Material self-evolution control method, system and terminal for intelligent laboratory
By combining process management, intelligent control, and data acquisition modules in the preparation of quantum materials, the problems of low efficiency of manual operation and lack of data closure in automated systems in existing technologies have been solved, thereby achieving automation and improved stability in the preparation of quantum materials.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-11-25
- Publication Date
- 2026-04-07
AI Technical Summary
Existing technologies for quantum material preparation suffer from problems such as low efficiency and poor repeatability of manual operations, as well as a lack of closed-loop data management between processes in automated systems, leading to unstable preparation and low efficiency.
The experimental task is decomposed into sub-tasks using a process management module. Combined with an intelligent control module, control commands are generated through a process parameter prediction unit and a model prediction control unit. These commands are then sent to the experimental equipment through a decision execution module. The experimental data is monitored by a data acquisition module. If the target is not met, the process parameters and control strategies are iteratively updated until the target is met.
The process of quantum material preparation has been automated, stabilized, and made more efficient. Experimental results and material properties have been optimized by adaptively adjusting process parameters and control strategies.
Smart Images
Figure CN121806569A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of artificial intelligence technology, and in particular to a material self-evolution control method, system, and terminal for intelligent laboratories. Background Technology
[0002] Quantum materials, as a class of functional materials with unique physicochemical properties, encompass several important branches, including transition metal dichalcogenides, quantum spin systems, and topological insulators. The preparation process of quantum materials typically involves a series of complex, multi-step operations such as weighing, mixing, pressing, and sintering, each requiring extremely precise control of process parameters. Currently, existing technologies in the field of quantum material preparation mainly fall into two categories: one is the traditional manual operation method, relying on researchers to manually complete each step and adjust parameters; the other is an automated system at the equipment level, which mechanizes individual operational steps by introducing specialized automated equipment such as automatic weighing machines and temperature-controlled furnaces.
[0003] However, both of these existing technologies have significant drawbacks, making it difficult to meet the demand for efficient and stable preparation of quantum materials. Traditional manual methods are affected by factors such as human error and fatigue, resulting in low experimental efficiency and poor repeatability, leading to significant differences in product performance between different batches and under different experimental conditions. Existing automated systems are limited to the independent automated operation of individual devices, failing to achieve systematic integration of the entire preparation process. They lack data interaction and closed-loop management capabilities between processes and cannot dynamically optimize subsequent preparation decisions based on experimental data.
[0004] Therefore, existing technologies still need improvement and development. Summary of the Invention
[0005] The technical problem to be solved by this invention is to provide a material self-evolution control method, system and terminal for smart laboratories, addressing the above-mentioned deficiencies of existing technologies. This aims to solve the problem that existing quantum material preparation technologies are limited to equipment-level automation, lack data closure between processes, and fail to achieve systematic integration of the entire preparation process.
[0006] The technical solution adopted by this invention to solve the problem is as follows: In a first aspect, embodiments of the present invention provide a material self-evolution control method for intelligent laboratories, the method comprising: The experimental task is broken down into several sub-tasks through the process management module; the experimental task includes the material preparation task. The intelligent control module uses a process parameter prediction unit to predict the combination of process parameters based on the performance of the target product, and a model prediction control unit uses a model prediction control unit to generate at least one control instruction corresponding to the sub-task based on the combination of process parameters and the current system state. The decision execution module receives the control commands and sends them to the experimental equipment. The experimental equipment is monitored through the data acquisition module to obtain the experimental data corresponding to the experimental task; If the experimental data does not meet the standard, the strategies of the process parameter prediction unit and the model prediction control unit are iteratively updated based on the experimental data and the multi-objective optimization model to obtain new combinations of process parameters and control commands until the experimental data meets the standard.
[0007] In one embodiment, the process management module is further configured to: determine the priority of each subtask and the scheduling order of each functional module; each functional module includes: the data acquisition module, the intelligent control module, the feedback optimization module, and the decision execution module.
[0008] In one implementation, the process management module is further configured to: manage multiple experimental tasks in parallel.
[0009] In one embodiment, the process parameter prediction unit is specifically used for: By learning the correlation between the combination of process parameters and product performance in advance, the combination of process parameters can be predicted based on the performance of the target product.
