, and the temperature parameters, the time parameters and the cooling conditions of the casting three-dimensional model and the mold model of the casting three-dimensional model are calculated. MAGMASOFT software is operated to carry out multi-physics field coupling simulation, TAG grid adaptive subdivision is adopted, potential defect types and areas are identified, a temperature field cloud picture and solidification animation are generated, the defect types comprise shrinkage cavities, shrinkage porosity, air holes and hot cracks, technological parameters are adjusted, simulation is carried out again, a DOE experimental design method and genetic algorithm iteration are adopted to carry out technological parameter adjustment, and the temperature field cloud picture and solidification animation are obtained. And finally, the required process conditions are met. According to the technical scheme, the casting simulation precision is improved, defects in actual production are reduced, and the production efficiency is improved by optimizing process parameters, reducing the production cost and shortening the process development period.

This invention discloses a method and system for predicting energy density in additive manufacturing based on spatiotemporal decoupling and hardware acceleration. The invention transforms continuous geometric space into a mesh through voxel discretization; constructs a total heat source tensor representing the cumulative energy distribution along the laser scanning path; performs spatiotemporal decoupling on the heat conduction model, transforming the four-dimensional transient problem into a three-dimensional static spatial integration problem; equivalences this integration problem to convolution operations, solving for the energy density field through the convolution of the total heat source tensor and the heat diffusion kernel tensor; utilizes GPU parallel acceleration for the convolution calculation; and finally, performs defect prediction and process optimization based on the energy density field. This invention avoids the time-consuming iterations of traditional transient simulations, achieving efficient and high-fidelity thermal process simulation at the part scale, significantly shortening the process development cycle.

The invention relates to the technical field of air conditioner accessories, and discloses a self-piercing riveting process optimization method and system.The method comprises the steps that firstly, a finite element model simulating the riveting forming process is constructed; secondly, a reference rivet is fixed, and a target riveting die enabling the joint forming quality to be optimal is screened out from the candidate riveting dies through simulation; then, on the basis of the target riveting die, riveting forming simulation and material object preparation are conducted on various candidate rivets, and corresponding simulation connector samples and actual connector samples are obtained; and finally, comprehensively analyzing the section morphology size of the simulation joint and the mechanical property data of the actual joint, screening out a target rivet with the optimal comprehensive performance, and outputting an optimized process parameter combination. According to the method, through the combination of simulation prediction and experimental verification and the step-by-step strategy of firstly optimizing the mold and then optimizing the rivet, accurate and efficient optimization of the process parameters is achieved, and the connection reliability of the base materials to be connected and the process development efficiency are remarkably improved.

The application belongs to the technical field of heat treatment, and provides a GCr15 bearing steel bearing ring heat treatment process parameter optimization method, which comprises the following steps: S1, obtaining the bearing type, geometric parameters, performance constraint conditions and target heat treatment equipment capability boundary of the GCr15 bearing steel bearing ring to be optimized; S2, establishing a collaborative process parameter set composed of independent process parameters and collaborative coupling parameters; S3, establishing a four-field full-coupling dynamic simulation model and verifying; S4, establishing a training data set; S5, establishing a double-level collaborative prediction model; S6, multi-objective collaborative optimization and three-level checking; S7, small-batch trial production and three-level triggering and double-dimension updating closed-loop iteration mechanism; S8, process database updating. Through the mapping relationship between the bearing geometric parameters and the collaborative process parameters, the application realizes the rapid migration and reuse of the heat treatment processes of different types of bearing rings, greatly shortens the process development cycle, and reduces the trial and error cost.

This invention discloses a method and system for macroscopic forming control in high-power single-mode laser welding, belonging to the field of laser welding quality monitoring and control technology. The method first establishes a complete quadratic polynomial mapping model between weld penetration depth and laser power, welding speed, and defocusing amount based on process experiments. Through time-scale analysis, a dimensionless number S is defined to characterize the relative magnitude of the evaporation and heat conduction effects at the top and bottom of the keyhole, establishing a keyhole stability criterion. Based on the threshold range of S and the penetration depth model, a process window for good macroscopic forming is defined. In actual welding, the S value is calculated according to process parameters, and combined with the process window, rapid judgment of weld forming quality is achieved. The dimensionless number S criterion provided by this invention quantitatively reflects keyhole stability from a physical mechanism perspective, avoiding the blindness of traditional empirical methods, and can efficiently achieve process parameter control, improving the forming quality and process development efficiency of thick plate laser welding.

