Chemical experiment monitoring optimization method and device based on digital twinning

Abstract

The application discloses a chemical experiment monitoring optimization method and device based on digital twinning and electronic equipment, and the method comprises the following steps: dividing a laboratory into multiple operation unit areas, and creating corresponding management shells for each operation unit area to control experiments and collect data; converting a current experiment process into experiment tasks and experiment operations, and mapping collected data to corresponding experiment objects; creating and training a prediction model, inputting the data mapped to the experiment objects into the prediction model to monitor the state of the current experiment; and optimizing the current experiment process according to the result of the current experiment process and the data mapped to the experiment objects. The application combines data collection, data mapping and a prediction model to establish a closed loop from data collection to experiment monitoring optimization in the laboratory, so that data collection, prediction, monitoring, index analysis, experiment optimization and simulation display of the experiment process are realized.

Description

Technical Field

[0001] This invention relates to the field of recommendation technology, and more specifically, to a method, apparatus, electronic device, and computer-readable medium for optimizing chemical experiment monitoring based on digital twins. Background Technology

[0002] Chemical experiments are primarily used to support chemical teaching or research. They typically require operators to follow experimental procedures, record data, and analyze that data. Clearly, this traditional approach relies excessively on manual labor. Furthermore, the diverse nature of chemical reactions in laboratories, with each reaction having its own unique operational procedures, makes it difficult to predict and monitor the experimental process. Additionally, optimizing experimental procedures requires repeated experimentation, resulting in significant expenditure of human and financial resources. Summary of the Invention

[0003] The present invention aims to solve the technical problem that the existing technology cannot effectively monitor, analyze and predict data from chemical experiments, resulting in huge losses of human and financial resources.

[0004] To address the aforementioned technical problems, the first aspect of this invention proposes a method for optimizing chemical experiment monitoring based on digital twins, the method comprising:

[0005] The laboratory is divided into multiple operating unit areas, and a corresponding management shell is created for each operating unit area for experimental control and data acquisition.

[0006] The current experimental process is converted into experimental tasks and experimental operations, and the collected data is mapped to the corresponding experimental objects; the experimental objects include experimental tasks and / or experimental operations.

[0007] Create and train a prediction model, and input the data mapped to the experimental objects into the prediction model to monitor the current experimental status;

[0008] The current experimental procedure is optimized based on the results and the data mapped onto the experimental objects.

[0009] According to a preferred embodiment of the present invention, mapping the collected data to the corresponding experimental object includes:

[0010] Create a mapping relationship between each instrument and the experimental object in the operation unit area;

[0011] Based on the mapping relationship, the data collected from each operation unit area is mapped to the corresponding experimental object.

[0012] According to a preferred embodiment of the present invention, the experimental object is an experimental task, and the experimental task includes a reaction. The data mapped onto the experimental object is a reaction formula. Then, a first model is created and trained, and the reaction formula is input into the first model to predict and monitor the feasibility probability of the current reaction.

[0013] If the experimental object is an experimental operation and the data mapped onto the experimental object is a time series of experimental operation parameters, then a second model is created and trained. The time series of experimental operation parameters is input into the second model to predict and monitor the probability of an anomaly occurring in the current experimental task.

[0014] The experimental object is an experimental operation, and the data mapped to the experimental object does not include the time series of experimental operation parameters; then the execution result of the current experimental task is predicted by the distribution of the historical execution results of the experimental operation.

[0015] According to a preferred embodiment of the present invention, optimizing the current experimental procedure based on the results of the current experimental procedure and the data mapped onto the experimental object includes:

[0016] Determine the results of the current experimental procedure and the parameters mapped to the experimental operation. Adjust the parameters of the experimental procedure and / or experimental operation according to the preset reaction parameters and the process optimization algorithm, and obtain the adjusted experimental procedure execution results.

[0017] Alternatively, determine the results of the current experimental procedure and the data mapped to the experimental operation, and adjust the experimental operation based on the analysis results of the current experimental procedure results and experimental operation data.

[0018] According to a preferred embodiment of the present invention, adjusting the experimental operation based on the analysis results of the current experimental procedure and experimental operation data includes:

[0019] Analyze the causal relationship between the current experimental results and the data mapped to the experimental operations, and make the first adjustment to the experimental operations;

[0020] And / or, compare the results of multiple executions of the same experimental operation in multiple experimental procedures, and make a second adjustment to the experimental operation based on the execution results.

[0021] According to a preferred embodiment of the present invention, the method further includes:

[0022] After making the first and second adjustments to the experimental operation, the experimental operation before and after the adjustment was executed multiple times, and the execution data was analyzed to verify the efficiency of the adjustment.

[0023] According to a preferred embodiment of the present invention, the method further includes:

[0024] The business metrics of the current experimental process are analyzed in real time based on the data mapped to the experimental objects; the business metrics include at least one of the following: experimental efficiency, manual workload, and automation rate.

[0025] According to a preferred embodiment of the present invention, the method further includes:

[0026] Collect the time distribution of similar experimental tasks during execution;

[0027] The execution time of the current experimental procedure is predicted based on the time consumption distribution.

[0028] According to a preferred embodiment of the present invention, the method further includes:

[0029] Send the monitoring results of the current experimental process, the business metrics of the current experimental process, and the execution time of the current experimental process to the user.

[0030] According to a preferred embodiment of the present invention, the method further includes:

[0031] The simulation demonstrates the execution effects before and after the optimization of the experimental process;

[0032] Receive the experimental support conditions input by the user, and find the target experimental process that matches the experimental support conditions based on the business indicators and execution time of each experimental process.

[0033] The system receives experimental expansion conditions input by the user and searches for experimental support conditions that match the experimental expansion conditions based on the business indicators of each experimental process.

[0034] To address the aforementioned technical problems, a second aspect of the present invention provides a chemical experiment monitoring and optimization device based on digital twins, the device comprising:

[0035] The data acquisition module is used to divide the laboratory into multiple operating unit areas and create a corresponding management shell for each operating unit area for experimental control and data acquisition.

