Experimental process instruction arrangement method, device, equipment, medium and program product

Abstract

The invention discloses an experimental process instruction arrangement method, device and equipment, a medium and a program product, and the method comprises the steps: obtaining an operation description of a target object, and inputting the operation description into a target large language model; analyzing the operation description into a script instruction based on a target knowledge base by utilizing a target large language model; the target knowledge base comprises a professional knowledge base and a local knowledge base; performing parallel task scheduling on the script instruction to obtain a standardized instruction set; and controlling the experimental equipment to execute the target operation according to the standardized instruction set. According to the method, natural language operation description is analyzed into an executable instruction through the synergistic effect of a large model and a multi-dimensional knowledge base, and efficient semantic programming of an experimental process is achieved; through multi-device parallel task scheduling, the device utilization rate is effectively improved. Through deep fusion of a large model and an automation technology, experimental process instruction arrangement can be efficiently realized, and the method can be widely applied to the technical field of data processing.

Description

Technical Field

[0001] The present invention relates to the field of data processing technology, and in particular to an experimental process instruction arrangement method, device, equipment, medium and program product. Background Art

[0002] The evolution of laboratory automation technology has always revolved around improving the efficiency of experimental process control, but existing solutions still face multiple bottlenecks in the instruction generation stage, which restricts their in-depth application in complex scientific research scenarios.

[0003] Traditional pre-programmed script methods have long dominated this field, relying on engineers to manually write code to achieve device control. This method requires converting experimental steps one by one into the underlying communication protocol of a specific device, resulting in a long script development cycle and high maintenance costs. For example, modifying centrifugation parameters requires reconstructing the serial communication logic, which takes several hours. In addition, static scripts cannot perceive dynamic changes in the experimental environment (such as depletion of reagent residues and sudden equipment failures), have poor fault tolerance during execution, and have a significant risk of interruption. Summary of the Invention

[0004] Aiming to solve at least one of the technical problems in the related art to a certain extent, the present invention proposes an experimental process instruction arrangement method, device, equipment, medium and program product, which can efficiently implement experimental process instruction arrangement.

[0005] In one aspect, an embodiment of the present invention provides a method for arranging experimental process instructions, comprising:

[0006] Obtain the operation description of the target object and input the operation description into the target large language model;

[0007] Utilize the target large language model to parse the operation description into script instructions based on the target knowledge base; the target knowledge base includes the professional knowledge base and the local knowledge base;

[0008] Perform parallel task scheduling on script instructions to obtain a standardized instruction set;

[0009] Control the experimental equipment to perform target operations according to the standardized instruction set.

[0010] Optionally, the professional knowledge base includes experimental specifications, and the local knowledge base includes device information of each experimental device; parsing the operation description into script instructions based on the target knowledge base using the target large language model includes the following steps:

[0011] Using the target large language model to convert the operation description into standardized operation logic based on the experimental specifications of the professional knowledge base; the standardized operation logic includes at least one logical step;

[0012] Based on the device information in the local knowledge base, the experimental equipment associated with each logical step in the standardized operation logic is determined. The target large language model is used to parse each logical step into the atomic operation sequence of the corresponding experimental equipment according to the device information, and then integrate them to obtain script instructions.

[0013] Optionally, the method further comprises the following steps:

[0014] Construct parameter constraints for each functional operation of the experimental equipment based on the equipment information; the equipment information includes the function, parameters, constraints of each experimental equipment, and the inventory information of the experimental items corresponding to each experimental equipment;

[0015] The target large language model is used to perform security parameter verification on the atomic operation sequence corresponding to each logical step based on the parameter constraint range.

[0016] Optionally, the method further comprises the following steps:

[0017] Obtaining professional knowledge text in the target professional field uploaded by the target subject; the professional knowledge text includes experimental specification data and operation guide data as well as equipment operation data of each experimental equipment;

[0018] The professional knowledge text is uploaded to the target cloud corresponding to the target large language model, and the professional knowledge text is processed into experimental specification knowledge vectors through embedding operations on the target cloud, and then the professional knowledge base is built or updated based on the knowledge vectors.

[0019] Optionally, the method further comprises the following steps:

[0020] Acquire real-time device information of each experimental device based on a preset framework; each experimental device is connected to the preset framework through a preset interface;

[0021] Build or update the local knowledge base through prompt word engineering based on real-time device information.

[0022] Optionally, before the step of performing parallel task scheduling on the script instructions, the method further includes the following steps:

[0023] Transmitting the script instruction to the target object to obtain a verification instruction of the target object for the script instruction; the verification instruction includes a confirmation instruction or an adjustment instruction;

[0024] When the verification instruction is an adjustment instruction, a modification operation is performed on the script instruction in response to the adjustment instruction to update the script instruction.

[0025] Optionally, the script instructions include multiple task instructions; performing parallel task scheduling on the script instructions to obtain a standardized instruction set includes the following steps:

[0026] Store each task instruction in the script instruction as a single-step task in the parallel list;

[0027] The target large language model is used to generate the predecessor tasks and constraints of each single-step task based on the target knowledge base; wherein the predecessor tasks of a single-step task are associated with other single-step tasks in a parallel list.

[0028] Optionally, controlling the experimental equipment to perform a target operation according to a standardized instruction set includes the following steps:

[0029] Traverse the unfinished single-step tasks in the parallel list and add multiple single-step tasks that meet the preset conditions to the parallel task set;

[0030] The preset conditions include that the predecessor task corresponding to the single-step task is completed, and the state parameters of the experimental equipment corresponding to the single-step task meet the constraint conditions corresponding to the single-step task;

[0031] Based on the task instructions corresponding to each single-step task in the parallel task set, the experimental equipment corresponding to the single-step task is controlled to perform the target task operation;

[0032] The single-step task that completes the target task operation in the parallel list is marked as completed, and the step of traversing the unfinished single-step tasks in the parallel list is returned until all the single-step tasks in the parallel list are marked as completed.

[0033] On the other hand, an embodiment of the present invention provides an experimental process instruction arrangement device, including:

[0034] The first module is used to obtain the operation description of the target object and input the operation description into the target large language model;

[0035] The second module is used to parse the operation description into script instructions based on the target knowledge base using the target large language model; the target knowledge base includes a professional knowledge base and a local knowledge base;

[0036] The third module is used to perform parallel task scheduling on script instructions to obtain a standardized instruction set;

[0037] The fourth module is used to control the experimental equipment to perform target operations according to the standardized instruction set.

[0038] Optionally, the apparatus further includes a fifth module, specifically configured to perform the following operations:

[0039] Construct parameter constraints for each functional operation of the experimental equipment based on the equipment information; the equipment information includes the function, parameters, constraints of each experimental equipment, and the inventory information of the experimental items corresponding to each experimental equipment;

[0040] The target large language model is used to perform security parameter verification on the atomic operation sequence corresponding to each logical step based on the parameter constraint range.

[0041] Optionally, the apparatus further includes a sixth module, specifically configured to perform the following operations:

[0042] Obtaining professional knowledge text in the target professional field uploaded by the target subject; the professional knowledge text includes experimental specification data and operation guide data as well as equipment operation data of each experimental equipment;

[0043] The professional knowledge text is uploaded to the target cloud corresponding to the target large language model, and the professional knowledge text is processed into experimental specification knowledge vectors through embedding operations on the target cloud, and then the professional knowledge base is built or updated based on the knowledge vectors.

