Manufacturing process variation simulator

Abstract

The computer system receives one or more process parameters to be optimized in a multi-step manufacturing process. The computer system starts a process prediction model based on the one or more process parameters. The computer system simulates the multi-step manufacturing process using multiple sets of different settings until the one or more process parameters are optimized. From these multiple sets of settings, the computer system identifies a first set of settings that optimizes the one or more process parameters. The computer system instructs the station controller to apply this first set of settings to one or more stations.

Description

Technical Field

[0001] [Cross - Reference to Related Applications] This application claims priority to U.S. Provisional Patent Application No. 63 / 485,616, filed on February 17, 2023, the entire disclosure of which is hereby incorporated by reference.

[0002] [Technical Field] The present disclosure generally relates to the field of control and optimization of manufacturing processes, and more particularly, to systems and methods for using deep learning models to simulate process variations and predict future effects by changing process parameters.

Background Art

[0003] In the field of control and optimization of manufacturing processes, various parameters can affect the results of the process. Such parameters include, but are not limited to, the temperature, pressure, humidity of the environment, the properties of the raw materials used, and the settings of the devices related to the process. The parameters can change over time due to many factors such as changes in raw materials, external conditions, or inherent variations in the devices or control systems used.

Summary of the Invention

Means for Solving the Problems

[0004] As part of an embodiment, a manufacturing system is disclosed. This manufacturing system includes one or more stations, a station controller, and a computer system. Each station is configured to perform at least one step in a multi-step manufacturing process for producing a product. The station controller communicates with one or more stations. The station controller defines setpoints for each of the one or more stations. The computer system is configured to optimize process parameters of the multi-step manufacturing process. The computer system is configured to perform an operation. This operation includes receiving one or more process parameters to be optimized in the multi-step manufacturing process. This operation further includes starting a process prediction model according to the one or more process parameters to be optimized in the multi-step manufacturing process. The operation further includes simulating the multi-step manufacturing process using the process prediction model with multiple sets of different setpoints until one or more process parameters are optimized. The operation further includes identifying a first set of setpoints from multiple sets of different setpoints that achieves optimization of one or more process parameters. The operation further includes causing the station controller to apply the first set of setpoints to one or more stations.

[0005] As part of an embodiment, a method for optimizing process parameters in a multi-step manufacturing process is disclosed. A computer system receives one or more process parameters to be optimized in a multi-step manufacturing process. This multi-step manufacturing process is executed in a manufacturing system having one or more stations and a station controller. Each station is configured to execute a step in the multi-step manufacturing process. The station controller communicates with one or more stations. The station controller defines setpoints for each of the one or more stations. The computer system starts a process prediction model according to the one or more process parameters to be optimized in the multi-step manufacturing process. The computer system uses the process prediction model to simulate the multi-step manufacturing process with multiple sets of different setpoints until one or more process parameters are optimized. From the multiple sets of different setpoints, the computer system identifies a first set of setpoints that achieves optimization of one or more process parameters. The computer system causes the station controller to apply the first set of setpoints to one or more stations.

[0006] As part of an embodiment, a non-temporary computer-readable medium is disclosed. The non-temporary computer-readable medium includes one or more stored instruction sequences that, when executed by a processor, cause a computer system to perform an operation. This operation includes the computer system receiving one or more process parameters to be optimized in a multi-step manufacturing process. The multi-step manufacturing process is performed in a manufacturing system having one or more stations and a station controller. Each station is configured to perform a step in the multi-step manufacturing process. The station controller communicates with one or more stations. The station controller defines setpoints for each of the one or more stations. The operation further includes the computer system initiating a process prediction model according to one or more process parameters to be optimized in the multi-step manufacturing process. The operation further includes the computer system simulating the multi-step manufacturing process using the process prediction model with multiple sets of different setpoints until one or more process parameters are optimized. The operation further includes the computer system identifying a first set of setpoints from multiple sets of different setpoints that achieves the optimization of one or more process parameters. The operation further includes the computer system instructing the station controller to apply the first set of setpoints to one or more stations.