[0036] The mapping module is used to convert the current experimental process into experimental tasks and experimental operations, and to map the collected data to the corresponding experimental objects; the experimental objects include experimental tasks and / or experimental operations.

[0037] The monitoring module is used to create and train a prediction model, and input the data mapped to the experimental object into the prediction model to monitor the current experimental status.

[0038] The optimization module is used to optimize the current experimental process based on the results of the current experimental process and the data mapped onto the experimental objects.

[0039] To address the aforementioned technical problems, a third aspect of the present invention provides an electronic device, comprising:

[0040] Processor; and

[0041] A memory that stores computer-executable instructions, which, when executed, cause the processor to perform the methods described above.

[0042] To address the aforementioned technical problems, a fourth aspect of the present invention provides a computer-readable storage medium, wherein the computer-readable storage medium stores one or more programs that, when executed by a processor, implement the aforementioned method.

[0043] This invention divides the laboratory into multiple operational unit areas and creates a corresponding management shell for each area to control experiments and collect data, achieving automated data acquisition. It converts the current experimental process into experimental tasks and operations, and maps the collected data to corresponding experimental objects; this completes the connection between experimental data and specific experimental objects, achieving business-level modeling. By creating and training a prediction model, the data mapped to the experimental objects is input into the prediction model to monitor the current experimental status; finally, the current experimental process is optimized based on the results of the current experimental process and the data mapped to the experimental objects. Ultimately, a closed loop from data acquisition to experimental monitoring and optimization is established in the laboratory. Compared with existing technologies, this invention has at least the following beneficial effects:

[0044] 1. Create a corresponding management shell for each operation unit area to control experiments and collect data, thereby achieving automated data acquisition;

[0045] 2. By creating and training predictive models, the physical simulation modeling process of chemical reactions can be bypassed, thereby enabling the prediction and monitoring of experimental results and processes.

[0046] 3. No need for repeated experiments for optimization; it can automatically optimize the current experimental process based on the results of the current experimental process and the data mapped to the experimental objects, effectively saving manpower and financial resources.

[0047] 4. It can analyze laboratory indicators based on data mapped to experimental objects, and obtain information such as laboratory execution efficiency, experimental time, and reaction success rate, providing support for subsequent experiments.

[0048] 5. Based on the data mapped to the experimental objects and the prediction model, it can provide experimental monitoring results, experimental business indicators, and experimental execution time for decision-making reference.

[0049] 6. Able to simulate experiments based on experimental monitoring results, experimental business indicators, and experimental execution time, and estimate the support and expansion conditions required for the experiment. Attached Figure Description

[0050] To make the technical problems solved by this invention, the technical means employed, and the technical effects achieved clearer, specific embodiments of this invention will be described in detail below with reference to the accompanying drawings. However, it should be noted that the drawings described below are merely drawings of exemplary embodiments of this invention. Those skilled in the art can obtain drawings of other embodiments based on these drawings without any creative effort.

[0051] Figure 1 This is a flowchart illustrating a chemical experiment monitoring and optimization method based on digital twins according to the present invention.

[0052] Figure 2 This is a schematic diagram of the implementation framework of the chemical experiment monitoring optimization based on digital twins of the present invention;

[0053] Figure 3 This is a schematic diagram of the structural framework of the chemical experiment monitoring and optimization device based on digital twins according to the present invention;

[0054] Figure 4 This is a structural block diagram of an exemplary embodiment of an electronic device according to the present invention;

[0055] Figure 5 This is a schematic diagram of an embodiment of the present invention in a computer-readable medium. Detailed Implementation

[0056] Exemplary embodiments of the invention will now be described more fully with reference to the accompanying drawings. While these exemplary embodiments can be implemented in various specific ways, they should not be construed as limiting the invention to the embodiments set forth herein. Rather, these exemplary embodiments are provided to make the content of the invention more complete and to facilitate a full communication of the inventive concept to those skilled in the art.

[0057] Subject to the inventive concept, the structures, performance, effects or other features described in a particular embodiment may be combined in any suitable manner with one or more other embodiments.

[0058] In the description of specific embodiments, detailed descriptions of structures, performance, effects, or other features are provided to enable those skilled in the art to fully understand the embodiments. However, this does not preclude those skilled in the art from implementing the present invention with technical solutions that do not contain the aforementioned structures, performance, effects, or other features under specific circumstances.

[0059] The flowchart in the accompanying drawings is merely an exemplary process demonstration and does not imply that the solution of this invention must include all the content, operations, and steps in the flowchart, nor does it imply that they must be executed in the order shown in the diagram. For example, some operations / steps in the flowchart can be decomposed, some operations / steps can be combined or partially combined, etc. Without departing from the inventive spirit of this invention, the execution order shown in the flowchart can be changed according to the actual situation.

[0060] The box in the attached diagram Figure 1 Generally, these refer to functional entities, and do not necessarily correspond to physically independent entities. That is, these functional entities can be implemented in software, in one or more hardware modules or integrated circuits, or in different network and / or processor devices and / or microcontroller devices.

[0061] The same reference numerals in the accompanying drawings denote the same or similar elements, components, or parts, and therefore, repeated descriptions of the same or similar elements, components, or parts may be omitted below. It should also be understood that although terms such as first, second, third, etc., indicating numbers may be used herein to describe various devices, elements, components, or parts, these devices, elements, components, or parts should not be limited by these terms. That is, these terms are only used to distinguish one from another. For example, a first device may also be referred to as a second device, without departing from the essential technical solution of the invention. Furthermore, the terms "and / or" and "and / or" refer to all combinations including any one or more of the listed items.

[0062] In this embodiment of the invention, digital twins fully utilize data such as physical models, sensor updates, and operational history to integrate multi-disciplinary, multi-physical-quantity, multi-scale, and multi-probability simulation processes, completing mapping in virtual space to reflect the entire lifecycle of the corresponding physical equipment. It constructs a precise physical object model in digital space based on modeling tools, then uses real-time IoT data to drive the model's operation, and finally builds comprehensive decision-making capabilities through data and model integration, promoting closed-loop optimization of the entire industrial business process.