[0010] In one implementation, the model prediction control unit is specifically used for: The control command is generated based on the combination of process parameters and the current system state using a sliding time window algorithm.
[0011] In one embodiment, the experimental data includes: process parameters, equipment status, and product characteristic data.
[0012] In one implementation, if the experimental data does not meet the target, the strategy of the process parameter prediction unit and the model prediction control unit is iteratively updated based on the experimental data and the multi-objective optimization model to obtain new process parameter combinations and control commands, until the experimental data meets the target. The steps include: Based on the experimental data, extract key performance indicators corresponding to the performance of the target product, and determine whether the experimental data meets the standards based on the key performance indicators. If the experimental data does not meet the standard, the process parameter prediction unit and the model prediction control unit are updated according to the experimental data and the multi-objective optimization model. After the strategy update is completed, the process continues to execute the steps of using the intelligent control module, employing the process parameter prediction unit to predict the combination of process parameters based on the target product performance, and employing the model prediction control unit to generate at least one control instruction corresponding to the sub-task based on the combination of process parameters and the current system state, until the experimental data meets the target.
[0013] Secondly, embodiments of the present invention also provide a material self-evolution control system for intelligent laboratories, the system comprising: The process management module is used to break down experimental tasks into several sub-tasks; the experimental tasks include material preparation tasks. The intelligent control module is used to use a process parameter prediction unit to predict the combination of process parameters based on the performance of the target product, and to use a model prediction control unit to generate at least one control command corresponding to the sub-task based on the combination of process parameters and the current system state. The decision execution module is used to receive the control commands and send them to the experimental equipment; The data acquisition module is used to monitor the experimental equipment and obtain experimental data corresponding to the experimental task. The intelligent control module is also used to iteratively update the strategies of the process parameter prediction unit and the model prediction control unit based on the experimental data and the multi-objective optimization model if the experimental data does not meet the standards, so as to obtain new combinations of process parameters and control commands until the experimental data meets the standards.
[0014] Thirdly, embodiments of the present invention also provide a terminal, the terminal including a memory and one or more processors; the memory stores one or more programs; the programs include instructions for executing the material self-evolution control method for smart laboratories as described above; the processor is used to execute the programs.
[0015] Fourthly, embodiments of the present invention also provide a computer-readable storage medium having stored thereon a plurality of instructions adapted to be loaded and executed by a processor to implement the steps of the material self-evolution control method for smart laboratories as described above.
[0016] The beneficial effects of this invention are as follows: In this embodiment, the experimental task is decomposed into several sub-tasks through a process management module; the experimental task includes a material preparation task; through an intelligent control module, a process parameter prediction unit predicts the combination of process parameters based on the target product performance, and a model prediction control unit generates at least one control instruction corresponding to the sub-task based on the process parameter combination and the current system state; through a decision execution module, the control instructions are received and sent to the experimental equipment; through a data acquisition module, the experimental equipment is monitored to obtain the experimental data corresponding to the experimental task; if the experimental data does not meet the standards, the strategies of the process parameter prediction unit and the model prediction control unit are iteratively updated based on the experimental data and the multi-objective optimization model to obtain new process parameter combinations and control instructions until the experimental data meets the standards. This invention does not rely on fixed process parameters and control strategies, but can automatically adjust process parameters and control strategies based on experimental results, achieving continuous optimization of experimental efficiency, result accuracy, and material performance. Attached Figure Description
[0017] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.
[0018] Figure 1 This is a flowchart illustrating the material self-evolution control method for smart laboratories provided in this embodiment of the invention.
[0019] Figure 2 This is a principle block diagram of the material self-evolution control method for smart laboratories provided in this embodiment of the invention.
[0020] Figure 3 This is a schematic diagram of a material self-evolution control system for smart laboratories provided in an embodiment of the present invention.
[0021] Figure 4 This is a schematic diagram of the terminal provided in the embodiment of the present invention. Detailed Implementation
[0022] This invention discloses a material self-evolution control method, system, and terminal for intelligent laboratories. To make the objectives, technical solutions, and effects of this invention clearer and more explicit, the invention is further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only for explaining the invention and are not intended to limit the invention.