[0007] [Brief description of the drawing] The accompanying drawings are incorporated herein as part of this specification and illustrate the disclosure, and together with this specification, further illustrate the principles of this disclosure and enable those skilled in the art to manufacture and use the embodiments described herein. [Brief explanation of the drawing]

[0008] [Figure 1] This is a block diagram showing a manufacturing environment relating to an embodiment of the present disclosure. [Figure 2]This flowchart shows a method for generating a predictive model of a manufacturing process and optimizing process parameters according to an embodiment of the present disclosure. [Figure 3] This flowchart shows a method for optimizing a manufacturing process using a trained predictive model according to an embodiment of the present disclosure. [Figure 4A] This is a block diagram of a computer device according to an embodiment of the present disclosure. [Figure 4B] This is a block diagram of a computer device according to an embodiment of the present disclosure. [Modes for carrying out the invention]

[0009] The features of this disclosure become even clearer when viewed in conjunction with the following detailed description, which includes drawings in which similar reference numerals indicate corresponding elements. In the drawings, similar reference numerals generally indicate elements that are identical, functionally similar, and / or structurally similar. Generally, the leftmost digit of a reference numeral identifies the drawing in which the reference numeral first appeared. Unless otherwise stated, the drawings in this disclosure should not be considered to scale.

[0010] [Detailed explanation] Manufacturing is complex and involves various process stations (also simply called "stations") that process raw materials to produce the final product (hereinafter referred to as the "final output"). Each process station, except for the final one, receives input to be processed, outputs an intermediate output, and sends this intermediate output to one or more subsequent (downstream) process stations for further processing. The final process station receives input to be processed and outputs the final output.

[0011] Each process station may include one or more tools / devices that perform a series of process steps on the raw materials received (which may be from the first or subsequent station in the manufacturing process) and / or the output received from the previous station (which may be from a subsequent station in the manufacturing process). Examples of process stations include, but are not limited to, conveyor belts, injection molding presses, cutting machines, stamping machines, extruders, CNC (computer numerical control) mills, grinders, assembly stations, 3D printers, robotic devices, and quality control / verification stations. Examples of process steps include transporting output from one place to another (such as by a conveyor belt), supplying material to an extruder, melting material, injecting it through a mold opening, and allowing it to cool and solidify to form the shape of the mold opening (such as by an injection molding press), cutting material to a specific shape or length (such as by a cutting machine), and pressing material to a specific shape (such as by a stamping machine).

[0012] In a manufacturing process, process stations can be configured in parallel or in series. When operating in parallel, a single process station can send its intermediate output to more than one station (e.g., one to N stations), and can also receive and combine intermediate outputs from more than one to N stations. Furthermore, a single process station can perform the same or different process steps on the received raw materials or intermediate outputs, either continuously or discontinuously, during a single iteration of the manufacturing process.

[0013] The operation of each process station can be managed by one or more process controllers. In some embodiments, each process station is provided with one or more process controllers (referred to as "station controllers") programmed to control the operation of the process station (the algorithm for this programming is referred to as the "control algorithm"). In some embodiments, one process controller can also be configured to control the operation of two or more process stations. An example of a factory controller is a programmable logic control unit (PLC). A PLC is programmable to operate manufacturing processes and systems. PLCs and other controllers receive information from connected sensors and input devices, process the data based on pre-programmed parameters and commands, and generate outputs (e.g., control signals for controlling the associated process station).

[0014] Depending on the operator and control algorithm, a station controller may be assigned a station controller setpoint (also called a "setpoint," "controller setpoint," or CSP) that represents a desired single value or range of values ​​for each control value. Measurable values ​​during the operation of the station's equipment or process can be classified as either control values ​​or station values. Here, values ​​controlled by the station controller are classified as control values, while other measured values ​​are classified as station values. Examples of control values ​​and / or station values ​​include, but are not limited to, speed, temperature, pressure, vacuum level, rotational speed, current, voltage, power, viscosity, materials / supplies used at the station, throughput rate, downtime, toxic gases, and the type and sequence of processes performed at the station. While the examples remain the same, whether a measured value is separated into a control value or classified as a station value depends on the station and whether the measured value is controlled by the station controller or is merely a by-product of the station's operation. During the manufacturing process, control values ​​are assigned to the station controller, but station values ​​are not.