This application provides a method and related equipment for verifying assembly code, belonging to the field of process development technology. It solves the problems in existing technologies where assembly code is unreadable, the compiler cannot verify it, leading to low efficiency in assembly code development, and difficulties in debugging and locating errors in the assembly code. The method includes: acquiring assembly code data, wherein the assembly code data includes assembly code and control instructions; simulating the execution of the assembly code based on the control instructions to obtain the execution result of the assembly code; converting the execution result into graphical data; and calling a graphical interface to display the graphical data.

This invention relates to the interdisciplinary field of artificial intelligence and welding process optimization, specifically disclosing a rapid optimization method for dissimilar metal welding process parameters based on transfer learning. This method aims to address the core problems of long optimization cycles and high costs in dissimilar metal welding due to scarce experimental data and reliance on trial and error in process development. Its core steps include: constructing a welding process knowledge base integrating multi-source data; generating a dynamic prediction model adapted to the target task using hierarchical transfer learning; establishing a digital twin integrating this model as a virtual optimization environment; applying multi-objective reinforcement learning in the twin for rapid optimization of process parameters; and finally achieving online dynamic control of the process strategy and system generalization across materials. This invention significantly reduces data dependence and achieves rapid and high-quality process optimization under small sample conditions.

This invention relates to the field of semiconductor device simulation and process development technology, and particularly to a calibration method and related apparatus for computer-aided design technology models (TCAD). The method includes constructing a TCAD device model and determining calibration parameters for the TCAD device model; training a random forest model based on an initial parameter set and labels of the initial parameter set to determine the effective parameter domain of the calibration parameters; and training a Gaussian process regression model within the effective parameter domain to determine the optimal parameter set of the calibration parameters. It can be seen that the proposed embodiment helps to solve the problems of low accuracy, low efficiency, slow convergence, and poor multi-objective optimization effects of traditional calibration methods, achieving rapid and high-precision calibration of TCAD models for subsequent research in areas such as radiation effects.

This invention proposes a motion control method for a PI coating machine controller based on Monte Carlo Markov chain sampling. A digital twin model of the coating process is constructed using a simulation platform. Model parameters are calibrated using Monte Carlo Markov chain sampling. The PI coating machine controller is then trained based on the calibrated digital twin model. After training, the optimal motion control method is obtained. The coating parameters of the PI coating machine are collected in real time and input into the optimal motion control method. The PI coating machine controller generates and executes real-time motion control commands. This method improves the accuracy and reliability of the digital twin model, providing a highly reliable foundation for PI coating machine controller training. It also reduces process development and control iteration costs, and enhances the model's adaptability and predictive capabilities.

The invention discloses a gray-scale photoetching ICP transfer modeling closed-loop method and system applied to the field of micro-nano optical master mold manufacturing and process development, and the scheme is developed around selection ratio modeling, three-sample development, outsourcing data return and version solidification. The system comprises an input module, a selection ratio modeling and parameter recommendation module, a three-sample planning and DOE management module, a metering and acceptance module, a process solidification and version management module and an output module. The method comprises the steps of PR / Si etching rate testing, selection ratio and reachable depth calculation, three-sample wafer DOE, measurement updating, full-size verification and formula curing output so as to improve contour retentivity, process reproducibility and female die mass production consistency.