[0044] Optionally, the device further includes a seventh module, specifically configured to perform the following operations:

[0045] Acquire real-time device information of each experimental device based on a preset framework; each experimental device is connected to the preset framework through a preset interface;

[0046] Build or update the local knowledge base through prompt word engineering based on real-time device information.

[0047] Optionally, the apparatus further includes an eighth module, specifically configured to perform the following operations:

[0048] Transmitting the script instruction to the target object to obtain a verification instruction of the target object for the script instruction; the verification instruction includes a confirmation instruction or an adjustment instruction;

[0049] When the verification instruction is an adjustment instruction, a modification operation is performed on the script instruction in response to the adjustment instruction to update the script instruction.

[0050] On the other hand, an embodiment of the present invention provides an electronic device, including: a processor and a memory; the memory is used to store programs; the processor executes the program to implement the above-mentioned experimental process instruction arrangement method.

[0051] On the other hand, an embodiment of the present invention provides a computer storage medium storing a program executable by a processor. The program executable by the processor is used to implement the above-mentioned experimental process instruction arrangement method when executed by the processor.

[0052] On the other hand, an embodiment of the present invention provides a computer program product, which includes a computer program. When the computer program is executed by a processor, it is used to implement the above-mentioned experimental process instruction arrangement method.

[0053] The embodiment of the present invention obtains an operation description of a target object and inputs the operation description into a target large language model; utilizes the target large language model to parse the operation description into script instructions based on a target knowledge base; the target knowledge base includes a professional knowledge base and a local knowledge base; performs parallel task scheduling on the script instructions to obtain a standardized instruction set; and controls the experimental equipment to execute the target operation according to the standardized instruction set.

[0054] The embodiments of the present invention have the following beneficial effects: through the synergy of large models and multi-dimensional knowledge bases, natural language operation descriptions are parsed into executable instructions, enabling efficient semantic programming of experimental processes; and through multi-device parallel task scheduling, equipment utilization is effectively improved. Through the deep integration of large models and automation technology, the embodiments of the present invention solve the problems of low instruction generation efficiency and weak equipment collaboration in traditional laboratory processes, providing an efficient and reliable automation solution for scientific research scenarios. BRIEF DESCRIPTION OF THE DRAWINGS

[0055] The accompanying drawings are used to provide a further understanding of the technical solution of the present invention and constitute a part of the specification. Together with the embodiments of the present invention, they are used to explain the technical solution of the present invention and do not constitute a limitation to the technical solution of the present invention.

[0056] Figure 1 A schematic diagram of an implementation environment for arranging experimental process instructions provided by an embodiment of the present invention;

[0057] Figure 2 A flow chart of a method for arranging experimental flow instructions provided by an embodiment of the present invention;

[0058] Figure 3 A schematic diagram of the expanded flow of step S200 provided in an embodiment of the present invention;

[0059] Figure 4 A schematic diagram of an example file structure of a device file provided in an embodiment of the present invention;

[0060] Figure 5 A schematic diagram of the expanded flow of step S300 provided in an embodiment of the present invention;

[0061] Figure 6 A schematic diagram of the expanded flow of step S400 provided in an embodiment of the present invention;

[0062] Figure 7 A schematic diagram of an example of a principle architecture for implementing experimental process instruction arrangement provided by an embodiment of the present invention;

[0063] Figure 8 A schematic diagram of an example process of the experimental process instruction arrangement method provided by an embodiment of the present invention;

[0064] Figure 9 A schematic diagram of the structure of an experimental process instruction arrangement device provided by an embodiment of the present invention;

[0065] Figure 10 A schematic structural diagram of an electronic device provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0066] In order to make the purpose, technical solutions and advantages of the present invention more clearly understood, the present invention will be 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 used to explain the present invention and are not intended to limit the present invention.

[0067] It should be noted that although the system diagrams illustrate functional module divisions and the flowcharts illustrate a logical sequence, in certain circumstances, the steps shown or described may be performed in a sequence that differs from the module divisions in the system or the sequence in the flowcharts. The terms "first / S100," "second / S200," and the like in the specification, claims, and drawings are used to distinguish similar objects and are not necessarily intended to describe a specific sequence or precedence.

[0068] References to "embodiments" in this disclosure mean that a particular feature, structure, or characteristic described in connection with the embodiment may be included in at least one embodiment of the disclosure. The appearance of this phrase in various places in the specification does not necessarily refer to the same embodiment, nor does it constitute an independent or alternative embodiment that is mutually exclusive of other embodiments. It is understood, both explicitly and implicitly, by those skilled in the art that the embodiments described herein may be combined with other embodiments.

[0069] like Figure 1 FIG. 1 is a schematic diagram of an implementation environment provided by an embodiment of the present invention. Figure 1 , the implementation environment includes at least one terminal 102 and a server 101. The terminal 102 and the server 101 can be connected to the network in a wireless or wired manner to complete data transmission and exchange.

[0070] Server 101 can be an independent physical server, or a server cluster or distributed system composed of multiple physical servers. It can also be a cloud server that provides basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communications, middleware services, domain name services, security services, CDN (Content Delivery Network), as well as big data and artificial intelligence platforms.

[0071] In addition, server 101 can also be a node server in a blockchain network. Blockchain is a new application model of computer technologies such as distributed data storage, peer-to-peer transmission, consensus mechanism, and encryption algorithm.

[0072] The terminal 102 may be a smart phone, tablet computer, laptop computer, desktop computer, smart speaker, smart watch, etc., but is not limited thereto. The terminal 102 and the server 101 may be connected directly or indirectly via wired or wireless communication, which is not limited in this embodiment of the present invention.

[0073] Based on the example Figure 1 In the implementation environment shown, an embodiment of the present invention provides an experimental process instruction arrangement method. The following is explained using the experimental process instruction arrangement method applied in the terminal 102 as an example. It can be understood that the experimental process instruction arrangement method can also be applied to the server 101.

[0074] Reference Figure 2 , Figure 2 The flowchart of the experimental process instruction arrangement method applied to the terminal provided by the embodiment of the present invention, the execution subject of the experimental process instruction arrangement method can be any of the aforementioned computer devices (including servers or terminals). Figure 2 , the method comprises the following steps:

[0075] S100, obtaining an operation description of a target object, and inputting the operation description into a target large language model;

[0076] For example, in some specific implementations, a user's description of the experimental procedure is first obtained, and then the relevant description is input into a target large language model. In the subsequent step, semantic parsing technology is used to convert the user's natural language description (such as "three-replicate PCR test") into standardized operation logic. Specifically, the target large language model can be a model such as ChatGPT or DeepSeek, or it can also be a preset or pre-trained natural language processing model or other large language model.

[0077] Among them, in some preferred implementations, the embodiments of the present invention also propose a feedback mechanism for ambiguous instructions input by the user's operation description. When a large language model encounters a problem that it cannot understand, it must interact with the user to obtain accurate execution information.