[0015] The control algorithm may include instructions for monitoring a control value, comparing it to a corresponding setpoint, and determining what action to take if the control value is not equal to (or outside the specified range of) the corresponding station controller setpoint. For example, if the current measured station temperature is lower than the setpoint, the station controller can send a signal to increase the heat source temperature of that station until the current station temperature equals the setpoint. Conventional process controllers used to control stations in a manufacturing process are limited because they follow static algorithms (e.g., on / off control, PI control, PID control, read-lag control) to instruct what action to take when the control value deviates from the setpoint.

[0016] Each process station may include or be linked to one or more sensors. Such sensors may be physical or virtual sensors that exist independently of the operation of the deep learning processor during the manufacturing process, as well as new sensors added to perform additional measurements required by the deep learning processor. Sensors can be used to measure values ​​generated in the manufacturing process, such as station values, control values, intermediate and final output values. Examples of sensors include, but are not limited to, rotary encoders for detecting position and velocity, sensors for detecting proximity, pressure, temperature, level, flow rate, current, and voltage, and limit switches for detecting states such as presence or absence, or limits of a stroke. In this specification, a sensor includes both a detection device and a signal processing device. For example, a detection device responds to station values ​​or control values, and a signal processing device converts that response into a signal that can be used and interpreted by the deep learning processor or station controller. Examples of temperature-sensitive sensors include RTDs (temperature-resistant detectors), thermocouples, and platinum resistance probes. Strain gauge sensors respond particularly to variations in pressure, vacuum, weight, and distance. Proximity sensors react to objects when they are within a predetermined distance of each other or a specific component. In all of the above examples, the reaction must be converted into a signal usable by a station controller or deep learning processor. Often, the sensor's signal processing function generates a digital signal that the station controller interprets. Signal processing devices also generate analog signals and TTL (transistor-transistor logic) signals, in particular. Virtual sensors, also known as soft sensors, smart sensors, or estimators, include system models that receive and process data from physical sensors.

[0017] In this specification, process values ​​mean station values ​​or control values ​​aggregated or averaged across a series of stations (or a subset of stations) that constitute a part of a manufacturing process. Process values ​​include, for example, total throughput time, all sources used, average temperature, average speed, etc.

[0018] In addition to station values ​​and process values, various characteristics of the process station's product output (i.e., intermediate and final output), such as temperature, weight, product dimensions, mechanical, chemical, optical, and / or electrical properties, the number of structural defects, and the presence or absence of defect types, can be measured. These various measurable characteristics are collectively referred to as "intermediate output values" and "final output values." Intermediate / final output values ​​can reflect a single measured characteristic of the intermediate / final output, or a comprehensive score based on a set of specific characteristics related to the intermediate / final output, with importance adjusted according to a predetermined formula.

[0019] Mechanical properties include hardness, compressibility, tackiness, density, and weight. Optical properties include absorbance, reflectance, transmittance, and refractive index. Electrical properties include electrical resistivity and electrical conductivity. Chemical properties include enthalpy of formation, toxicity, chemical stability under specified environmental conditions, flammability, preferred oxidation state, pH (acidity / alkalinity), chemical composition, boiling point, and vapor point. The disclosed mechanical, optical, chemical, and electrical properties are merely examples and not intended to be limiting.

[0020] All processes typically involve inherent variability, even under static conditions. Each process station can vary within limits. Such limits can arise naturally as a result of normal control algorithms, such as proportional-integral-derivative (PID) controllers. Each process may also vary due to conditions outside the control of the process station. For example, raw materials can change over time and affect the process. Variations in raw materials can generally be limited by defining limits on the properties of the raw materials. For example, plastics can be specified to have upper and lower viscosity limits within the range of 47-53 cP. The effects of variations in raw material properties can be used to train artificial intelligence models used in process simulators. In another example, external conditions such as factory temperature and humidity may vary. Such inputs to artificial intelligence models can be considered universal inputs.

[0021] A process parameter can broadly refer to various characteristics and values that can be measured and observed in a manufacturing process. For example, a process parameter may be a characteristic of an intermediate output or a final output generated in a manufacturing process.

[0022] In some embodiments, the inherent variability of a process can be characterized by the limits of variability as described above. For example, a process parameter can be indicated by the average and standard deviation of the parameter over time. In the expression of statistical process control, the performance of a process can be expressed as the limits of parameter specification / process variability (e.g., process capability Cp, Cpk). In some embodiments, the inherent variability of a process can also be characterized by the frequency of variability. For example, the temperature of a press under control may vary by ±2°C in 2 minutes. In another example, the raw material supplied to the press may vary by + / 3 cP in 1 hour. In yet another example, the temperature / humidity in the factory may vary by + / 5°C, ±10 RH in 12 hours. In yet another example, the "total" variability can be the cumulative sum of the variability ranges of normal variability and time-dependent variability.