[0023] Those skilled in the art will understand that, unless specifically stated otherwise, the singular forms “a,” “an,” “the,” and “the” used herein may also include the plural forms. It should be further understood that the term “comprising” as used in this specification means the presence of the stated features, integers, steps, operations, elements, and / or components, but does not exclude the presence or addition of one or more other features, integers, steps, operations, elements, components, and / or groups thereof. It should be understood that when we say an element is “connected” or “coupled” to another element, it can be directly connected or coupled to the other element, or there may be intermediate elements. Furthermore, “connected” or “coupled” as used herein can include wireless connections or wireless coupling. The term “and / or” as used herein includes all or any units and all combinations of one or more associated listed items.
[0024] It will be understood by those skilled in the art that, unless otherwise defined, all terms used herein (including technical and scientific terms) have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains. It should also be understood that terms such as those defined in general dictionaries should be understood to have the same meaning as in the context of the prior art, and should not be interpreted in an idealized or overly formal sense unless specifically defined as herein.
[0025] To address the aforementioned shortcomings of existing technologies, this invention provides a material self-evolution control method for smart laboratories, such as... Figure 1 As shown, the method specifically includes the following steps: Step S100: The experimental task is decomposed into several sub-tasks through the process management module; the experimental task includes the material preparation task.
[0026] Specifically, the method in this embodiment is applicable to intelligent laboratories, which refer to experimental environments that integrate automated equipment, sensor networks, and artificial intelligence algorithms, possessing the capabilities for autonomous decision-making, task scheduling, experimental execution, and result analysis. This embodiment pre-sets a process management module for decomposing the various stages of the experimental task. For example, if the experimental task is a materials preparation task, and it pertains to quantum materials, the various stages of the materials preparation task include, but are not limited to, weighing, mixing, pressing, and sintering.
[0027] In one implementation, the process management module is further configured to: determine the priority of each subtask and the scheduling order of each functional module; each functional module includes: the data acquisition module, the intelligent control module, the feedback optimization module, and the decision execution module.
[0028] Specifically, the process management module is also responsible for the scheduling and dependency management of the entire experimental task. In other words, the process management module not only decomposes the operations at each stage, but also sets priorities for each stage and performs unified scheduling of various functional modules based on time sequences, ultimately achieving automated logical control of each stage's operations. In short, the process management module is responsible not only for task breakdown but also for overall system scheduling. By clearly defining priorities and scheduling sequences, it ensures that the progress of each subtask and the collaboration of each functional module can be executed in an orderly manner, avoiding chaos.
[0029] In one implementation, the process management module is further used to: manage multiple experimental tasks in parallel.
[0030] The process management module also supports parallel management of multiple experimental tasks (or experimental exploration tasks). Specifically, the module can use a distributed scheduling mechanism to run multiple experimental tasks in parallel. This distributed scheduling mechanism involves breaking down multiple experimental tasks and distributing them across multiple computing nodes or functional modules of the system for simultaneous execution. The module can also adaptively allocate experimental and computing resources through a performance evaluation function. This function comprehensively evaluates the importance, urgency, resource consumption intensity, and execution efficiency of each experimental task based on factors such as task priority, potential value, and resource requirements. This allows for the adaptive allocation of resources such as computing power, device usage time, and data storage bandwidth, thereby improving experimental efficiency and product consistency. In practical applications, the process management module can achieve coordination and concurrent management through process state machines or hierarchical control algorithms.
[0031] Step S200: Through the intelligent control module, the process parameter prediction unit predicts the combination of process parameters based on the performance of the target product, and the model prediction control unit generates at least one control instruction corresponding to the sub-task based on the combination of process parameters and the current system state.