[0078] S200, parsing the operation description into script instructions based on the target knowledge base using the target large language model;

[0079] Among them, the target knowledge base includes professional knowledge base and local knowledge base;

[0080] It should be noted that the professional knowledge base includes experimental specifications, and the local knowledge base includes the equipment information of each experimental equipment; in some embodiments, such as Figure 3 As shown, step S200 may include the following steps:

[0081] S201, using the target large language model to convert the operation description into standardized operation logic based on the experimental specifications of the professional knowledge base;

[0082] Wherein, the standardized operation logic includes at least one logical step;

[0083] For example, in some specific embodiments, the semantic parsing technology of the LLM large model (which is designed with a localized dynamic knowledge base, i.e., the target knowledge base) can be used to convert the user's natural language description (such as "three-repeat PCR detection") into standardized operation logic, and automatically associate the dependencies between experimental steps (such as sterilization takes precedence over sample processing) to ensure process compliance and safety.

[0084] It should also be noted that in some embodiments, the method may also include the following steps: obtaining professional knowledge text in the target professional field uploaded by the target object; the professional knowledge text includes experimental specification data and operation guide data as well as equipment operation data of each experimental equipment; uploading the professional knowledge text to the target cloud corresponding to the target large language model, and processing the professional knowledge text into a knowledge vector of the experimental specification through the target cloud using embedding operations, and then constructing or updating the professional knowledge base based on the knowledge vector.

[0085] For example, in some specific implementations, the professional knowledge base provides professional knowledge for artificial intelligence (such as the LLM large model) by integrating existing biochemical experimental standard process operations, professional field literature, laboratory safety manuals and other professional-related documents; in some specific application scenarios, the embodiment of the present invention integrates biomedical experimental standard specifications (SOPs), safety operation guidelines (MSDS) and equipment operation manuals to build a structured domain knowledge base (i.e., professional knowledge base), covering dimensions such as logical constraints of experimental steps, chemical reagent characteristics, and equipment safety rules. Specifically, the files of the professional knowledge base can be manually updated, and users can upload experimental process text files in a specific field to the cloud of the model provider, which then uses Embedding technology to automatically disassemble the text and integrate it into corresponding vectors to provide information to AI (such as the LLM large model).

[0086] S202. Determine the experimental equipment associated with each logical step in the standardized operation logic based on the equipment information in the local knowledge base, and use the target large language model to parse each logical step into the atomic operation sequence of the corresponding experimental equipment according to the equipment information, and then integrate them to obtain script instructions.

[0087] For example, in some specific implementations, the large model is further linked to the device information in the local knowledge base to achieve automated decomposition of intent into instructions: complex experimental processes (such as "high-throughput screening") are parsed into sequences of device-level atomic operations. Specifically, the local knowledge base is synchronized in real time based on the environmental status of each experimental device: dynamic tracking of reagent and consumable inventory, equipment operating status, and spatial coordinate information (such as the safety point of the robotic arm) ensures that the instruction generation strictly matches the physical environment.

[0088] In some specific application scenarios, all operable functions can be defined in the device file (i.e., device information) under the local knowledge base. Standard descriptive language is used to limit the scope of large language model generation. If the device file does not contain certain functions, the corresponding key call words will not be generated. The file structure example is as follows: Figure 4 shown; specifically:

[0089] The device directory stores the device function files of each module, in which the parameters of the corresponding functions are constrained. The following is an illustration with reference to a specific example (for illustration only and not to be construed as limiting the embodiments of the present invention):

[0090] For some functions, such as the robot arm movement function, the following parameters are required (coord, speed, roughly):

[0091] "MoveJ":{

[0092] "description":"MoveJ can control the four-axis robot arm to move to the specified position."

[0093] "required_params":{

[0094] "coord":{

[0095] "type":"variant",

[0096] "example":"$p",

[0097] "description":"Variable name, the p after $ is the actual variable name, the corresponding name needs to be found and replaced in the variable definition area"

[0098] },

[0099] "speed":{

[0100] "type":"range",

[0101] "value_type":"float",

[0102] "min":5,

[0103] "max":200,

[0104] "unit":"cm / s",

[0105] "description":"The speed of the robot arm movement, the default is 100cm / s"

[0106] },

[0107] "roughly":{

[0108] "type":"enum",

[0109] "enum_values":[0,1],

[0110] "description":"Specify the point speed, the default is 0"

[0111] }

[0112] },

[0113] }

[0114] In summary, the above description allows parameters to be restricted to a certain range. In view of this, in some embodiments, the method may further include the following steps: constructing a parameter constraint range for each functional operation of the experimental equipment based on device information; the device information includes the function, parameters, and constraints of each experimental equipment, as well as inventory information of experimental items corresponding to each experimental equipment; and utilizing the target large language model to perform safety parameter verification on the atomic operation sequence corresponding to each logical step based on the parameter constraint range.

[0115] For example, in some specific implementations, the generated instruction may be subjected to security parameter verification to ensure that the device function parameters it calls are kept within the constraints of local knowledge about the corresponding device to avoid potential security risks.

[0116] It should also be noted that, in some embodiments, the method may further include the following steps: obtaining real-time device information of each experimental device based on a preset framework; connecting each experimental device to the preset framework through a preset interface; and constructing or updating a local knowledge base through a prompt word project based on the real-time device information.

[0117] For example, in some specific implementations, the local knowledge base (knowledge) stores relevant information of all available equipment of the automation platform, including functions, parameters, restrictions, etc., as well as the inventory, location and key point information of items. Through standardized prompt word engineering, this information is converted into standard prompt words and passed to the large model, so that the large model can grasp the existing platform information. Specifically, the local knowledge base can realize dynamic adaptation based on the real-time equipment information of the experimental equipment. Dynamic adaptation means that the local knowledge base file can be dynamically adjusted according to environmental changes, including the inventory of reagents and consumables, adding or updating equipment modules and available key points, etc.

[0118] In some specific application scenarios, the localized dynamic knowledge base (Knowledge architecture) designed in the embodiments of the present invention includes the following core capabilities:

[0119] ① Plug-and-play device management: This supports the rapid integration and deployment of multi-brand devices through a standardized description framework. Specifically, the architecture here is built on a third-party platform, which opens up the interfaces of each module and uses the MQTT IoT to control all modular devices. For example, when adding a new modular device (i.e., experimental equipment), its functions must first be integrated into the third-party platform, and the corresponding functional interfaces must be opened. These functions can then be uniformly called using AI-generated instructions. (Subsequent control of individual experimental devices can also be achieved based on this architecture.)

[0120] ② Real-time synchronization of environmental status: Dynamically track the inventory of reagents and consumables, equipment operating status, and spatial coordinate information (such as the safety point of the robotic arm) to ensure that command generation strictly matches the physical environment;

[0121] ③Interactive teaching support: allows users to dynamically calibrate or update key parameters (such as pipetting target position) through a visual interface to achieve zero-code environment configuration.