[0063] Currently, digital twin technology cannot be applied in chemical laboratories, mainly for the following reasons:

[0064] 1. In existing chemical laboratory settings, most operations rely on manual labor, and some equipment (such as rotary evaporators) lacks data acquisition capabilities, thus failing to provide the data required for modeling.

[0065] 2. The diversity of chemical reactions in the laboratory and the different operation procedures for each reaction make it difficult to model and predict the experimental process.

[0066] 3. The success of an experiment or task in a chemistry laboratory is not determined solely by physical equipment and its operation. For example, the success of a chemical reaction is primarily determined by the reaction mechanism, the selection of reaction conditions, raw materials, operational steps (such as the process and sequence of material addition), and specific operational decisions (whether there are any operational errors). In a reactor, information such as reaction time and temperature can be read from instruments, but other information is difficult to simulate through physical simulation. Therefore, the process of a chemical reaction is difficult to predict using physical models. Chemical reaction simulation software like COMSOL is only used to simulate large-scale, mature reactions, and lacks solutions for new chemical reactions.

[0067] Based on this, the present invention provides a method for monitoring and optimizing chemical experiments based on digital twins, so as to realize data acquisition, prediction, monitoring and optimization of chemical experiments.

[0068] Figure 1 This is a flowchart illustrating a chemical experiment monitoring and optimization method based on digital twins provided by the present invention; as shown below. Figure 1 As shown, the method includes:

[0069] S1. Divide the laboratory into multiple operating unit areas and create a corresponding management shell for each operating unit area for experimental control and data acquisition;

[0070] Among them, the management shell is a "shell" that adds a digital model to the physical device. Multiple devices in the management shell receive cloud instructions and report the actual execution results through a certain connection method (including but not limited to WebSocket long connection and MQTT), thereby realizing data exchange.

[0071] For example, the laboratory can be divided into three areas based on the operating entity: automated operation area, manual operation area, and mixed operation area. The automated operation area includes instruments that can perform experiments automatically, such as automated reactors, automated weighing equipment, and liquid chromatography-mass spectrometry (LC-MS) systems. Figure 2 Data generated by instruments within the automated operation area can be collected via protocols such as RS485, Modbus, and OPC-UA, and reported to the cloud via MQTT.

[0072] The manual operation area includes instruments requiring manual experimental operation, such as ordinary fume hoods. In this embodiment, an interactive terminal (such as a tablet) can be installed in the manual operation area. The operator works according to the instructions on the terminal and records the process and results of the instrument operation into the terminal. Figure 2The data is reported to the cloud via an HTTPS interface.

[0073] The mixed operation area includes instruments that can perform automated experiments but require manual assistance, such as semi-automatic column chromatography machines. Data generated by these instruments can be collected via protocols such as RS485, Modbus, and OPC-UA and reported to the cloud via MQTT. Data entered manually by the terminal can be reported to the cloud via HTTPS interface.

[0074] To facilitate subsequent data processing, the collected data can be further processed, such as... Figure 2 In the data layer, the raw data layer stores the data reported by instruments and interactive terminals. This includes instrument-reported data such as instrument parameters and logs; and data from the interactive terminals such as operator interactions and input. The Data Detail Layer (DWD) cleanses the raw data from the raw data layer, for example, deleting or filling in missing data. For instance, if a reactor has a log at a certain point in time, a normal log should include both a timestamp and content, but the data only contains the timestamp, then this data is useless and should be deleted. The Common Data Model Layer (CDM) defines the data model based on business entities and the relationships between them. ETL is used to normalize the data from the Data Detail Layer and store it in the Common Model Layer. For example, different reactor models may return logs in different formats, requiring conversion to a unified data content before storage. The Data Service Layer (DWS) distributes the cleaned data from the Data Detail Layer, performs light aggregation by topic using ETL, and forms large, wide tables for downstream services to query and use.

[0075] S2. Convert the current experimental process into experimental tasks and experimental operations, and map the collected data to the corresponding experimental objects; the experimental objects include experimental tasks and / or experimental operations.

[0076] In this embodiment, the experimental process includes all procedures for completing the experiment, which can consist of multiple experimental tasks, each of which can be completed by multiple experimental operations. Optionally, the experimental process can be represented by a directed graph constructed from the experimental tasks and their sequence. Each node in the graph represents an experimental task, and each experimental task has corresponding multiple experimental operations for completion. Simple arrows indicate that the previous task is performed first, and the next task is executed after the previous task is completed. Parallel tasks controlled by parallel task switches indicate that the process between the start and end of a parallel task is carried out simultaneously.

[0077] Among them: Experimental task refers to completing a work with a chemical purpose, such as: material requisition, material return, weighing, feeding, reaction, quenching, recrystallization, column chromatography, extraction, filtration, liquid-liquid separation, crystallization, column chromatography, solvent removal, dehydration with desiccant, pH adjustment, etc. Experimental operation refers to a series of operations to complete an experimental task. For example, dehydration with desiccant includes four experimental operations: 1. Add desiccant until no more clumping occurs; 2. Let stand; 3. Filter; 4. Wash.

[0078] In one example, the experimental procedure can be described in text. A conversion model can be pre-trained to convert the experimental procedure text into an experimental flowchart. Using this conversion model, the experimental flowchart is obtained, and experimental tasks and operations can be directly extracted from the flowchart nodes. Specifically, the conversion model can include: a classification module for classifying text at the sentence level to determine the task type of each node; an extraction module for extracting experimental substances and parameters from the sentences; a construction module connected to the classification module for extracting keywords indicating the order of experimental tasks (e.g., before, after, simultaneously, etc.) from the sentences and constructing the sequence of tasks based on these keywords; and a generation module connected to the classification, extraction, and construction modules for generating a flowchart from the nodes, experimental substances, and parameters according to the sequence of tasks. Optionally, the classification, extraction, and construction modules can be implemented by training corresponding models. For example: In the classification module, text representing experimental tasks can be annotated in existing experimental procedure text (e.g., a text describing how to add ingredients can be labeled as "adding ingredients"). Then, a large amount of labeled experimental procedure text is used as a training set to input into the classification model for supervised training. This allows the trained classification model to identify experimental tasks (i.e., the task type of a node) based on the input text. In the extraction module, text representing experimental substances and parameters can be annotated in existing experimental procedure text. Then, a large amount of labeled experimental procedure text is used as a training set to input into the extraction model for supervised training. This allows the trained extraction model to extract experimental substances and parameters from the text. In the construction module, text representing the sequence of experimental tasks can be annotated in existing experimental procedure text. Then, a large amount of labeled experimental procedure text and the experimental tasks within that text are used as a training set to input into the construction model for supervised training. This allows the trained extraction model to extract sequential keywords from the text and construct the sequence of experimental tasks based on these keywords. The generation module can be combined with a flowchart editor, allowing professional scientists to adjust the nodes, experimental substances, parameters, and the order of tasks in a specific chemical experiment to generate a flowchart for that experiment.