[0032] This embodiment includes an intelligent control module that, through a process parameter prediction unit and a model predictive control unit (MPC), automates and self-learns the operation of each stage of the experimental task. It can also automatically optimize process parameters and update the control strategy based on experimental feedback. Specifically, the process parameter prediction unit can be established using a data-based self-learning unit, such as a reinforcement learning (RL) algorithm, which learns the relationship between different process parameters and product performance by exploring the experimental parameter space. Exploring the experimental parameter space refers to the process by which the algorithm finds the optimal combination of experimental parameters in a multi-dimensional parameter space using methods such as Bayesian optimization and reinforcement learning. The model predictive control unit calculates the next optimal control command based on the process parameter combination predicted by the process parameter prediction unit and the real-time system state. This control command corresponds to at least one subtask and is used to advance the experimental task. In practical applications, the process parameter prediction unit and the model predictive control unit can be combined to form an exploration-prediction-correction self-evolving control loop.
[0033] In one implementation, the process parameter prediction unit is specifically used for: By learning the correlation between the combination of process parameters and product performance in advance, the combination of process parameters can be predicted based on the performance of the target product.
[0034] Specifically, in practical applications, the process parameter prediction unit can be used to reverse-engineer parameters, that is, starting from the desired target product performance, to calculate the combination of process parameters that can achieve that target product performance. The combination of process parameters is not a single parameter, but a set of parameters adapted to the target product performance. For example, in the preparation of quantum materials, the proportions of different raw materials, mixing time, sintering temperature, and holding time are clearly defined, and these parameters constitute the combination of process parameters.
[0035] In one implementation, the model prediction control unit is specifically used for: The control command is generated based on the combination of process parameters and the current system state using a sliding time window algorithm.
[0036] Specifically, in practical applications, the model predictive control unit (MMCU) can be used for parameter implementation, transforming abstract combinations of process parameters into control commands that experimental equipment can directly execute and adapt to real-time conditions. The MMCU first receives the process parameter combinations output by the process parameter prediction unit, and then collects real-time system status data, such as equipment operating status, ambient temperature and humidity, execution results of previous subtasks, and real-time material status. Based on these two types of information—process parameter combinations and current system status—accurate control commands are generated, such as weighing commands for automatic weighing machines and heating commands for temperature-controlled furnaces. The MMCU can also optimize future control sequences using a sliding time window algorithm: starting from the current moment, the sliding time window algorithm defines a short-term future time window, continuously predicts system state changes within that window, and dynamically adjusts control commands based on the prediction results. Taking the sintering curve and mixing ratio as an example, if the mixing subtask is delayed by 2 minutes due to material status, the sliding time window algorithm will adjust the heating start time and holding time of the sintering curve to ensure the overall process flow remains seamless and guarantees product performance stability.
[0037] Step S300: Receive the control command through the decision execution module and send it to the experimental equipment.
[0038] Specifically, control commands are sent to the experimental equipment (i.e., the equipment execution end) through the decision execution module. The decision execution module interacts with the experimental equipment through a unified communication interface to achieve automatic execution and feedback reception of experimental commands. For example, the decision execution module can interact with the experimental equipment through industrial automation communication protocols (such as Modbus Communication Protocol, MODBUS) and transmission control protocols (such as Transmission Control Protocol, TCP).
[0039] Step S400: Monitor the experimental equipment through the data acquisition module to obtain the experimental data corresponding to the experimental task.
[0040] Specifically, the data acquisition module can collect multimodal experimental data in real time. Multimodal data refers to various types of experimental states and results, such as process parameters (e.g., temperature, time), equipment status, and product characteristic data (e.g., quality, spectral features). The data acquired by the acquisition module is preprocessed and stored in the experimental database, which can be used for algorithm model training and feedback, or as input for decision support. Through the data acquisition module, the system can monitor the execution status and abnormal signals in real time, achieving safe closed-loop control. Closed-loop control refers to the formation of a feedback loop between experimental execution and data analysis. The system dynamically adjusts the next experimental operation or experimental conditions based on the real-time detected experimental results, achieving automatic iterative optimization.
[0041] Step S500: If the experimental data does not meet the standard, the strategies of the process parameter prediction unit and the model prediction control unit are iteratively updated based on the experimental data and the multi-objective optimization model to obtain new process parameter combinations and control commands until the experimental data meets the standard.