[0023] The future timeline of process performance is predictable, and the probability of its occurrence can be determined by knowing or obtaining the range and frequency of variability for all process conditions. In some embodiments, by knowing or obtaining the future timeline and probability of occurrence, the process parameters can be changed to reduce the variability of process performance. To evaluate the future effects obtained by changing the process parameters, "new" process parameters can be supplied to a process emulator, which can then change the process characteristics determined by the process simulator.

[0024] This disclosure relates to systems and methods for evaluating the future effects of changing process parameters in a manufacturing system. In some embodiments, the system and method include the use of a process prediction model and a plurality of process stations, each having inherent variability. This variability may be due to a variety of factors, including, but is not limited to, external conditions, raw material properties, and control algorithms. The system and method can be designed to take such variability into account, thereby enabling more accurate and reliable prediction of future process performance.

[0025] In some embodiments, the system may include a station controller configured to communicate bidirectionally with process stations and process prediction models. The station controller sends and receives parameters to and from each process station via bidirectional communication, and the deep learning processor can exchange parameters bidirectionally with each process station. This communication allows for a more dynamic and responsive system that can adjust to real-time changes in process parameters.

[0026] Overall, the systems and methods of this disclosure can provide a more accurate and reliable method for predicting future process performance in a manufacturing system. By considering inherent process variations and using deep learning techniques, the systems and methods of this disclosure can provide a more comprehensive and detailed manufacturing process model, thereby improving process performance and efficiency.

[0027] Figure 1 is a block diagram showing a manufacturing system 100 according to an embodiment of the present disclosure. The manufacturing system 100 may include a station controller 110. In some embodiments, the station controller 110 may be represented by a PLC (Programmable Logic Controller). In some embodiments, the station controller 110 can be configured to control one or more process stations, or one or more control valves associated with one or more process stations. As shown in the figure, the station controller 110 can communicate bidirectionally with a plurality of process stations. For example, the station controller 110 can send and receive parameters bidirectionally with process station 130 via communication path 116, the station controller 110 can send and receive parameters bidirectionally with process station 140 via communication path 114, the station controller 110 can send and receive parameters bidirectionally with process station 150 via communication path 112, the station controller 110 can send and receive parameters bidirectionally with process station 160 via communication path 116, and the station controller 110 can send and receive parameters bidirectionally with process station 170 via communication path 118. Although the figure shows a single station controller 110, those skilled in the art will understand that multiple station controllers can be used.

[0028] As shown in the figure, process station 130 can receive input material 131. Input material 131 broadly represents raw materials that can be used in a manufacturing process. Examples of raw materials include plastics, metals, rubber, etc. Process station 130 can be configured to perform manufacturing process steps on input material 131. As an output, process station 130 can produce output material 132. Output material 132 can then be provided as input to process station 170.

[0029] The process station 140 can receive input material 141. Input material 141 broadly represents raw materials that can be used in a manufacturing process. Examples of raw materials include plastics, metals, rubber, etc. The process station 140 can be configured to perform manufacturing process steps on the input material 141. As an output, the process station 140 can produce output material 142. The output material 142 can then be provided as input to the process station 160.

[0030] The process station 150 can receive input material 151. Input material 151 broadly represents raw materials that can be used in a manufacturing process. Examples of raw materials include plastics, metals, rubber, etc. The process station 150 can be configured to perform manufacturing process steps on the input material 151. As an output, the process station 150 can produce output material 152. The output material 152 can then be provided as input to the process station 170.

[0031] Process station 160 can receive output material 142 as input from process station 140. Process station 160 can be configured to perform manufacturing process steps on output material 142. As output, process station 160 can produce output material 162. Output material 162 can then be provided as input to process station 170.

[0032] The process station 170 can be configured to perform manufacturing process steps on output material 132, output material 162, and output material 152. As an output, the process station 170 can produce output material 172, which represents the final output of the manufacturing process.