[0079] An experimental object refers to the mapping of a real-world experimental entity onto a data model. An experimental object is tracked by a unique identifier (ID), ensuring it remains the same object even if its state and data change. For example, lab employees have employee IDs as unique identifiers; age, years of service, and job level are employee data that may change, but the employee ID remains constant. An experiment has a fixed experiment number, which tracks the experiment's states (never started, in progress, successful, etc.) and its data. This facilitates tracking the lifecycle data and state of an experimental object. In this embodiment, an experimental object can be simply an experimental task, an experimental operation, or both. Optionally, before data mapping, a mapping relationship between each instrument and the experimental object can be created in the operation unit area; that is, different instruments will correspond to different experimental tasks and operations. For example, in one mapping relationship, reactor data in the data service layer is mapped to corresponding experimental tasks and stirring operations. The experimental tasks are determined based on central control signals. For instance, when the central control issues an instruction for experimental task A to the reactor, it records a timestamp t1; when experimental task A is completed, it records a timestamp t2. The reactor's log data between t1 and t2 is then mapped to experimental task A. This step maps the data collected from each operation unit area to the corresponding experimental object based on this mapping relationship.

[0080] Optionally, the present invention can also pre-configure the labeled data for each experimental object. In this case, this step can further determine the execution status of the experimental object by comparing the data mapped to each experimental object with the labeled data of each experimental object. For example, in a reaction task, the expected reaction time is 3 hours, but if the data mapped to the experimental task shows that the reaction stopped after 10 minutes, then the execution status of the reaction task is considered a failure. Furthermore, the reason for the failure of the experimental task can be diagnosed based on the subsequent mapped data. For example, if the subsequent mapped data of the experimental task shows an experimental cancellation status (subsequent experimental data is empty), it can be determined that an abnormal event occurred in the reaction. Further examination of the reactor log can determine the cause of the abnormal termination, such as liquid spraying.

[0081] This step maps the collected data onto data objects, completing the process. Figure 2 The establishment of the business model.

[0082] S3. Create and train a prediction model, and input the data mapped to the experimental object into the prediction model to monitor the current experimental status.

[0083] To address the problem that chemical experiments cannot be simulated and predicted using physical models, this invention employs artificial intelligence technology. By creating and training a prediction model, data mapped onto the experimental object is input into the prediction model to monitor the current experimental status.

[0084] In one example, when the experimental object is an experimental task, and the experimental task includes reactions, and the data mapped to the experimental object is a reaction formula, a first model is created and trained. The reaction formula is input into the first model to predict and monitor the feasibility probability of the current reaction. Considering that the reaction is the most critical part of an experimental task, the first model, which can judge the feasibility probability of a reaction, can be pre-trained using a large number of reaction texts. After training, a test set is established, which includes feasible reactions (reported in the literature) and infeasible reactions (reactions that failed in actual experiments and were judged infeasible by human experts). The performance of the first model is tested using the test set.

[0085] Based on the feasibility of the reaction, this step can also predict the execution status of the experimental task. When the experimental object is an experimental operation, and the data mapped to the experimental object is a time series of experimental operation parameters, a second model is created and trained. The time series of experimental operation parameters is input into the second model to predict and monitor the probability of anomalies occurring in the current experimental task. The experimental operation parameters can be duration, temperature, pressure, etc., and the changes in the reaction parameters are trained through the time series of these parameters. The second model can be pre-trained based on the historical time series of experimental operation parameters and the results of the experimental task, allowing the second model to predict the probability of anomalies occurring in the experimental task based on the time series of experimental operation parameters. For example, when the thermocouple reading rises sharply, the experimental task often experiences an abnormal phenomenon of overheating.

[0086] When the experimental object is an experimental operation, and the data mapped to the experimental object does not include the time series of experimental operation parameters; for situations where it is impossible to determine whether an experimental task has encountered an anomaly from the operational parameters (such as manually executed tasks), the execution result of the current experimental task can be predicted by analyzing the distribution of the historical execution results of that experimental operation. For example, the execution results of the same experimental task can be collected in advance, and the distribution of the execution results can be analyzed through big data, and the execution result of the current experimental task can be predicted based on the distribution.

[0087] As a preferred approach, after training the first and second models, they can be validated to ensure the accuracy of the model outputs. Specifically, the results output by the first / second model are compared with the actual experimental results, running the model in shadow mode: that is, the model runs simultaneously with the experiment, but the model's output is not perceived by the experimenters, in order to assess whether the model's predictions closely match reality. Validation revealed that although there is a large variance in the predictions of a single or a few experiments (the model predicts a high probability of success, but actually fails), in a large number of experiments, the success rate of the model's predictions and the occurrence of various anomalies closely match the actual execution results.

[0088] This step establishes the first and second models for predicting and monitoring experimental states, thus completing the process. Figure 2 The creation of AI models in China.

[0089] Corresponding to Figure 2 This step of monitoring the current experimental status may include: while the experiment is in progress, using the data from the acquisition layer, data layer, and business model mapped to the experimental object to monitor the status of each experimental task in real time, and using the AI ​​model in the model layer to predict the feasibility of the ongoing experimental tasks and the probability of various anomalies.

[0090] S4. Optimize the current experimental procedure based on the results of the current experimental procedure and the data mapped onto the experimental objects.