[0042] like Figure 2 As shown, this embodiment can also dynamically optimize process parameters based on experimental data fed back from the data acquisition module, combined with a multi-objective optimization model, forming a closed-loop control mechanism from experimental data acquisition to parameter redistribution, thereby achieving self-evolution of the experimental path and performance optimization. Specifically, substandard experimental data can reflect defects in the current combination of process parameters and / or control commands. For example, data showing insufficient product purity may indicate insufficient raw material weighing accuracy or a low sintering temperature. Therefore, substandard experimental data can provide specific correction directions for the strategy optimization of the process parameter prediction unit and the model prediction control unit. The multi-objective optimization model will comprehensively balance multiple optimization objectives of the experimental task, such as material performance, energy consumption, and time cost, to avoid the situation where the optimization of a single objective excessively affects other indicators. The multi-objective optimization model will calculate the optimal balance point, providing optimization constraints for the strategy optimization of the process parameter prediction unit and the model prediction control unit. Strategy optimization refers to the iterative update of the strategy, the key of which lies in correcting the parameter prediction logic and command generation logic, rather than simply adjusting the values.
[0043] In one implementation, a multi-objective optimization model can be built based on historical experimental databases and performance evaluation results to dynamically correct experimental paths and execution instructions.
[0044] In one implementation, if the experimental data does not meet the target, the strategy of the process parameter prediction unit and the model prediction control unit is iteratively updated based on the experimental data and the multi-objective optimization model to obtain new process parameter combinations and control commands, until the experimental data meets the target. The steps include: Based on the experimental data, extract key performance indicators corresponding to the performance of the target product, and determine whether the experimental data meets the standards based on the key performance indicators. If the experimental data does not meet the standard, the process parameter prediction unit and the model prediction control unit are updated according to the experimental data and the multi-objective optimization model. After the strategy update is completed, the process continues to execute the steps of using the intelligent control module, employing the process parameter prediction unit to predict the combination of process parameters based on the target product performance, and employing the model prediction control unit to generate at least one control instruction corresponding to the sub-task based on the combination of process parameters and the current system state, until the experimental data meets the target.
[0045] Specifically, such as Figure 2As shown, key performance indicators, such as conductivity, magnetization, Raman and PXRD related peaks, are extracted from experimental data to determine whether the current product meets the performance standards corresponding to the target product. If the standards are not met, the subsequent strategy update process is directly triggered. The strategy update process mainly updates the parameters of the process parameter prediction unit and the model prediction control unit to ensure that the next generated combination of process parameters and control commands can specifically address the problems that have not met the standards, thereby guiding the next round of experiments. After multiple rounds of iterative updates, the intelligent control module can automatically select the optimal parameter region based on the material target (such as conductivity, magnetism, optical response, etc.) to prepare a product with the required performance.
[0046] In one implementation, the strategy update of the intelligent control module is based on a self-learning model; the process parameters and control weights are continuously updated through the self-learning model, thereby achieving self-evolution of the experimental process and multi-objective optimal control.
[0047] Specifically, the multi-objective optimization model first determines multiple optimization objectives for the experimental task, such as material performance, energy consumption, and time cost, and then dynamically calculates the priority weight of each optimization objective based on the needs of the experimental stage. This ensures that the updated strategy prioritizes the optimization objectives with higher weights, without sacrificing the constraints of other optimization objectives.
[0048] The process parameter prediction unit updates its strategy based on the experimental environment, executed actions, and performance feedback signals. The essence of this strategy update is adjusting the prediction logic for process parameter combinations. Similarly, the model prediction control unit's strategy update essentially adjusts the processing logic that transforms process parameter combinations into control commands adapted to the real-time state.
[0049] The strategy updates of the process parameter prediction unit and the model prediction control unit are not performed independently, but rather collaboratively through a self-learning model to avoid strategy conflicts. Specifically, the updated process parameter combinations from the process parameter prediction unit serve as input constraints for the model prediction control unit, ensuring that the generated control commands do not deviate from the parameter targets. Conversely, the control commands fed back by the model prediction control unit also serve as boundary conditions for the action space of the process parameter prediction unit, preventing the process parameter prediction unit from predicting process parameters that the experimental equipment cannot execute.
[0050] The weights of each optimization objective output by the multi-objective optimization model are simultaneously transmitted to both units, ensuring that the parameter prediction of the process parameter prediction unit and the instruction generation of the model prediction control unit both revolve around a unified multi-objective optimization direction.