[0033] As those skilled in the art will understand, raw materials, by their very nature, exhibit variations in their physical properties. However, such variations are not limited to the properties of the raw materials themselves, but also depend on the environmental conditions of the manufacturing system 100. Due to these variations, fluctuations in the goods being manufactured accumulate during the manufacturing process, resulting in highly volatile output from the manufacturing system 100. Taking this into consideration, the manufacturing system 100 may utilize a computer system 120, including a process prediction model 125, to optimize process parameters and reduce or minimize variations in the manufacturing process. In some embodiments, the process prediction model 125 is represented by a deep learning model.

[0034] As shown in the figure, the station controller 110 can also communicate bidirectionally with the computer system 120. For example, the station controller 110 can be configured to exchange parameters bidirectionally with the computer system 120 via the path 115. The computer system 120 can also be configured to analyze the output from any of the process stations 130, 140, 150, 160, and 170. The process prediction model 175 is trained with the outputs generated from the process stations 130, 140, 150, 160, and 170. The process prediction model 175 can be run considering the variations of all process stations. In some embodiments, if the optimized parameters meet a predetermined confidence threshold, the process prediction model 175 can be further trained using those optimized parameters.

[0035] In some embodiments, the computer system 120 may collect station parameters for each of the multiple process stations 130-170 via one or more paths 122, 124, 126, 127, 128. Generally, the station parameters may include at least one of setpoints, control values, intermediate output values, and other station values. In some embodiments, process variability for each of the multiple process stations 130-170 can be characterized by the mean and standard deviation of the station parameters over time. In some embodiments, process variability for each of the multiple process stations 130-170 can be further characterized by the frequency of variation of the station parameters.

[0036] Figure 2 is a flowchart of a method 200 according to an embodiment of the present disclosure, which generates a process prediction model to optimize process parameters and reduce variability in the manufacturing process. Method 200 can begin with step 202.

[0037] In step 202, the computer system 120 can receive station parameters for the first station of the manufacturing system. For example, the station controller 110 can transmit the station parameters of the first station to the computer system 120. These station parameters can broadly be settings that the station controller 110 has set for the first station to perform the first step of the manufacturing process. In some embodiments, the computer system 120 can also receive information relating to variations in the input material to the first station.

[0038] In step 204, the computer system 120 can receive the intermediate output generated by the first station. For example, the first station can perform the first step of the manufacturing process on the raw materials according to the setpoints communicated by the station controller 110. The first station communicates the intermediate output to the computer system 120. In some embodiments, the intermediate output may include an evaluation of the variation in the intermediate output. The process at the first station may continue until the process variation of the process station can be fully characterized. In some embodiments, fully characterizing the process variation may include analyzing the parameters under control and the corresponding final output and intermediate output values.

[0039] In step 206, the computer system 120 can determine whether the received output is the final output of the manufacturing process. In other words, the computer system 120 can determine whether the received output is from the final stage of the manufacturing process. If the computer system 120 determines in step 206 that the output is an intermediate output and not the final output, method 200 proceeds to step 208. In step 208, station parameters for the next process station may be received. In that case, method 200 returns from step 208 to step 204 for further processing.

[0040] However, if the computer system 120 determines in step 206 that the received output is the final output, method 200 can proceed to step 210. In step 210, the computer system 120 can generate a trained process prediction model based on the station parameters and corresponding output values ​​collected in steps 202 to 208. For example, to train a process prediction model to optimize the process parameters of a manufacturing process, the computer system 120 may generate a training dataset containing multiple examples of station parameters, corresponding output values, and variation information.

[0041] In step 212, the computer system 120 can output a well-trained process prediction model that can be deployed to the manufacturing system 100.

[0042] Figure 3 is a flowchart of a method 300 according to an embodiment of the present disclosure, which optimizes process parameters to reduce or minimize variability in a manufacturing system. Method 300 can begin with step 302.

[0043] In step 302, the computer system 120 receives the process parameters to be optimized. In specific embodiments, such as additive manufacturing, the operator can specify to the computer system 120 that the process parameters to be optimized are the tensile strength of each layer in additive manufacturing, while also specifying that the amount of material used be minimized. In some embodiments, the operator or user can set a hierarchy of process parameters to be optimized. Continuing the above example, the operator or user may define the hierarchy as first optimizing the tensile strength, and then optimizing the amount of material used. As those skilled in the art will understand, this hierarchy can be rearranged. For example, the operator may instruct the computer system 120 to first optimize the amount of material used, and then optimize the tensile strength.