[0122] In some specific application scenarios, step S300 mainly depends on the generation capability of a large language model. Among them, the embodiment of the present invention provides the AI ​​(such as the LLM large model) with the functional description of the existing platform equipment (i.e., the constituent elements of the target knowledge base) in the form of prompt words in advance. For example, the platform includes a PCR module device. In this process, a standardized prompt word project can be used in advance to convey the descriptive language such as "open (opening the lid)", "start (starting)", "setParam (setting working parameters)" of the PCR module to the AI, so that it can master the operations that can be performed by the PCR device and its restrictions. Subsequently, the large language model will generate script instructions that meet the platform requirements and follow a specific format based on the knowledge provided by the professional knowledge base and the locally achievable functions.

[0123] S300, performing parallel task scheduling on the script instructions to obtain a standardized instruction set;

[0124] In some embodiments, before executing step S300, the method may further include the following steps: transmitting the script instruction to the target object to obtain the target object's verification instruction for the script instruction; the verification instruction includes a confirmation instruction or an adjustment instruction; when the verification instruction is an adjustment instruction, modifying the script instruction in response to the adjustment instruction to update the script instruction.

[0125] For example, in some specific implementations, the command parameters of the footstep instructions can also be verified and feedback can be provided to the user. If the user confirms that the process is correct, the process can be executed. Specifically, for script instructions generated by a large language model (i.e., the target large language model), the large language model can be used to further implement parameter verification. In addition, the user's feedback can be combined to make a final decision on whether the generated instructions are correct and suitable for execution.

[0126] It should be noted that the script instruction includes multiple task instructions; in some embodiments, such as Figure 5 As shown, step S300 may include the following steps: S301, storing each task instruction in the script instruction as a single-step task in a parallel list; S302, using the target large language model to generate the predecessor task and constraint conditions of each single-step task based on the target knowledge base; wherein the predecessor task of a single-step task is associated with other single-step tasks in the parallel list.

[0127] For example, in some specific implementations, task instructions can be stored in a parallel list. The large model then generates each instruction while adding the necessary prerequisite tasks and constraints to each single-step task. Specifically, each single-step task instruction in the LLM-generated instructions can be parsed and stored in a parallel list, which primarily includes the prerequisite tasks and constraints for the single-step task.

[0128] S400: Control the experimental equipment to execute the target operation according to the standardized instruction set.

[0129] It should be noted that, in some embodiments, Figure 6As shown, step S400 may include the following steps: S401, traverse the unfinished single-step tasks in the parallel list, and add multiple single-step tasks that meet the preset conditions to the parallel task set; wherein the preset conditions include that the predecessor task corresponding to the single-step task is completed, and the state parameters of the experimental equipment corresponding to the single-step task meet the constraint conditions corresponding to the single-step task; S402, based on the task instructions corresponding to each single-step task in the parallel task set, control the experimental equipment corresponding to the single-step task to perform the target task operation; S403, mark the single-step task that completes the target task operation in the parallel list as completed, and return to the step of traversing the unfinished single-step tasks in the parallel list until all single-step tasks in the parallel list are marked as completed. In some extreme cases, two single-step tasks may be executed by the same functional component of the same experimental equipment. Therefore, in some preferred implementations, the condition that the execution components of each single-step task added to the parallel task set are different can also be set in the preset conditions. Specifically, it should be noted that in some possible implementation methods, when a single-step task loops through the state parameters of its corresponding experimental equipment for multiple times (such as 3 times, which can be adjusted according to actual needs) and none of them meet the corresponding constraints (for example, the experimental materials do not meet the experimental requirements), the target large language model can be called to further generate the equipment state adjustment task corresponding to the single-step task as a predecessor task (the predecessor task may not be associated with other single-step tasks in the parallel list. If the corresponding experimental materials do not meet the experimental requirements, a material replenishment task can be generated to control the experimental equipment to replenish materials first). In addition, direct feedback can also be given to the target object to manually adjust the experimental equipment to meet the constraints.

[0130] For example, in some specific implementations, single-step tasks whose predecessors have been completed and whose constraints allow them can be executed synchronously to maximize experimental efficiency. Specifically, after executing single-step tasks whose predecessors have been completed and are not subject to current constraints in parallel, they are further marked after completion and then retrieved and executed in a loop, thereby achieving high experimental execution efficiency.

[0131] In order to explain the principle of the technical solution of the present invention in detail, the overall process of the present invention is described below in combination with some specific embodiments. It is easy to understand that the following is an explanation of the technical principle of the present invention and cannot be regarded as a limitation of the present invention.

[0132] First of all, it should be noted that the main problems in the related technologies of laboratory automation are as follows:

[0133] Robotic process automation (RPA) tools attempt to simplify process design by recording user interface operations, but they have significant limitations. Recorded actions are tightly bound to device models, making them difficult to reuse across platforms. They also only record superficial operations (such as button clicks) and fail to parse the physical relationship between experimental intent and device parameters. For example, "Centrifuge at 3000 rpm for 5 minutes to remove sediment" requires manual mapping of the speed and centrifugal force. The system lacks automated conversion capabilities, leaving operational accuracy reliant on manual experience.

[0134] Initial natural language processing (NLP) solutions explored converting command text into device calls, but were limited by the rigid architecture of the rules engine. Their ability to parse specialized terminology (such as "Suzuki coupling reaction under anaerobic conditions") was limited, requiring the pre-definition of numerous keyword-to-device action mapping rules and resulting in poor scalability. The integration of new devices necessitated a redesign of the syntax parsing logic, and development costs grew linearly with the scale of the device fleet.

[0135] Large language model (LLM)-driven approaches (such as ChatGPT) have recently been introduced into this field, generating control code through natural language interaction, lowering the operational barrier. However, their actual deployment still faces severe challenges: the generated code requires manual review for safety, making it difficult to directly use for closed-loop control; it lacks the ability to dynamically integrate with real-time sensor data (such as incubator temperature and robotic arm posture), making it impossible to optimize instruction sequences based on environmental changes; and its output is mostly high-level task descriptions that need to be broken down into device-level atomic operations, limiting the degree of automation.

[0136] In summary, existing technology systems suffer from common shortcomings in command generation accuracy, adaptability to dynamic environments, and efficiency in disassembling complex processes. The heavy reliance on manual labor from experimental design to command deployment makes full process automation difficult in cutting-edge fields like high-throughput screening and synthetic biology. A new technology framework is urgently needed that can deeply integrate domain knowledge, perceive environmental conditions in real time, and automatically generate reliable commands.

[0137] In this regard, the following technical problems need to be urgently solved for laboratory automation technology:

[0138] 1. Lack of real-time semantic mapping between dynamic experimental environments and device capabilities:

[0139] Traditional instruction generation systems rely on static device configurations and manually maintained environmental parameters, making them difficult to cope with dynamic changes during the experiment process:

[0140] ① Inefficient management of device heterogeneity: Adding or replacing new devices requires redesigning interface adaptation logic, resulting in a development cycle lasting several weeks.

[0141] ② Delayed environmental status perception: Key information such as reagent and consumable inventory (e.g., buffer volume) and equipment real-time status (e.g., centrifuge load rate) is not synchronized to the instruction generation logic in real time, resulting in a disconnect between the script and the physical environment (e.g., performing a subpackaging operation using depleted reagents).