The invention discloses an Agent construction and operation system and method based on the AI large model technology, and relates to the technical field of AI large models, and the Agent construction and operation system based on the AI large model technology comprises an interactive node configuration system and a visual process development system which are in a layered development system. The interactive node configuration system outputs a standardized component unit, and the visual process development system realizes the combined arrangement of a business process through a component reference mechanism; the interactive node configuration system and the visual process development system are provided with dynamic update links; the interactive node configuration system and the visual process development system adopt an interactive closed-loop design; based on the configured node process, automatically generating delivery forms adaptive to different types; the interactive node configuration system adapts to full-scene development requirements; the layered development system reduces the development cost; the complex process arrangement capability breaks through the traditional limitation; the system can automatically generate delivery forms adaptive to different requirements, and the efficiency is improved.

The invention relates to the technical field of manufacturing execution systems, in particular to a digital collaborative management and control method and system for connection of pharmaceutical process development and production, and the method comprises the following steps: collecting process development data of a process development system and production resource state data of a production execution system, and analyzing the process development data to extract process parameter design space and intermediate stability data. According to the method, the production scheduling optimization model is constructed, the target function of the minimum total completion time cost and the process parameter deviation cost is calculated, combined optimization of process parameter setting and production plan scheduling is achieved, and an instruction giving consideration to both quality stability and efficiency optimization is output. According to the instruction, collaborative management and control are carried out on pharmaceutical production equipment, it is guaranteed that pharmaceutical production strictly follows the process design range, the comprehensive utilization rate of the equipment and the compactness of the production rhythm are improved, the scrap risk caused by degradation of intermediates is reduced, and fusion of process science and production management is achieved.

The invention discloses a lightweight remote EOL detection system, which comprises detection terminal equipment, a task detection subsystem, a process tracking subsystem and a visual interaction module, the task detection subsystem sets and issues a detection task to the process tracking subsystem, the process tracking subsystem generates an executable detection instruction and then transmits the executable detection instruction to the detection terminal device, and the detection terminal device is connected with a vehicle machine system of a vehicle to be detected and executes the detection task. Monitoring the execution condition of the detection task executed by the detection terminal equipment in real time through a process tracking subsystem; and the visual interaction module is used for checking the execution condition of the detection task and vehicle data information detected by the detection terminal equipment in real time. According to the method, a'process development-task management-terminal execution-result feedback 'closed loop is constructed, production systems such as MES are seamlessly connected, and detection is deeply integrated into production; multi-dimensional abnormal early warning of order progress, equipment, communication and the like is covered, batch quality problems are reduced from the source, and the risk that unqualified products flow out is reduced.

This invention discloses a method and system for optimizing the cooling process of continuously cast billets with columnar crystal length as the control target, belonging to the field of continuous casting process control in iron and steel metallurgy. The method includes: S1. Establishing a quantitative relationship model for the target steel grade and the cross-section of the billet; S2. Determining the target columnar crystal length range based on product quality requirements and the quantitative relationship model; S3. Based on the principle of solidification heat transfer, calculating and determining the optimal cooling process parameter set required for the secondary cooling zone of continuous casting to reach the target columnar crystal length range through numerical simulation or empirical formulas; S4. Applying the optimal cooling process parameter set to production and verification. This invention uses the key intermediate metallurgical state variable of "columnar crystal length" as a direct control target, changing the traditional indirect control mode that uses surface temperature as the target. It achieves precise and scientific design from final quality requirements to process parameters, significantly improving the stability and consistency of the internal quality of the billet and greatly increasing process development efficiency.

The application discloses an intelligent manufacturing method for the positive temperature coefficient characteristic of a fast recovery diode. The method takes a partition physical information agent model as a first fidelity evaluation layer, takes process simulation as a second fidelity evaluation layer, and determines a process design parameter set through layered fidelity constraint Bayesian search; a P+ anode layer is formed through ion implantation and annealing, and an agent model is updated online; a Pareto optimal irradiation parameter is searched through deep Gaussian process multi-target Bayesian optimization, and after proton irradiation and electron beam irradiation, combined annealing is performed to stabilize two types of defects simultaneously under a single heat budget; the agent model is continuously updated in increments through a layered test strategy, process correction suggestions are output through statistical process control, and a self-evolution process knowledge base is constructed. The application significantly reduces the experimental cost of process development, and realizes continuous improvement of batch-to-batch consistency of the positive temperature coefficient characteristic.