[0044] In step 304, the computer system 120 can start a process prediction model 125 based on the received process parameters. For example, the computer system 120 may instruct the process prediction model 125 to optimize the width and thickness of each layer of additive manufacturing to increase tensile strength while minimizing the amount of material used, according to the hierarchy set by the operator.

[0045] In step 306, the computer system 120 can run the process prediction model 125 to simulate a multi-step manufacturing process and optimize one or more process parameters. For example, the computer system 120 may repeatedly simulate the multi-step manufacturing process using different sets of station settings until it can identify a set of settings that optimizes one or more process parameters. For example, the computer system 120 may run the process prediction model 125 using a first set of process station settings to simulate the manufacturing process. Based on this first set of process station settings, the process prediction model 125 can generate and evaluate intermediate outputs generated in the simulation process using the first set of process station settings. The computer system 120 may continuously adjust the settings input to the process prediction model 125 until the process prediction model 125 determines that one or more process parameters have been optimized.

[0046] In some embodiments, if the variation in one or more process parameters is reduced or minimized, those one or more process parameters may be considered optimized.

[0047] In step 308, the computer system 120 can identify a first set of station setting values ​​that resulted in one or more optimized process parameters. For example, the computer system 120 may identify one or more setting values ​​used by the process prediction model 125 that resulted in an optimal balance between tensile strength and the amount of material used.

[0048] In step 310, the computer system 120 can instruct the station controller 110 to apply the process station settings to the actual process station.

[0049] Figure 4A shows the system bus architecture of a computer system 400 according to an embodiment of the present disclosure. System 400 may represent at least a station controller 110 or a computer system 120. One or more components of system 400 can communicate with each other using a bus 405. System 400 may include a processing unit (CPU or processor) 410 and a system bus 405 that connects various system components, including read-only memory (ROM) 420 and random access memory (RAM) 425, to the processor 410.

[0050] System 400 may include a high-speed memory cache that is directly connected to, adjacent to, or integrated as part of the processor 410. System 400 can copy data from memory 415 and / or storage device 430 to cache 412 so that the processor 410 can quickly access the data. In this way, cache 412 can improve performance so that the processor 410 is not delayed by waiting for data. Such modules can control or be configured to control the processor 410 to perform various operations. Other system memories 415 can be used in the same way. Memory 415 may include multiple different types of memory with different performance characteristics. The processor 410 may include various general-purpose processors and hardware modules or software modules, such as service 1 (432), service 2 (434), and service 3 (436), which are configured to control the processor 410 and stored in storage device 430, as well as dedicated processors in which software instructions are incorporated into the actual processor design. The Processor 410 can essentially be a completely self-contained computer system incorporating multiple cores or processors, buses, memory controllers, caches, etc. The multi-core processor can be symmetrical or asymmetrical.

[0051] To enable the user to interact with the computer system 400, the input device 445 can represent several input mechanisms, such as a microphone for voice, a touch screen for gesture and graphic input, a keyboard, a mouse, motion input, and voice. The output device 435 can also be one or more of the many output mechanisms known to those skilled in the art. In some cases, a multimodal system can be used to allow the user to provide many types of input for interacting with the computer system 400. The communication interface 440 can typically adjust and manage user input and system output. Since there is no restriction that operations must be performed with a specific hardware configuration, this basic feature can be easily replaced with improved hardware or firmware configurations as they are developed.

[0052] The storage device 430 can be non-volatile memory and can be a hard disk or other form of computer-readable media capable of storing computer-accessible data, such as a magnetic cassette, flash memory card, solid memory device, digital versatile disk, cartridge, random access memory (RAM) 425, read-only memory (ROM) 420, and hybrids thereof.

[0053] The storage device 430 may include services 432, 434, and 436 for controlling the processor 410. Other hardware and software modules are also possible. The storage device 430 may be connected to the system bus 405. According to certain features, a hardware module that performs a particular function may include software components stored on a computer-readable medium in association with the necessary hardware components, such as the processor 410, the bus 405, and the output device 435 (e.g., a display), in order to perform that function.