[0142] ③ Weak interactive teaching support: Key operation points (such as robotic arm grabbing coordinates and detector sampling ports) rely on manual entry by engineers, lacking a unified management mechanism, making it difficult to quickly respond to the needs of experimental plan iteration.

[0143] 2. Insufficient automated decomposition and parallel optimization capabilities for complex experimental logic:

[0144] Existing solutions rely on manual experience to generate instructions for multi-device collaborative experiments, which has the following systemic flaws:

[0145] ① Decoupling of domain knowledge and equipment capabilities: Experimental processes (e.g., "cell lysis → nucleic acid extraction → qPCR detection") need to be manually disassembled into device-level atomic operations, lacking automated logic verification based on professional knowledge (e.g., biosafety regulations, thermal cycling parameters);

[0146] ② Rigid parallel scheduling strategy: Task allocation relies on fixed rules (such as sequential execution), and it is impossible to dynamically optimize instruction timing based on the real-time load of the equipment (such as pipette channel occupancy and incubator preheating progress), resulting in a high equipment idle rate.

[0147] ③ Lack of physical constraint pre-verification: The generated instructions do not automatically verify the device capability boundaries (such as whether the pipette range matches the target volume, whether the robot arm path collides), and the risk of triggering parameter out of bounds during the execution phase is high.

[0148] In view of this, the present invention has constructed a technical solution for intelligent experimental instruction generation. Specifically, the embodiment of the present invention realizes the automation of the entire process from experimental intention description to executable instructions through a layered architecture. The solution includes three core modules, forming a closed-loop technical system of "knowledge integration - dynamic adaptation - intelligent compilation":

[0149] 1. Multimodal knowledge fusion and semantic understanding:

[0150] The system integrates standard procedures (SOPs), safety data sheets (MSDS), and equipment manuals for biomedical experiments to build a structured domain knowledge base covering logical constraints on experimental procedures, chemical reagent characteristics, and equipment safety rules. Using semantic parsing technology, it converts user natural language descriptions (e.g., "triplicate PCR tests") into standardized operational logic and automatically associates dependencies between experimental steps (e.g., sterilization takes precedence over sample processing) to ensure process compliance and safety.

[0151] 2. Dynamic environment perception and device adaptation:

[0152] Design a localized dynamic knowledge base (Knowledge architecture), including the following core capabilities:

[0153] ① Plug-and-play device management: supports rapid access and deployment of multi-brand devices through a standardized description framework;

[0154] ② Real-time synchronization of environmental status: Dynamically track the inventory of reagents and consumables, equipment operating status, and spatial coordinate information (such as the safety point of the robotic arm) to ensure that command generation strictly matches the physical environment;

[0155] ③Interactive teaching support: allows users to dynamically calibrate or update key parameters (such as pipetting target position) through a visual interface, achieving zero-code environment configuration.

[0156] 3. Intelligent compilation and optimized output:

[0157] Build an AI-driven instruction generation engine based on a dynamic knowledge base to implement the following core functions:

[0158] ① Automated decomposition from intent to instructions: parse complex experimental processes (such as "high-throughput screening") into device-level atomic operation sequences and pre-verify resource feasibility (such as reagent remaining amount and equipment load);

[0159] ② Parallel task scheduling optimization: Dynamically plan execution paths based on real-time device status (such as incubator preheating progress and pipette channel occupancy) to maximize multi-device collaborative efficiency;

[0160] ③ Standardized instruction output: Generate a JSON instruction set that complies with the interface specifications of the third-party execution platform, retain interpretable semantic labels (such as "sample transferred to well plate B3") and device-independent parameter mapping to ensure cross-platform compatibility.

[0161] In some specific application scenarios, such as Figure 7 and Figure 8 As shown, the embodiment of the present invention can be implemented through the following process:

[0162] First, the professional knowledge base provides AI with specialized knowledge by integrating existing biochemical experimental standard procedures, professional field literature, laboratory safety manuals, and other relevant documents. The local "knowledge base" stores information about all available equipment on the automation platform, including functions, parameters, restrictions, and item inventory, location, and key points. Through standardized prompt word engineering, this information is converted into standard prompt words and passed to the large model, allowing it to grasp the existing platform information.

[0163] Furthermore, intelligent compilation means that according to the experimental requirements proposed by the user, the large model generates experimental process script instructions in a specified format. The system will first perform security parameter verification on the generated instructions to ensure that the device function parameters it calls are kept within the constraints of the local knowledge about the corresponding equipment to avoid potential safety hazards; then the task instructions are stored in a parallel list. While the large model generates each instruction, the necessary predecessor tasks and constraints will be added to each single-step task. The system will synchronously execute those single-step tasks whose predecessor tasks have been completed and are allowed by the current constraints, so as to maximize the experimental efficiency.

[0164] Specifically, the main process of the technical solution of the embodiment of the present invention can be implemented as follows: the user puts forward requirements → the big model generates script instructions in a specific format based on the professional knowledge base and the available functions of the equipment in the local knowledge base → security parameter verification → parallel engine drives task execution.

[0165] In summary, the core innovation of the embodiments of the present invention lies in the construction of an instruction generation system that deeply integrates dynamic perception and intelligent scheduling, and achieves precise control of the experimental process and efficient resource utilization through a two-way technical closed loop. The specific implementation is as follows:

[0166] 1. Dynamically aware local knowledge architecture:

[0167] A real-time updated equipment-material database (knowledge) is constructed through structured JSON Schema, dynamically integrating equipment capability parameters, the spatial distribution of reagents and consumables, and key points of user teaching into a unified data model. This architecture supports plug-and-play expansion of equipment - new equipment only needs to register a standardized description file to automatically connect to the system. At the same time, RFID scanning and machine vision are used to track material consumption (such as PBS buffer remaining amount) and equipment operating status (such as centrifuge rotor load rate) in real time to ensure that the generated instruction script strictly matches the current physical environment and avoid resource conflicts and execution failure risks. The generated standardized JSON instructions are seamlessly connected to third-party execution platforms through interpretable semantic tags and device-independent interface design, realizing seamless collaboration of cross-brand equipment workflows.

[0168] 2. Parallel scheduling engine for real-time compilation:

[0169] Dynamically optimize the timing logic and resource allocation of instruction sequences based on the real-time status of the equipment (including load rate and availability) and experimental constraints. Intelligently decompose task chains to synchronize equipment operations that can be executed in parallel (such as incubator preheating and sample transfer) to shorten the entire experimental cycle. During operation, monitor equipment failures or environmental fluctuations (such as temperature deviations) and automatically initiate backup plans (such as switching to backup pipetting channels). Dynamically allocate tasks based on equipment performance to avoid overloading individual equipment and extend the life of the hardware.

[0170] For example, in some specific implementations, taking the safe path driven robotic arm operation as an example, the technical solution of the present invention is compared with the existing solution to illustrate:

[0171] 1. Implementation of the present invention:

[0172] User input: Transfer sample from reservoir A to assay plate B.

[0173] System process:

[0174] ①The dynamic knowledge base calls the device library (device) to query the robot arm's callable interfaces and path restriction rules; ②The coordinate library (position) provides pre-calibrated points ($Grasp_A, $Place_B);

[0175] ③ The material warehouse (object) verifies that the sample balance of liquid storage rack A is sufficient;

[0176] The AI ​​engine generates instructions and inserts a safe fallback path ($Air_Safe).