[0091] In this embodiment, the reaction process and reaction conditions in the experimental procedure can be optimized, and it can be applied in process optimization scenarios. For example... Figure 2 The data acquisition layer determines the results of the current experimental process and the parameters mapped to the experimental operation. Based on preset reaction parameters and a process optimization algorithm, the parameters of the experimental process and / or experimental operation are adjusted, and the adjusted experimental process execution results are obtained, thereby achieving the goal of optimizing the reaction process and reaction conditions. For example, the preset reaction parameters and process optimization algorithm first train a surrogate model for a class of reactions based on training data and a predetermined optimization objective (e.g., yield). Then, the following operations are repeated: 1. Set the acquisition function, and select a series of parameter points by maximizing the acquisition function.

[0092] 2. Conduct experiments based on the parameter points, and update the surrogateposterior model as the new surrogate model based on the experimental results.

[0093] The parameter points obtained until the predetermined optimization target (such as productivity) no longer increases or resources are exhausted are the "optimal parameters" after optimization.

[0094] Furthermore, the experimental procedure of the present invention can also be parameterized, and then the above-mentioned preset reaction parameters and process optimization algorithm are used to optimize and adjust the parameterized experimental procedure.

[0095] This embodiment can also optimize individual experimental operations. This involves determining the results of the current experimental process and the data mapped to the experimental operation, and adjusting the experimental operation based on the analysis of the current experimental process results and experimental operation data. Specifically, the experimental operation can also be parameterized, and then the aforementioned preset reaction parameters and process optimization algorithm are used to optimize and adjust the parameterized experimental operation.

[0096] In one example, the causal relationship between the current experimental procedure results and the data mapped to the experimental operation can be analyzed to make initial adjustments to the experimental operation; this can be applied to optimize the experimental operation. For example, if the experimental operation data reveals that some operators prefer to directly place the heated and dissolved solution into an ice bath during recrystallization, resulting in the abnormal phenomenon of crystal bursting, the experimental operation can be optimized by cooling it to room temperature before slowly cooling it into an ice bath. Preferably, the correspondence between the above causal relationship and the optimized experimental operation can be pre-stored, so that after analyzing the causal relationship, the optimized experimental operation can be directly obtained.

[0097] In another example, the results of multiple executions of the same experimental operation across various experimental procedures can be compared, and the operation can be adjusted based on these results. This approach can be applied to diagnose bottlenecks in experimental operations. For instance, in the data analysis of 56 experimental operations, it was found that weighing operations accounted for 7% of all tasks performed, which is a significant proportion. Automated weighing methods can be developed to replace manual weighing and improve efficiency.

[0098] Furthermore, after optimizing a single experimental operation using the two methods described above, the adjustment efficiency can be compared, and the experimental operation with the highest adjustment efficiency can be selected as the final optimized experimental operation. This step can also involve performing a first and second adjustment on the experimental operation, then executing the un-adjusted and post-adjusted experimental operations multiple times, and analyzing the execution data to verify the adjustment efficiency. For example, after optimizing an experimental operation using the two adjustment methods described above, the un-optimized and post-optimized experimental operations can be executed multiple times, and the improved efficiency can be verified through execution data analysis.

[0099] Thus, step S4 has been completed. Figure 2 Optimization of the application layer.

[0100] Furthermore, the present invention can also automatically analyze the laboratory's business indicators to provide support for operational decisions, in which case the method further includes:

[0101] S5. Analyze the business metrics of the current experimental process in real time based on the data mapped to the experimental objects;

[0102] The business indicators may include at least one of the following: experimental efficiency, manual workload, and automation rate.

[0103] Experimental efficiency refers to the number of experiments completed per person per day, which can be calculated by dividing the number of experiments completed within a predetermined time period by the number of days.

[0104] Human workload can be calculated by dividing the time spent operating the equipment by the number of hours worked.

[0105] Automation rate can be measured by the ratio of the time spent on automatically performed experimental operations to the total time spent on all experimental operations.

[0106] Thus, step S5 has been completed. Figure 2 Metrics for the application layer.

[0107] Furthermore, the present invention can also predict the execution time required for a new experimental procedure based on data mapped onto the experimental object, providing support for the experiment. In this case, the method further includes:

[0108] S6. Collect the time consumption distribution of similar experimental tasks during execution; predict the execution time of the current experimental process based on the time consumption distribution.

[0109] Specifically, the current experimental process can be converted into multiple target experimental tasks. The time distribution of each target experimental task can be queried and the execution time of each target experimental task can be determined. Then the execution time of the current experimental process is the sum of the execution times of all target experimental tasks.

[0110] Furthermore, since the present invention also supports interaction with the user, the method further includes:

[0111] S7. Send the monitoring results of the current experimental process, the business indicators of the current experimental process, and the execution time of the current experimental process to the user.

[0112] The monitoring results may include: predictions of the feasibility of the response, predictions of the probability of anomalies in the experimental task, and monitoring results of data mapped onto the experimental objects.

[0113] Optionally, the monitoring results, business metrics, and execution time of the experimental process can be provided to users in the form of reports to support decision-making.

[0114] Thus, step S7 has been completed. Figure 2 Reports from the middle-level decision-making body.

[0115] Furthermore, the present invention also supports a virtual laboratory cockpit function to simulate the progress of the experimental process, and the method further includes:

[0116] S8. Simulate and demonstrate the execution effect before and after the optimization of the experimental process;

[0117] This step demonstrates to decision-makers the effects of optimizing the reaction and operational processes: for example, how much efficiency can be improved by changing the recrystallization operation; and how much efficiency can be improved by introducing automated weighing equipment in weighing.

[0118] Specifically, the system receives optimization metrics input by the user. These metrics may include: the optimization object and the optimization method. For example, the optimization object might be weighing operations, and the optimization method might be introducing automated weighing equipment. The execution effect before and after optimization is then determined based on both manual and automated weighing data.

[0119] S9. Receive the experimental support conditions input by the user, and find the target experimental process that matches the experimental support conditions based on the business indicators and execution time of each experimental process.