[0051] The self-learning model integrates the optimization results of the policy updates of the two units to form a global optimization result. Subsequent updates refer to the historical optimization results to avoid repeated trial and error and accelerate policy convergence.
[0052] In one implementation, the intelligent control module also has a feedback optimization function, which can realize knowledge transfer across material systems by establishing a mapping relationship between experimental feature database and process parameters.
[0053] For example, taking the experimental task of preparing Fe-Ni-based quantum single crystals with high spin-state stability as an example, the key steps of the method in this embodiment include: 1. The system automatically calls the weighing module and dispenses ingredients according to the algorithm-suggested proportions: The algorithm-suggested ratio is a process parameter predicted by the intelligent control module. The process management module, based on the priority of the preparation logic, schedules the decision-making execution module to ensure the weighing module executes this parameter, guaranteeing accurate raw material proportions.
[0054] 2. The mixing and grinding module automatically selects the rotation speed and time based on historical feedback: The feedback history consists of correlated data recorded by the data acquisition module during the preceding experiments. The model prediction control unit adaptively generates control commands (including speed and time) based on the process parameters output by the process parameter prediction unit and the current system state, ensuring that the raw materials are accurately proportioned and mixed uniformly.
[0055] 3. The pressure is adaptively adjusted during the tablet compression process based on the sample's adhesion: Sample adhesion is measured in real-time using system state data. The model predictive control unit can generate corresponding control commands based on this system state data to dynamically adjust pressure parameters, preventing tablet cracking or weak adhesion.
[0056] 4. During the sintering stage, the heating rate and holding time are controlled by the model prediction and control unit: Sintering is a crucial step in the growth of Fe-Ni-based quantum single crystals, determining crystal integrity and spin-state stability. The model prediction and control unit uses a data-based self-learning unit to predict the heating rate and holding time, combined with real-time data on system states such as temperature uniformity within the sintering furnace and real-time sample temperature, to dynamically optimize the control sequence using a sliding time window algorithm.
[0057] 5. After the experiment, the sample magnetization curve and crystal integrity were uploaded through the characterization module.
[0058] 6. The process parameter prediction unit updates the strategy network with performance indicators as reward signals and updates the strategy of the model control prediction unit to achieve feedback optimization.
[0059] 7. The updated intelligent control module calculates the new combination of process parameters and proceeds to the next experimental cycle.
[0060] After multiple self-evolutionary cycles, the system can effectively improve the target magnetization and experimental reproducibility.
[0061] Possible technical variations of this invention include: 1. Regarding the intelligent control module: A. Replace the algorithm kernel: The process parameter prediction unit mostly uses reinforcement learning algorithms, but reinforcement learning can be replaced by genetic algorithms, Bayesian optimization, deep Q-network (DQN) or self-attention Transformer controllers.
[0062] B. Optimization strategy fusion method: Instead of using RL+MPC, a collaborative optimization structure is used (e.g., reinforcement learning + simulated annealing or proportional-integral-derivative control (PID) self-adjustment).
[0063] C. Model adaptive update mechanism: from round-by-round optimization to continuous online learning or batch update mode; can be used as a subordinate variation of the algorithm implementation method.
[0064] 2. Regarding the system structure A. Adjustment of module quantity and function: Weighing, mixing, tableting and calcination can be combined into a multi-stage process execution module; or an in-situ characterization module and a visual recognition module can be added.
[0065] B. Interface and Communication Method: The communication method can be changed from Ethernet control to RS485, CAN or cloud application programming interface (API) mode.
[0066] C. Deployment variations: The system can be deployed on a single experimental device, a distributed control cluster, or a cloud-based laboratory platform.
[0067] 3. For application scenarios: A. Expansion of material categories: from quantum single crystals to two-dimensional materials, solid electrolytes, and optoelectronic functional crystals.
[0068] B. Platform Application Expansion: Adaptable to various experimental systems such as Chemical Vapor Deposition (CVD), Metal-Organic Chemical Vapor Deposition (MOCVD), flux method, and Czochralski method; forming a general intelligent control architecture.