[0177] Generate instructions (brief description):

[0178] {

[0179] "work_info":[

[0180] {

[0181] "data":{

[0182] "actions":[

[0183] {

[0184] "DeviceID":"99301",

[0185] "Function":"Move",

[0186] "Set":{

[0187] "point":"$Grasp_A",

[0188] "mode":"1",

[0189] "speed":"30",

[0190] "acc":"30",

[0191] "rough":"-1"

[0192] }}]},

[0193] "name":"The robotic arm moves to the position to grab sample A"

[0194] },{

[0195] "data":{

[0196] "actions":[

[0197] {

[0198] "DeviceID":"14001",

[0199] "Function":"OpenGrip",

[0200] "Set":{

[0201] "GripPos":"0"

[0202] }}]},

[0203] "name":"Gripper closing action"

[0204] },{

[0205] "data":{

[0206] "actions":[

[0207] {

[0208] "DeviceID":"99301",

[0209] "Function":"Move",

[0210] "Set":{

[0211] "point":"$Air_Safe",

[0212] "mode":"1",

[0213] "speed":"30",

[0214] "acc":"30",

[0215] "rough":"-1"

[0216] }}]},

[0217] "name":"The robotic arm moves to a safe position in the air"

[0218] },{

[0219] "data":{

[0220] "actions":[

[0221] {

[0222] "DeviceID":"99301",

[0223] "Function":"Move",

[0224] "Set":{

[0225] "point":"$Place_B",

[0226] "mode":"1",

[0227] "speed":"30",

[0228] "acc":"30",

[0229] "rough":"-1"

[0230] }}]},

[0231] "name":"The robotic arm moves to the placement area B"

[0232] },{

[0233] "data":{

[0234] "actions":[

[0235] {

[0236] "DeviceID":"14001",

[0237] "Function":"OpenGrip",

[0238] "Set":{

[0239] "GripPos":"20"

[0240] }}]},

[0241] "name":"Gripper opening action"

[0242] },{

[0243] "data":{

[0244] "actions":[

[0245] {

[0246] "DeviceID":"99301",

[0247] "Function":"Move",

[0248] "Set":{

[0249] "point":"$Air_Safe",

[0250] "mode":"1",

[0251] "speed":"30",

[0252] "acc":"30",

[0253] "rough":"-1"

[0254] }}]},

[0255] "name":"The robotic arm moves to a safe position in the air"

[0256] }

[0257] ],

[0258] "var_info":["$Air_Safe","$Grasp_A","$Place_B"]

[0259] }

[0260] 2. Comparative Example: Traditional Pre-Programming Script Method:

[0261] Engineers manually wrote Python scripts to control the movement and grasping of the robotic arm, as shown in (brief description):

[0262] robot.move_to("Grasp_A")#hard-coded coordinates

[0263] robot.grasp()

[0264] robot.move_to("Place_B")

[0265] 3. Conclusion:

[0266] Avoid safety verification steps that may be missed during manual operation (such as moving directly without rolling back);

[0267] The instruction generation time is greatly shortened and the risk of path conflict is reduced to zero.

[0268] In summary, the embodiments of the present invention achieve significant technological advancement and practical value at multiple levels through the intelligent experimental instruction compilation system. The core beneficial effects can be summarized as follows:

[0269] 1. Improve the level of experimental process automation: Rapidly convert experimental plans from natural language descriptions to device-executable instructions, significantly shortening the experimental preparation cycle; support dynamic scheduling of multi-device collaboration, significantly improving laboratory resource utilization and experimental throughput.

[0270] 2. Enhance the accuracy and reliability of command generation: Through semantic parsing technology enhanced by domain knowledge, it accurately understands complex experimental requirements and avoids misoperations caused by semantic ambiguity. It also automatically verifies the physical limitations and safety rules of the equipment during the command generation stage to ensure the feasibility of the operation plan.

[0271] 3. Break through the limitations of device heterogeneity: It can generate specific JSON format text according to the rules of different platforms, thereby realizing the rapid transplantation and reuse of experimental plans; it supports the plug-and-play expansion function of new devices, greatly reducing the cost of system upgrades and maintenance.

[0272] On the other hand, Figure 9 As shown, an embodiment of the present invention provides an experimental process instruction arrangement device 900, which may include:

[0273] The first module 901 is used to obtain an operation description of a target object and input the operation description into a target large language model;

[0274] The second module 902 is configured to parse the operation description into script instructions based on a target knowledge base using a target large language model; the target knowledge base includes a professional knowledge base and a local knowledge base;

[0275] The third module 903 is used to perform parallel task scheduling on the script instructions to obtain a standardized instruction set;

[0276] The fourth module 904 is used to control the experimental equipment to perform the target operation according to the standardized instruction set.

[0277] In some embodiments, the apparatus may further include a fifth module, specifically configured to perform the following operations:

[0278] Construct parameter constraints for each functional operation of the experimental equipment based on the equipment information; the equipment information includes the function, parameters, constraints of each experimental equipment, and the inventory information of the experimental items corresponding to each experimental equipment;

[0279] The target large language model is used to perform security parameter verification on the atomic operation sequence corresponding to each logical step based on the parameter constraint range.

[0280] In some embodiments, the apparatus may further include a sixth module, specifically configured to perform the following operations:

[0281] Obtaining professional knowledge text in the target professional field uploaded by the target subject; the professional knowledge text includes experimental specification data and operation guide data as well as equipment operation data of each experimental equipment;

[0282] The professional knowledge text is uploaded to the target cloud corresponding to the target large language model, and the professional knowledge text is processed into experimental specification knowledge vectors through embedding operations on the target cloud, and then the professional knowledge base is built or updated based on the knowledge vectors.

[0283] In some embodiments, the apparatus may further include a seventh module, specifically configured to perform the following operations:

[0284] Acquire real-time device information of each experimental device based on a preset framework; each experimental device is connected to the preset framework through a preset interface;

[0285] Build or update the local knowledge base through prompt word engineering based on real-time device information.

[0286] In some embodiments, the apparatus may further include an eighth module, specifically configured to perform the following operations:

[0287] Transmitting the script instruction to the target object to obtain a verification instruction of the target object for the script instruction; the verification instruction includes a confirmation instruction or an adjustment instruction;

[0288] When the verification instruction is an adjustment instruction, a modification operation is performed on the script instruction in response to the adjustment instruction to update the script instruction.

[0289] The contents of the method embodiments of the present invention are all applicable to the device embodiments. The functions specifically implemented by the device embodiments are the same as those of the above method embodiments, and the beneficial effects achieved are also the same as those achieved by the above method embodiments.

[0290] In another aspect, an embodiment of the present invention further provides an electronic device comprising a memory and a processor, wherein the memory stores a computer program, and the processor implements the sensitive information protection method described above when executing the computer program. The electronic device can be any intelligent terminal, including a tablet computer and an in-vehicle computer.

[0291] It can be understood that the contents of the above method embodiments are applicable to the present device embodiments, the functions specifically implemented by the present device embodiments are the same as those of the above method embodiments, and the beneficial effects achieved are also the same as those achieved by the above method embodiments.