[0054] Figure 4B shows a computer system 450 having a chipset architecture that may represent at least the station controller 110 or the computer system 120. The computer system 450 is an example of computer hardware, software, and firmware that can be used to perform the technology disclosed herein. The system 450 may include a processor 455 that represents several physically and / or logically distinct resources capable of performing software, firmware, and hardware configured to perform specified calculations. The processor 455 may communicate with a chipset 460 that can control the input and output of the processor 455.

[0055] In this example, the chipset 460 can output information to an output 465 such as a display, and can read and write information to a storage device 470 including magnetic media, solid media, etc. The chipset 460 can also read and write data to a storage device 475 (e.g., RAM). A bridge 480 may be provided for communication with the chipset 460, which interacts with various user interface components 485. Such user interface components 485 may include a keyboard, microphone, contact detection and processing circuits, pointing devices such as a mouse, etc. In general, the input to the system 450 may come from any of various sources, both machine-generated and / or human-generated.

[0056] The chipset 460 can also communicate with one or more communication interfaces 490, which may have different physical interfaces. Such communication interfaces may include interfaces for wired / wireless local area networks, broadband wireless networks, and personal area networks. Applications of the methods for generating, displaying, and using the GUI disclosed herein may include receiving sequential datasets via physical interfaces or sequential datasets generated by the machine itself, via a processor 455 that analyzes data stored in storage device 470 or storage device 475. Furthermore, the machine may receive input from a user via a user interface component 485 and interpret this input using the processor 455 to perform appropriate functions such as browsing.

[0057] It should be noted that the illustrated systems 400 and 450 may have more than one processor 410, or they may be part of a group or cluster of computing devices connected to each other by a network to provide higher processing capabilities.

[0058] The above relates to embodiments described herein, but other embodiments are conceivable without departing from the basic scope. For example, the features of this disclosure can be implemented in hardware, software, or a combination of hardware and software. One embodiment described herein can be implemented as a program product for use with a computer system. The program of the program product defines the function of the embodiment (including the method described herein) and may be contained in various computer-readable storage media. Exemplary computer-readable storage media include, but are not limited to, (i) non-writable storage media for permanently storing information (e.g., read-only memory (ROM) devices in a computer, such as CD-ROM disks readable by a CD-ROM drive, flash memory, ROM chips, or various solid-state non-volatile memories), and (ii) writable recording media for storing erasable and rewritable information (e.g., floppy disks in disk drives or hard disk drives, or various solid-state random-access memories). Such computer-readable storage media constitute an embodiment of this disclosure if it stores computer-readable instructions that direct the function of the disclosed embodiment.

[0059] Those skilled in the art will understand that the above examples are illustrative and not limiting. All substitutions, enhancements, equivalents, and improvements thereto will be apparent to those skilled in the art upon reading the specification and examining the drawings, and are intended to be included in the true spirit and scope of this disclosure. Accordingly, the following claims are intended to include all modifications, substitutions, and equivalents that fall within the true spirit and scope of this teaching.

Claims

1. One or more stations, each configured to perform at least one step in a multi-step manufacturing process for producing a product, A station controller that communicates with one or more of the aforementioned stations and defines setting values ​​for each of them, A computer system configured to optimize the process parameters of the aforementioned multi-step manufacturing process and to perform operations. The operation is provided, The process of receiving one or more process parameters to be optimized in the aforementioned multi-step manufacturing process, The process prediction model is started according to one or more process parameters to be optimized in the aforementioned multi-step manufacturing process, Using the process prediction model, the manufacturing process of the multiple steps is simulated using multiple sets of different settings until one or more process parameters are optimized. From the plurality of sets of different setting values, identify a first set of setting values ​​that optimizes one or more process parameters, The station controller is instructed to apply the first set of settings to one or more stations. A manufacturing system that includes the above.

2. Using the process prediction model, the manufacturing process of the multiple steps is simulated using multiple sets of different settings until one or more process parameters are optimized. The first simulation is performed using the aforementioned multiple sets of different settings to achieve the first optimization process parameters, The second simulation is performed using the aforementioned multiple sets of different settings to achieve the second optimization process parameters. A manufacturing system according to claim 1, including the above.

3. The manufacturing system according to claim 2, wherein the operator defines a hierarchy between the first optimization process parameter and the second optimization process parameter.