[0069] C. AI-driven automated experiment networking: Connecting multiple experimental sites to form a collaborative self-learning experimental cloud.
[0070] Based on the above embodiments, the present invention also provides a material self-evolution control system for smart laboratories, such as... Figure 3As shown, the system includes: Process management module 01 is used to decompose experimental tasks into several sub-tasks; the experimental tasks include material preparation tasks; The intelligent control module 02 is used to use a process parameter prediction unit to predict the combination of process parameters based on the performance of the target product, and to use a model prediction control unit to generate at least one control instruction corresponding to the sub-task based on the combination of process parameters and the current system state. The decision execution module 03 is used to receive the control commands and send them to the experimental equipment; The data acquisition module 04 is used to monitor the experimental equipment and obtain the experimental data corresponding to the experimental task. The intelligent control module 02 is also used to iteratively update the strategies of the process parameter prediction unit and the model prediction control unit based on the experimental data and the multi-objective optimization model if the experimental data does not meet the standards, so as to obtain new combinations of process parameters and control commands until the experimental data meets the standards.
[0071] The system in this embodiment is a self-evolving control system. That is, the system in this embodiment does not rely on fixed manual rules, but can automatically adjust experimental parameters and control strategies based on experimental feedback data to achieve continuous optimization of experimental efficiency, result accuracy or material properties.
[0072] The system in this embodiment can adopt a multi-layer control architecture, dividing the control logic into multiple parts such as the task layer, scheduling layer, and execution layer, which are respectively responsible for setting experimental objectives, coordinating processes, and executing at the lower level.
[0073] The system in this embodiment has high versatility and adaptability, and is suitable for the automated preparation and laboratory-level intelligent control of quantum single crystals, two-dimensional quantum materials and other solid-state quantum functional materials.
[0074] Based on the above embodiments, the present invention also provides a terminal, the principle block diagram of which can be as follows: Figure 4 As shown, the terminal includes a processor, memory, network interface, and display screen connected via a system bus. The processor provides computing and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system and computer programs. The internal memory provides the environment for the operation of the operating system and computer programs in the non-volatile storage media. The network interface is used to communicate with external terminals via a network connection. When the computer program is executed by the processor, it implements a material self-evolution control method for smart laboratories. The display screen can be a liquid crystal display (LCD) or an e-ink display.
[0075] Those skilled in the art will understand that Figure 4The schematic diagram shown is merely a partial structural diagram related to the present invention and does not constitute a limitation on the terminal to which the present invention is applied. A specific terminal may include more or fewer components than those shown in the figure, or combine certain components, or have different component arrangements.
[0076] In one implementation, the terminal's memory stores one or more programs, and these programs are configured to be executed by one or more processors, and the programs contain instructions for performing a material self-evolution control method for a smart laboratory.
[0077] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium. When executed, the computer program can include the processes of the embodiments of the above methods. Any references to memory, storage, databases, or other media used in the embodiments provided by this invention can include non-volatile and / or volatile memory. Non-volatile memory can include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memory can include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in various forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), dual data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link DRAM (SLDRAM), Rambus direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM), etc.
[0078] In summary, this invention discloses a material self-evolution control method, system, and terminal for intelligent laboratories. The method includes: decomposing an experimental task into several sub-tasks through a process management module; the experimental task includes a material preparation task; through an intelligent control module, a process parameter prediction unit predicts a combination of process parameters based on the target product performance, and a model prediction control unit generates at least one control instruction corresponding to the sub-task based on the process parameter combination and the current system state; through a decision execution module, the control instruction is received and sent to the experimental equipment; through a data acquisition module, the experimental equipment is monitored to obtain experimental data corresponding to the experimental task; if the experimental data does not meet the standards, the strategies of the process parameter prediction unit and the model prediction control unit are iteratively updated based on the experimental data and a multi-objective optimization model to obtain new process parameter combinations and control instructions until the experimental data meets the standards. This invention does not rely on fixed process parameters and control strategies, but can automatically adjust process parameters and control strategies based on experimental results, achieving continuous optimization of experimental efficiency, result accuracy, and material performance.