[0120] Among them, experimental support conditions may include: the manpower required for the experiment, the status of equipment, the time required for the experiment, etc.

[0121] S10. Receive the experimental expansion conditions input by the user, and find the experimental support conditions that match the experimental expansion conditions according to the business indicators of each experimental process.

[0122] Among them, experimental expansion conditions may include the expansion of experimental instruments.

[0123] Thus, steps S8 to S10 have been completed. Figure 2 A virtual laboratory cockpit for the middle-level decision-makers.

[0124] In summary, this invention divides the laboratory into multiple operational unit areas and creates a corresponding management shell for each area to control experiments and collect data, thereby achieving automated data acquisition. It converts the current experimental process into experimental tasks and operations, and maps the collected data to corresponding experimental objects, completing the connection between experimental data and specific experimental objects, and realizing business-level modeling. By creating and training a prediction model, the data mapped to the experimental objects is input into the prediction model to monitor the current experimental status. Finally, the current experimental process is optimized based on the results of the current experimental process and the data mapped to the experimental objects. Ultimately, a closed loop from data acquisition to experimental monitoring and optimization is established in the laboratory. Compared with existing technologies, this invention has at least the following beneficial effects:

[0125] 1. Create a corresponding management shell for each operation unit area to control experiments and collect data, thereby achieving automated data acquisition;

[0126] 2. By creating and training predictive models, the physical simulation modeling process of chemical reactions can be bypassed, thereby enabling the prediction and monitoring of experimental results and processes.

[0127] 3. No need for repeated experiments for optimization; it can automatically optimize the current experimental process based on the results of the current experimental process and the data mapped to the experimental objects, effectively saving manpower and financial resources.

[0128] 4. It can analyze laboratory indicators based on data mapped to experimental objects, and obtain information such as laboratory execution efficiency, experimental time, and reaction success rate, providing support for subsequent experiments.

[0129] 5. Based on the data mapped to the experimental objects and the prediction model, it can provide experimental monitoring results, experimental business indicators, and experimental execution time for decision-making reference.

[0130] 6. Able to simulate experiments based on experimental monitoring results, experimental business indicators, and experimental execution time, and estimate the support and expansion conditions required for the experiment.

[0131] Figure 3 This is a schematic diagram of the structural framework of a chemical experiment monitoring and optimization device based on digital twins according to the present invention, as shown below. Figure 3 As shown, the device includes:

[0132] The acquisition module 31 is used to divide the laboratory into multiple operating unit areas and create a corresponding management shell for each operating unit area for experimental control and data acquisition.

[0133] Mapping module 32 is used to convert the current experimental process into experimental tasks and experimental operations, and to map the collected data to the corresponding experimental objects; the experimental objects include experimental tasks and / or experimental operations.

[0134] Monitoring module 33 is used to create and train a prediction model, and input the data mapped to the experimental object into the prediction model to monitor the current experimental status.

[0135] The optimization module 34 is used to optimize the current experimental process based on the results of the current experimental process and the data mapped onto the experimental object.

[0136] In one specific embodiment, the mapping module 32 includes:

[0137] The creation module is used to create the mapping relationship between each instrument and the experimental object in the operation unit area;

[0138] The sub-mapping module is used to map the data collected from each operation unit area to the corresponding experimental object according to the mapping relationship.

[0139] Optionally, if the experimental object is an experimental task, and the experimental task includes a reaction, and the data mapped onto the experimental object is a reaction formula, then the monitoring module 33 is used to create and train a first model, input the reaction formula into the first model to predict and monitor the feasibility probability of the current reaction.

[0140] Optionally, if the experimental object is an experimental operation and the data mapped onto the experimental object is a time series of experimental operation parameters, then the monitoring module 33 is used to create and train a second model, input the time series of experimental operation parameters into the second model to predict and monitor the probability of an anomaly occurring in the current experimental task.

[0141] Optionally, the experimental object is an experimental operation, and the data mapped to the experimental object does not include the time series of experimental operation parameters; then the monitoring module 33 is used to predict the execution result of the current experimental task through the distribution of the historical execution results of the experimental operation.

[0142] In one example, the optimization module 34 includes:

[0143] The first optimization module is used to determine the results of the current experimental process and the parameters mapped to the experimental operation, adjust the parameters of the experimental process and / or experimental operation according to the preset reaction parameters and the process optimization algorithm, and obtain the adjusted experimental process execution results.

[0144] Alternatively, the second optimization module is used to determine the results of the current experimental process and the data mapped to the experimental operation, and to adjust the experimental operation based on the analysis results of the current experimental process results and experimental operation data.

[0145] Furthermore, the second optimization module includes:

[0146] The analysis and adjustment module is used to analyze the causal relationship between the current experimental process results and the data mapped to the experimental operations, and to make the first adjustment to the experimental operations.

[0147] And / or, the comparison and adjustment module is used to compare the results of multiple executions of the same experimental operation in multiple experimental processes, and to make a second adjustment to the experimental operation based on the execution results.

[0148] Furthermore, the device also includes:

[0149] The analysis and verification module is used to perform the experimental operation before and after the first and second adjustments multiple times, and to analyze and verify the adjustment efficiency based on the execution data.

[0150] The indicator analysis module is used to analyze the business indicators of the current experimental process in real time based on the data mapped to the experimental objects; the business indicators include at least one of: experimental efficiency, manual workload, and automation rate.

[0151] The first data acquisition module is used to collect the time distribution of similar experimental tasks during execution.

[0152] The first prediction module is used to predict the execution time of the current experimental process based on the time consumption distribution.

[0153] The sending module is used to send the monitoring results of the current experimental process, the business indicators of the current experimental process, and the execution time of the current experimental process to the user.

[0154] The demonstration module is used to simulate and showcase the execution effects before and after the optimization of the experimental process;

[0155] The first matching module is used to receive the experimental support conditions input by the user, and find the target experimental process that matches the experimental support conditions based on the business indicators and execution time of each experimental process.

[0156] The second matching module is used to receive the experimental expansion conditions input by the user, and to find experimental support conditions that match the experimental expansion conditions based on the business indicators of each experimental process.