4. The manufacturing system according to claim 1, wherein the manufacturing process of the multiple steps is additive manufacturing.

5. The manufacturing system according to claim 4, wherein one or more process parameters to be optimized in the multi-step manufacturing process are the width and thickness of each layer of the additive manufacturing process.

6. The manufacturing system according to claim 1, wherein the process prediction model is trained by the actual output generated by the manufacturing system in the multi-step manufacturing process.

7. The manufacturing system according to claim 1, wherein the variation of each of the one or more stations is characterized by the mean and standard deviation over time of the station parameters.

8. A method for optimizing process parameters in a multi-step manufacturing process, A computer system receives one or more process parameters to be optimized in a multi-step manufacturing process performed in a manufacturing system having one or more stations, wherein each station is configured to perform a step in the multi-step manufacturing process, and a station controller communicates with the one or more stations and receives one or more process parameters to define a set value for each of the one or more stations. The computer system initiates a process prediction model according to one or more process parameters to be optimized in the multi-step manufacturing process, The computer system performs a process prediction model to simulate the manufacturing process of the multiple steps using multiple sets of different settings until one or more process parameters are optimized. The computer system performs the steps of identifying a first set of setting values ​​from the plurality of sets of different setting values ​​that optimizes one or more process parameters, The computer system provides a step of causing the station controller to apply the first set of setting values ​​to one or more stations. A method for optimizing process parameters, including [specific parameters].

9. The computer system simulates the manufacturing process of the multiple steps using a process prediction model, using multiple sets of different settings until one or more process parameters are optimized. A step of performing a first simulation using the aforementioned multiple sets of different settings to achieve a first optimization process parameter, The process of performing a second simulation using the aforementioned multiple sets of different settings to achieve the second optimization process parameters The method according to claim 8, including the method described in claim 8.

10. The method according to claim 9, wherein the operator defines a hierarchy between the first optimization process parameter and the second optimization process parameter.

11. The method according to claim 8, wherein the manufacturing process of the multiple steps is additive manufacturing.

12. The method according to claim 11, wherein the one or more process parameters to be optimized in the multi-step manufacturing process are the width and thickness of each layer of the additive manufacturing.

13. The method according to claim 8, wherein the process prediction model is trained by the actual output generated by the manufacturing system in the multi-step manufacturing process.

14. The method according to claim 8, wherein the variation of each of the one or more stations is characterized by the mean and standard deviation over time of the station parameters.

15. A non-temporary computer-readable medium containing one or more stored instruction sequences that, when executed by a processor, cause a computer system to perform an operation, wherein the operation is: The process of receiving one or more process parameters to be optimized in the multi-step manufacturing process by a computer system, wherein the multi-step manufacturing process is performed in a manufacturing system having one or more stations configured to execute the steps in the multi-step manufacturing process, and a station controller that communicates with the one or more stations and defines set values ​​for each of them, The computer system starts a process prediction model according to one or more process parameters to be optimized in the multi-step manufacturing process, The computer system uses the process prediction model to simulate the multi-step manufacturing process using multiple sets of different settings until one or more process parameters are optimized. The computer system identifies a first set of setting values ​​that optimizes one or more process parameters from the plurality of sets of different setting values, The computer system causes the station controller to apply the first set of setting values ​​to one or more stations. Non-temporary computer-readable media, including [specific examples of such media].

16. Using the process prediction model, the manufacturing process of the multiple steps is simulated using multiple sets of different settings until one or more process parameters are optimized. The first simulation is performed using the aforementioned multiple sets of different settings to achieve the first optimization process parameters, The second simulation is performed using the aforementioned multiple sets of different settings to achieve the second optimization process parameters. A non-temporary computer-readable medium according to claim 15, including the following:

17. The non-temporary computer-readable medium according to claim 16, wherein the operator defines a hierarchy between the first optimization process parameter and the second optimization process parameter.

18. The non-temporary computer-readable medium according to claim 15, wherein the manufacturing process of the multiple steps is additive manufacturing.

19. The non-temporary computer-readable medium according to claim 18, wherein one or more process parameters to be optimized in the multi-step manufacturing process are the width and thickness of each layer of the additive manufacturing.

20. The non-temporary computer-readable medium according to claim 15, wherein the process prediction model is trained by the actual output generated by the manufacturing system in the multi-step manufacturing process.

Images