[0292] like Figure 10 As shown, Figure 10 A specific example of the hardware structure of an electronic device 1000 according to an embodiment is shown. The electronic device 1000 includes:

[0293] The processor 1001 may be implemented as a general-purpose CPU (Central Processing Unit), a microprocessor, an application-specific integrated circuit (ASIC), or one or more integrated circuits, and is configured to execute relevant programs to implement the technical solutions provided by the embodiments of the present invention.

[0294] The memory 1002 can be implemented in the form of a read-only memory (ROM), a static storage device, a dynamic storage device, or a random access memory (RAM). The memory 1002 can store an operating system and other application programs. When the technical solutions provided in the embodiments of this specification are implemented through software or firmware, the relevant program codes are stored in the memory 1002 and are called by the processor 1001 to execute the experimental process instruction arrangement method of the embodiments of the present invention.

[0295] Input / output interface 1003, used to implement information input and output;

[0296] Communication interface 1004, used to implement communication interaction between this device and other devices, which can be achieved through wired means (such as USB, network cable, etc.) or wireless means (such as mobile network, WiFi, Bluetooth, etc.);

[0297] Bus 1005 , which transmits information between various components of the device (e.g., processor 1001 , memory 1002 , input / output interface 1003 , and communication interface 1004 );

[0298] The processor 1001 , the memory 1002 , the input / output interface 1003 and the communication interface 1004 are connected to each other in communication within the device via the bus 1005 .

[0299] The electronic device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate, i.e., they may be located in one location or distributed across multiple network units. Some or all of these modules may be selected based on actual needs to achieve the objectives of this embodiment.

[0300] The contents of the method embodiments of the present invention are all applicable to the electronic device embodiments. The functions specifically implemented by the electronic device embodiments are the same as those of the above method embodiments, and the beneficial effects achieved are also the same as those achieved by the above method embodiments.

[0301] Another aspect of an embodiment of the present invention further provides a computer-readable storage medium, wherein the storage medium stores a program, and the program is executed by a processor to implement the above method.

[0302] It should be noted that the computer-readable medium shown in the embodiments of the present invention may be a computer-readable signal medium or a computer-readable storage medium, or any combination thereof. The computer-readable storage medium may be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, device, or component, or any combination thereof. More specific examples of computer-readable storage media may include, but are not limited to: an electrical connection with one or more wires, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM), a flash memory, an optical fiber, a portable compact disc read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination thereof. In the present invention, a computer-readable storage medium may be any tangible medium containing or storing a program that can be used by or in conjunction with an instruction execution system, device, or component. In the present invention, a computer-readable signal medium may include a data signal propagated in baseband or as part of a carrier wave, which carries computer-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. A computer-readable signal medium may also be any computer-readable medium other than a computer-readable storage medium that can transmit, propagate, or transport a program for use by or in connection with an instruction execution system, apparatus, or device. Program code embodied on a computer-readable medium may be transmitted using any suitable medium, including but not limited to wireless, wired, or any suitable combination thereof.

[0303] The contents of the method embodiments of the present invention are all applicable to the computer-readable storage medium embodiments. The functions specifically implemented by the computer-readable storage medium embodiments are the same as those of the above method embodiments, and the beneficial effects achieved are also the same as those achieved by the above method embodiments.

[0304] The present invention also discloses a computer program product or computer program, which includes computer instructions stored in a computer-readable storage medium. A processor of a computer device can read the computer instructions from the computer-readable storage medium and execute the computer instructions, causing the computer device to perform the above method.

[0305] The flowcharts and block diagrams in the accompanying drawings illustrate the possible implementation architecture, functions and operations of the systems, methods and computer program products according to various embodiments of the present invention. In this regard, each box in the flowchart or block diagram can represent a module, program segment, or a part of code, and the above-mentioned module, program segment, or a part of code contains one or more executable instructions for implementing the specified logical function. It should also be noted that in some alternative implementations, the functions marked in the box can also occur in an order different from that marked in the accompanying drawings. For example, two boxes represented in succession can actually be executed substantially in parallel, and they can sometimes be executed in the opposite order, depending on the functions involved. It should also be noted that each box in the block diagram or flowchart, and the combination of boxes in the block diagram or flowchart, can be implemented with a dedicated hardware-based system that performs the specified function or operation, or can be implemented with a combination of dedicated hardware and computer instructions.

[0306] It should be noted that although several modules of the device for action execution are mentioned in the above detailed description, this division is not mandatory. In fact, according to an embodiment of the present invention, the features and functions of two or more modules or units described above can be concretized in one module or unit. Conversely, the features and functions of one module or unit described above can be further divided into multiple modules or units to be concretized.

[0307] Through the description of the above embodiments, it is easy for those skilled in the art to understand that the example embodiments described herein can be implemented by software or by combining software with necessary hardware. Therefore, the technical solution according to the embodiments of the present invention can be embodied in the form of a software product, which can be stored in a non-volatile storage medium (which can be a CD to ROM, a USB flash drive, a mobile hard disk, etc.) or on a network, and includes a number of instructions to enable a computing device (which can be a personal computer, a server, a touch terminal, or a network device, etc.) to execute the method according to the embodiments of the present invention.

[0308] In some optional embodiments, the function / operation mentioned in the block diagram may not occur in the order mentioned in the operation diagram. For example, depending on the function / operation involved, the two boxes shown in succession can actually be executed substantially simultaneously or the boxes can sometimes be executed in reverse order. In addition, the embodiment presented and described in the flow chart of the present invention is provided in an exemplary manner for the purpose of providing a more comprehensive understanding of the technology. The disclosed method is not limited to the operation and logic flow presented by the present invention. Optional embodiments are contemplated in which the order of the various operations is changed and the sub-operations described as a part of a larger operation are performed independently.

[0309] Furthermore, while the present invention has been described in the context of functional modules, it should be understood that, unless otherwise indicated, one or more of the functions and / or features may be integrated into a single physical device and / or software module, or one or more functions and / or features may be implemented in separate physical devices or software modules. It will also be understood that a detailed discussion of the actual implementation of each module is not necessary for understanding the present invention. More specifically, given the properties, functions, and internal relationships of the various functional modules in the devices disclosed in the present invention, the actual implementation of the module will be understood within the ordinary skill of an engineer. Therefore, a person skilled in the art will be able to implement the present invention as set forth in the claims using ordinary skill without undue experimentation. It will also be understood that the specific concepts disclosed are merely illustrative and are not intended to limit the scope of the present invention, which is determined by the full scope of the appended claims and their equivalents.

[0310] If the function is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art, or the part of the technical solution, can be embodied in the form of a software product. The computer software product is stored in a storage medium and includes several instructions for enabling a computer device (which can be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods of each embodiment of the present invention. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM, Read Only Memory), random access memory (RAM, Random Access Memory), disk or optical disk, and other media that can store program code.