[0157] Those skilled in the art will understand that the modules in the above-described device embodiments can be distributed throughout the device as described, or they can be modified accordingly and distributed in one or more devices different from the above embodiments. The modules in the above embodiments can be combined into one module, or they can be further divided into multiple sub-modules.

[0158] The following describes embodiments of the electronic device of the present invention, which can be considered as implementations of the physical form of the methods and apparatus embodiments of the present invention described above. Details described in the embodiments of the electronic device of the present invention should be considered as supplements to the methods or apparatus embodiments described above; details not disclosed in the embodiments of the electronic device of the present invention can be implemented with reference to the methods or apparatus embodiments described above.

[0159] Figure 4 This is a structural block diagram of an exemplary embodiment of an electronic device according to the present invention. Figure 4 The electronic device shown is merely an example and should not impose any limitation on the functionality and scope of use of the embodiments of the present invention.

[0160] like Figure 4As shown, the electronic device 400 of this exemplary embodiment is manifested in the form of a general data processing device. The components of the electronic device 400 may include, but are not limited to: at least one processing unit 410, at least one storage unit 420, a bus 430 connecting different electronic device components (including storage unit 420 and processing unit 410), a display unit 440, etc.

[0161] The storage unit 420 stores a computer-readable program, which may be source code or read-only program code. The program can be executed by the processing unit 410, causing the processing unit 410 to perform the steps of various embodiments of the present invention. For example, the processing unit 410 can perform actions such as... Figure 1 The steps are shown.

[0162] The storage unit 420 may include a readable medium in the form of a volatile storage unit, such as a random access memory (RAM) unit 4201 and / or a cache memory unit 4202, and may further include a read-only memory unit (ROM) unit 4203. The storage unit 420 may also include a program / utility 4204 having a set (at least one) of program modules 4205, such program modules 4205 including, but not limited to: operating electronic devices, one or more application programs, other program modules, and program data; each or some combination of these examples may include an implementation of a network environment.

[0163] Bus 430 can represent one or more of several types of bus structures, including a memory cell bus or memory cell controller, a peripheral bus, a graphics acceleration port, a processing unit, or a local bus using any of the various bus structures.

[0164] Electronic device 400 can also communicate with one or more external devices 300 (e.g., keyboard, display, network device, Bluetooth device, etc.), enabling users to interact with electronic device 400 via these external devices 400, and / or enabling electronic device 400 to communicate with one or more other data processing devices (e.g., router, modem, etc.). This communication can be made via input / output (I / O) interface 450, and also via network adapter 460 to one or more networks (e.g., local area network (LAN), wide area network (WAN), and / or public networks, such as the Internet). Network adapter 460 can communicate with other modules of electronic device 400 via bus 430. It should be understood that, although... Figure 4 As not shown, other hardware and / or software modules may be used in electronic device 400, including but not limited to: microcode, device drivers, redundancy processing units, external disk drive arrays, RAID electronics, tape drives, and data backup storage electronics.

[0165] Figure 5 This is a schematic diagram of a computer-readable medium embodiment of the present invention. Figure 5 As shown, the computer program can be stored on one or more computer-readable media. A computer-readable medium can be a readable signal medium or a readable storage medium. A readable storage medium can be, for example, but not limited to, an electronic device, apparatus, or device that is electrical, magnetic, optical, electromagnetic, infrared, or semiconductor, or any combination thereof. More specific examples of readable storage media (a non-exhaustive list) include: an electrical connection having one or more wires, a portable disk, a hard disk, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fiber, portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination thereof. When the computer program is executed by one or more data processing devices, it enables the computer-readable medium to implement the above-described method of the present invention, namely: dividing the laboratory into multiple operating unit areas and creating a corresponding management shell for each operating unit area for experimental control and data acquisition; converting the current experimental process into experimental tasks and experimental operations, and mapping the acquired data to corresponding experimental objects; the experimental objects include experimental tasks and / or experimental operations; creating and training a prediction model, inputting the data mapped to the experimental objects into the prediction model to monitor the current experimental status; and optimizing the current experimental process based on the results of the current experimental process and the data mapped to the experimental objects.

[0166] From the above description of the embodiments, those skilled in the art will readily understand that the exemplary embodiments described in this invention can be implemented by software or by combining software with necessary hardware. Therefore, the technical solutions according to the embodiments of this invention can be embodied in the form of a software product, which can be stored in a computer-readable storage medium (such as a CD-ROM, USB flash drive, portable hard drive, etc.) or on a network, including several instructions to cause a data processing device (such as a personal computer, server, or network device, etc.) to execute the methods described above according to this invention.

[0167] The computer-readable storage medium may include data signals propagated in baseband or as part of a carrier wave, carrying readable program code. Such propagated data signals may take various forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination thereof. The readable storage medium may also be any readable medium other than a readable storage medium, capable of transmitting, propagating, or transmitting programs for use by or in conjunction with an electronic device, apparatus, or device that executes instructions. The program code contained on the readable storage medium may be transmitted using any suitable medium, including but not limited to wireless, wired, optical fiber, RF, etc., or any suitable combination thereof.

[0168] Program code for performing the operations of this invention can be written in any combination of one or more programming languages, including object-oriented programming languages ​​such as Java and C++, and conventional procedural programming languages ​​such as C or similar languages. The program code can execute entirely on the user's computing device, partially on the user's device, as a standalone software package, partially on the user's computing device and partially on a remote computing device, or entirely on a remote computing device or server. In cases involving remote computing devices, the remote computing device can be connected to the user's computing device via any type of network, including a local area network (LAN) or a wide area network (WAN), or it can be connected to an external computing device (e.g., via the Internet using an Internet service provider).

[0169] In summary, the present invention can be implemented by methods, apparatus, electronic devices, or computer-readable media that execute computer programs. In practice, some or all of the functions of the present invention can be implemented using general-purpose data processing devices such as microprocessors or digital signal processors (DSPs).