[0079] It should be understood that the application of the present invention is not limited to the examples above. Those skilled in the art can make improvements or modifications based on the above description, and all such improvements and modifications should fall within the protection scope of the appended claims.
Claims
1. A material self-evolution control method for intelligent laboratories, characterized in that, The method includes: The experimental task is broken down into several sub-tasks through the process management module; the experimental task includes the material preparation task. The intelligent control module uses a process parameter prediction unit to predict the combination of process parameters based on the performance of the target product, and a model prediction control unit uses a model prediction control unit to generate at least one control instruction corresponding to the sub-task based on the combination of process parameters and the current system state. The decision execution module receives the control commands and sends them to the experimental equipment. The experimental equipment is monitored through the data acquisition module to obtain the experimental data corresponding to the experimental task; If the experimental data does not meet the standard, the strategies of the process parameter prediction unit and the model prediction control unit are iteratively updated based on the experimental data and the multi-objective optimization model to obtain new combinations of process parameters and control commands until the experimental data meets the standard.
2. The material self-evolution control method for intelligent laboratories according to claim 1, characterized in that, The process management module is also used to: determine the priority of each subtask and the scheduling order of each functional module; each functional module includes: the data acquisition module, the intelligent control module, the feedback optimization module, and the decision execution module.
3. The material self-evolution control method for intelligent laboratories according to claim 1, characterized in that, The process management module is also used to manage multiple experimental tasks in parallel.
4. The material self-evolution control method for intelligent laboratories according to claim 1, characterized in that, The process parameter prediction unit is specifically used for: By learning the correlation between the combination of process parameters and product performance in advance, the combination of process parameters can be predicted based on the performance of the target product.
5. The material self-evolution control method for intelligent laboratories according to claim 1, characterized in that, The model prediction control unit is specifically used for: The control command is generated based on the combination of process parameters and the current system state using a sliding time window algorithm.
6. The material self-evolution control method for intelligent laboratories according to claim 1, characterized in that, The experimental data includes: process parameters, equipment status, and product characteristic data.
7. The material self-evolution control method for intelligent laboratories according to claim 1, characterized in that, If the experimental data does not meet the standards, the strategy of the process parameter prediction unit and the model prediction control unit is iteratively updated based on the experimental data and the multi-objective optimization model to obtain new combinations of process parameters and control commands, until the experimental data meets the standards. The steps include: Based on the experimental data, extract key performance indicators corresponding to the performance of the target product, and determine whether the experimental data meets the standards based on the key performance indicators. If the experimental data does not meet the standard, the process parameter prediction unit and the model prediction control unit are updated according to the experimental data and the multi-objective optimization model. After the strategy update is completed, the process continues to execute the steps of using the intelligent control module, employing the process parameter prediction unit to predict the combination of process parameters based on the target product performance, and employing the model prediction control unit to generate at least one control instruction corresponding to the sub-task based on the combination of process parameters and the current system state, until the experimental data meets the target.
8. A material self-evolution control system for intelligent laboratories, characterized in that, The system includes: The process management module is used to break down experimental tasks into several sub-tasks; the experimental tasks include material preparation tasks. The intelligent control module is used to use a process parameter prediction unit to predict the combination of process parameters based on the performance of the target product, and to use a model prediction control unit to generate at least one control command corresponding to the sub-task based on the combination of process parameters and the current system state. The decision execution module is used to receive the control commands and send them to the experimental equipment; The data acquisition module is used to monitor the experimental equipment and obtain experimental data corresponding to the experimental task. The intelligent control module is also used to iteratively update the strategies of the process parameter prediction unit and the model prediction control unit based on the experimental data and the multi-objective optimization model if the experimental data does not meet the standards, so as to obtain new combinations of process parameters and control commands until the experimental data meets the standards.
9. A terminal, characterized in that, The terminal includes a memory and one or more processors; the memory stores one or more programs; the programs contain instructions for executing the material self-evolution control method for smart laboratories as described in any one of claims 1 to 7; the processors are used to execute the programs.
10. A computer-readable storage medium storing a plurality of instructions thereon, characterized in that, The instructions are applicable to being loaded and executed by a processor to implement the steps of the material self-evolution control method for smart laboratories as described in any one of claims 1 to 7.