[0311] The logic and / or steps represented in the flowcharts or otherwise described herein, for example, can be considered as an ordered list of executable instructions for implementing the logical functions, and can be embodied in any computer-readable medium for use by, or in conjunction with, an instruction execution apparatus, device, or apparatus (e.g., a computer-based apparatus, a device including a processor, or other apparatus that can fetch instructions from and execute instructions on an instruction execution apparatus, device, or apparatus). For purposes of this specification, a "computer-readable medium" can be any device that can contain, store, communicate, propagate, or transport a program for use by, or in conjunction with, an instruction execution apparatus, device, or apparatus.

[0312] More specific examples (a non-exhaustive list) of computer-readable media include the following: an electrical connection with one or more wires (electronic devices), a portable computer disk cartridge (magnetic devices), a random access memory (RAM), a read-only memory (ROM), an erasable and programmable read-only memory (EPROM or flash memory), a fiber optic device, and a portable compact disc read-only memory (CDROM). In addition, the computer-readable medium may even be paper or other suitable medium on which the program is printed, since the program may be obtained electronically, for example, by optically scanning the paper or other medium, followed by editing, deciphering, or processing in another suitable manner as necessary, and then stored in a computer memory.

[0313] It should be understood that various parts of the present invention can be implemented using hardware, software, firmware, or a combination thereof. In the above-described embodiments, multiple steps or methods can be implemented using software or firmware stored in a memory and executed by a suitable instruction execution device. For example, if implemented using hardware, as in another embodiment, any one of the following technologies known in the art or a combination thereof can be used: a discrete logic circuit having a logic gate circuit for implementing a logic function on a data signal, an application-specific integrated circuit having a suitable combination of logic gate circuits, a programmable gate array (PGA), a field programmable gate array (FPGA), etc.

[0314] Throughout this specification, reference to terms such as "one embodiment," "some embodiments," "examples," "specific examples," or "some examples" means that a specific feature, structure, material, or characteristic described in conjunction with that embodiment or example is included in at least one embodiment or example of the present invention. In this specification, schematic representations of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in any one or more embodiments or examples.

[0315] While embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions, and variations may be made to the embodiments without departing from the principles and spirit of the invention, and that the scope of the invention is defined by the claims and their equivalents.

[0316] The above is a specific description of the preferred implementation of the present invention, but the present invention is not limited to the embodiments. Those skilled in the art can make various equivalent modifications or substitutions without violating the spirit of the present invention. These equivalent modifications or substitutions are all included in the scope defined by the claims of the present invention.

Claims

1. A method for arranging experimental process instructions, characterized in that: The following steps are involved: Obtaining an operation description of a target object, and inputting the operation description into a target large language model; Parsing the operation description into script instructions using the target large language model based on a target knowledge base; the target knowledge base includes a professional knowledge base and a local knowledge base; Performing parallel task scheduling on the script instructions to obtain a standardized instruction set; The experimental equipment is controlled to perform target operations according to the standardized instruction set.

2. The experimental process instruction arrangement method according to claim 1, characterized in that: The professional knowledge base includes experimental specifications, and the local knowledge base includes device information of each experimental device; parsing the operation description into script instructions based on the target knowledge base using the target large language model includes the following steps: Using the target large language model to convert the operation description into standardized operation logic based on the experimental specifications of the professional knowledge base; the standardized operation logic includes at least one logical step; Based on the device information in the local knowledge base, the experimental device associated with executing each logical step in the standardized operation logic is determined, and each logical step is parsed into an atomic operation sequence corresponding to the experimental device according to the device information using the target large language model, and then the script instructions are integrated.

3. The experimental process instruction arrangement method according to claim 2, characterized in that: The method further comprises the following steps: Constructing a parameter constraint range for each functional operation of the experimental equipment based on the equipment information; the equipment information includes the function, parameters, and constraints of each of the experimental equipment and the inventory information of the experimental items corresponding to each of the experimental equipment; The target large language model is used to perform security parameter verification on the atomic operation sequence corresponding to each of the logical steps based on the parameter constraint range.

4. The experimental process instruction arrangement method according to claim 1, characterized in that: The method further comprises the following steps: Acquire professional knowledge text in the target professional field uploaded by the target subject; the professional knowledge text includes experimental specification data and operation guide data and equipment operation data of each of the experimental equipment; The professional knowledge text is uploaded to the target cloud corresponding to the target large language model, and the professional knowledge text is processed into an experimental specification knowledge vector by using an embedding operation on the target cloud, and then the professional knowledge base is constructed or updated based on the knowledge vector.

5. The experimental process instruction arrangement method according to claim 1, characterized in that: The method further comprises the following steps: Acquire real-time device information of each of the experimental devices based on a preset framework; wherein each of the experimental devices is connected to the preset framework through a preset interface; The local knowledge base is constructed or updated through prompt word engineering based on the real-time device information.

6. The experimental process instruction arrangement method according to claim 1, characterized in that: Before the step of performing parallel task scheduling on the script instructions, the method further includes the following steps: Transmitting the script instruction to the target object to obtain a verification instruction of the target object for the script instruction; the verification instruction includes a confirmation instruction or an adjustment instruction; When the verification instruction is the adjustment instruction, a modification operation is performed on the script instruction in response to the adjustment instruction to update the script instruction.

7. The experimental process instruction arrangement method according to claim 1, characterized in that: The script instructions include a plurality of task instructions; and performing parallel task scheduling on the script instructions to obtain a standardized instruction set includes the following steps: Storing each of the task instructions in the script instructions as a single-step task in a parallel list; The target large language model is used to generate a predecessor task and constraint conditions for each single-step task based on the target knowledge base; wherein the predecessor task of a single-step task is associated with other single-step tasks in the parallel list.

8. The experimental process instruction arrangement method according to claim 7, characterized in that: The step of controlling the experimental equipment to perform the target operation according to the standardized instruction set includes the following steps: Traversing the unfinished single-step tasks in the parallel list, and adding multiple single-step tasks that meet preset conditions to the parallel task set; The preset conditions include that the preceding task corresponding to the single-step task is completed, and the state parameters of the experimental equipment corresponding to the single-step task meet the constraint conditions corresponding to the single-step task; Controlling the experimental equipment corresponding to each of the single-step tasks in the parallel task set to perform a target task operation based on the task instruction corresponding to each of the single-step tasks; The single-step task that completes the target task operation in the parallel list is marked as completed, and the step of traversing the unfinished single-step tasks in the parallel list is returned to be executed until all the single-step tasks in the parallel list are marked as completed.

9. An experimental process instruction arrangement device, characterized in that: include: The first module is used to obtain an operation description of the target object and input the operation description into the target large language model; The second module is configured to parse the operation description into script instructions based on a target knowledge base using the target large language model; the target knowledge base includes a professional knowledge base and a local knowledge base; The third module is used to perform parallel task scheduling on the script instructions to obtain a standardized instruction set; The fourth module is used to control the experimental equipment to perform target operations according to the standardized instruction set.

10. An electronic device, characterized in that: including a processor and a memory; The memory is used to store programs; The processor executes the program to implement the method according to any one of claims 1 to 8.

11. A computer-readable storage medium storing a program, characterized in that: When the program is executed by a processor, the method according to any one of claims 1 to 8 is implemented.

12. A computer program product comprising a computer program, characterized in that When the computer program is executed by a processor, the method according to any one of claims 1 to 8 is implemented.

Images