[0170] The specific embodiments described above further illustrate the purpose, technical solution, and beneficial effects of the present invention. It should be understood that the present invention is not inherently related to any specific computer, virtual device, or electronic device, and various general-purpose devices can also implement the present invention. The above descriptions are merely specific embodiments of the present invention and are not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A method for optimizing chemical experiment monitoring based on digital twins, characterized in that, The method includes: The laboratory is divided into multiple operating unit areas, and a corresponding management shell is created for each operating unit area for experimental control and data acquisition. The current experimental process is converted into experimental tasks and operations, and the collected data is mapped to the corresponding experimental objects. This includes: creating a mapping relationship between each instrument and the experimental object in the operation unit area, and mapping the data collected in each operation unit area to the corresponding experimental object according to the mapping relationship; the experimental object includes experimental tasks and / or experimental operations. If the experimental object is an experimental task, and the experimental task includes a reaction, and the data mapped onto the experimental object is a reaction formula, then a first model is created and trained. The reaction formula is input into the first model to predict and monitor the feasibility probability of the current reaction. The first model is pre-trained using a large amount of reaction text and is used to determine the feasibility probability of the reaction. If the experimental object is an experimental operation, and the data mapped onto the experimental object is a time series of experimental operation parameters, then a second model is created and trained. The time series of experimental operation parameters is input into the second model to predict and monitor the probability of an anomaly occurring in the current experimental task. The second model is pre-trained based on the time series of historical experimental operation parameters and the results of the experimental task, and is used to predict the probability of an anomaly occurring in the experimental task based on the time series of experimental operation parameters. If the experimental object is an experimental operation, and the data mapped onto the experimental object does not include the time series of experimental operation parameters, then the execution result of the current experimental task is predicted based on the distribution of the historical execution results of the experimental operation. The current experimental procedure is optimized based on the results and data mapped to the experimental objects. This includes: determining the results of the current experimental procedure and the parameters mapped to the experimental operations, adjusting the parameters of the experimental procedure and / or experimental operations according to preset reaction parameters and a process optimization algorithm, and obtaining the adjusted experimental procedure execution results; or, determining the results of the current experimental procedure and the data mapped to the experimental operations, and adjusting the experimental operations based on the analysis results of the current experimental procedure results and experimental operation data.

2. The method according to claim 1, characterized in that, The adjustment of experimental operations based on the analysis results of the current experimental process and experimental operation data includes: Analyze the causal relationship between the current experimental results and the data mapped to the experimental operations, and make the first adjustment to the experimental operations; And / or, compare the results of multiple executions of the same experimental operation in multiple experimental procedures, and make a second adjustment to the experimental operation based on the execution results.

3. The method according to claim 2, characterized in that, The method further includes: After making the first and second adjustments to the experimental operation, the experimental operation before and after the adjustment was executed multiple times, and the execution data was analyzed to verify the efficiency of the adjustment.

4. The method according to claim 1, characterized in that, The method further includes: The business metrics of the current experimental process are analyzed in real time based on the data mapped to the experimental objects; the business metrics include at least one of the following: experimental efficiency, manual workload, and automation rate.

5. The method according to claim 4, characterized in that, The method further includes: Collect the time distribution of similar experimental tasks during execution; The execution time of the current experimental procedure is predicted based on the time consumption distribution.

6. The method according to claim 5, characterized in that, The method further includes: Send the monitoring results of the current experimental process, the business metrics of the current experimental process, and the execution time of the current experimental process to the user.

7. The method according to claim 5, characterized in that, The method further includes: The simulation demonstrates the execution effects before and after the optimization of the experimental process; Receive the experimental support conditions input by the user, and find the target experimental process that matches the experimental support conditions based on the business indicators and execution time of each experimental process. The system receives experimental expansion conditions input by the user and searches for experimental support conditions that match the experimental expansion conditions based on the business indicators of each experimental process.

8. A chemical experiment monitoring and optimization device based on digital twins, characterized in that, The device includes: The data acquisition module is used to divide the laboratory into multiple operating unit areas and create a corresponding management shell for each operating unit area for experimental control and data acquisition. The mapping module is used to convert the current experimental process into experimental tasks and operations, and to map the collected data to the corresponding experimental objects. It includes: a creation module, which creates a mapping relationship between each instrument and the experimental object in the operation unit area; and a sub-mapping module, which maps the data collected in each operation unit area to the corresponding experimental object according to the mapping relationship; the experimental object includes experimental tasks and / or experimental operations. The monitoring module is configured to: if the experimental object is an experimental task, and the experimental task includes a reaction, and the data mapped to the experimental object is a reaction formula, then the monitoring module creates and trains a first model, inputs the reaction formula into the first model to predict and monitor the feasibility probability of the current reaction; the first model is pre-trained using a large amount of reaction text to determine the feasibility probability of the reaction; if the experimental object is an experimental operation, and the data mapped to the experimental object is a time series of experimental operation parameters, then the monitoring module creates and trains a second model, inputs the time series of experimental operation parameters into the second model to predict and monitor the probability of an anomaly occurring in the current experimental task; the second model is pre-trained based on the time series of historical experimental operation parameters and the results of experimental tasks, and is used to predict the probability of an anomaly occurring in the experimental task based on the time series of experimental operation parameters; if the experimental object is an experimental operation, and the data mapped to the experimental object does not include the time series of experimental operation parameters, then the monitoring module predicts the execution result of the current experimental task based on the distribution of the historical execution results of the experimental operation. An optimization module is used to optimize the current experimental process based on the results of the current experimental process and the data mapped to the experimental object. This includes: a first optimization module, which determines the results of the current experimental process and the parameters mapped to the experimental operation, adjusts the parameters of the experimental process and / or the experimental operation according to preset reaction parameters and a process optimization algorithm, and obtains the adjusted experimental process execution result; or a second optimization module, which determines the results of the current experimental process and the data mapped to the experimental operation, and adjusts the experimental operation based on the analysis results of the current experimental process results and the experimental operation data.

9. An electronic device, comprising: processor; as well as A memory storing computer-executable instructions, which, when executed, cause the processor to perform the method according to any one of claims 1-7.

10. A computer-readable storage medium, wherein, The computer-readable storage medium stores one or more programs that, when executed by a processor, implement the method of any one of claims 1-7.

Images