A high-throughput preparation system and method for lithium iron phosphate based on artificial intelligence

Abstract

This invention provides an artificial intelligence-based high-throughput lithium iron phosphate preparation system, characterized by comprising: a synthesis module, a mobile robot, a process optimization module, and a central control module; wherein, the synthesis module includes: a raw material workstation, a ball milling workstation, a sintering workstation, and a testing workstation; the mobile robot is used to complete material transfer between the workstations in the synthesis module; the process optimization module includes a waiting time optimization unit; the central control module is used for data storage and analysis, receiving feedback information from the testing workstation and the process optimization module, and controlling the movement of the mobile robot.

Description

Technical Field

[0001] This invention relates to the field of lithium iron phosphate cathode materials, and more particularly to an artificial intelligence-based high-throughput lithium iron phosphate preparation system, and a method for preparing lithium iron phosphate using this system. Background Technology

[0002] The research and development of novel materials has long been hampered by an inefficient model characterized by low throughput, trial-and-error, and labor-intensive processes, severely limiting its progress. In recent years, the rapid development of internet big data, artificial intelligence (AI), and robotics has provided a significant historical opportunity for intelligent, high-throughput, automated, and efficient research and development of novel materials. Globally, relying on various experimental platforms, rapid progress has been achieved in areas such as catalysts and chemical synthesis. In the field of energy storage materials, particularly battery materials, machine learning technology has been widely used to predict and discover material performance. In the past, researchers constantly tried different materials and processes, a method that was inefficient and unable to meet the demands of rapidly developing high-tech industries. Artificial intelligence, with its powerful high-speed, massive data processing capabilities, is the most promising technology to overcome the aforementioned research bottlenecks, greatly promoting the development of battery material research, battery device design and manufacturing, and material and device characterization. Battery research and application generate massive amounts of data daily; artificial intelligence and machine learning can assist researchers in solving the parameter and data challenges of lithium-ion batteries, greatly advancing the industrialization of large-scale and high-performance electrochemical energy devices.

[0003] Lithium iron phosphate (LFP) is considered the most promising cathode material for power lithium-ion batteries due to its non-toxicity, environmental friendliness, abundant raw material sources, high specific capacity, and good cycle performance. High-temperature solid-state polymerization (HSP) is currently the most commonly used and mature method for LFP preparation. Raw material mixing and high-temperature calcination are key factors affecting product performance, and these two steps also involve significant waiting times for researchers. Currently, the emergence of AI-powered chemical robots has brought new hope to battery material research, with their automation levels continuously improving. However, lacking management and scheduling systems, these robots can only perform tasks at each workstation sequentially during chemical experiments. This leads to prolonged periods of inactivity, increasing the overall experimental time and making the robots sluggish and inefficient in automated chemical experimental procedures. Summary of the Invention

[0004] The purpose of this invention is to solve at least one of the following technical problems:

[0005] (1) The traditional high-temperature solid-state method for preparing lithium iron phosphate mainly relies on manual, intensive, and high-intensity experiments, which have low throughput, poor targeting, and low efficiency. Most experiments are based on the experience of the experimenters and are improved, resulting in a slow innovation process.

[0006] (2) The prepared lithium iron phosphate products need to be manually sampled and tested to measure their compaction density, specific surface area, electrical properties, etc. A large amount of test results data still need to be manually analyzed one by one, which requires a lot of effort.

[0007] (3) The sand milling and sintering processes in the high-temperature solid-state method require a long time, and the conventional chemical robot has a long waiting time, which reduces the experimental efficiency.

[0008] Therefore, this invention provides a high-throughput lithium iron phosphate preparation system based on artificial intelligence, comprising: a synthesis module, a mobile robot, a process optimization module, and a central control module, wherein:

[0009] The synthesis module includes: a raw material workstation, a ball milling workstation, a sintering workstation, and a testing workstation;

[0010] The mobile robot is used to transfer materials between workstations in the synthesis module.

[0011] The process optimization module includes a waiting time optimization unit;

[0012] The central control module is used for data storage and analysis, receiving feedback information from the testing workstation and process optimization module, and controlling the movement of the mobile robot.

[0013] In a specific embodiment of this application, the central control module is the central hub of the preparation system provided by this invention, controlling the operation of all modules and communicating with other modules. Specifically, the central control module is connected to the synthesis module, the mobile robot, and the process optimization module. The central control module sends the initial experimental plan to the process optimization module, which then feeds back the optimized plan to the central control module. The mobile robot completes the experimental operation according to the instructions of the central control module, and the central control module then derives the optimal experimental plan based on the test data from the analysis and testing workstation.

[0014] In a specific embodiment of this application, the central control module can be connected to the Internet and search for relevant literature based on Internet big data, perform data extraction and data cleaning on the literature content based on natural language learning methods, and form a lithium iron phosphate material database.

[0015] Preferably, the central control module extracts key experimental preparation information from the lithium iron phosphate material database and obtains an initial experimental plan based on artificial intelligence algorithms.

[0016] In this application, the central control module provides multiple initial experimental schemes. The mobile robot must complete all the initial experimental schemes. The central control module then selects the optimal experimental scheme based on the test data. The waiting time optimization unit is used to optimize the waiting time of the mobile robot during the experimental operation, minimize the waiting time, and speed up the experimental progress.

[0017] In this application, the optimization method of the waiting time optimization unit requires decomposing the initial experimental scheme into multiple sequential workstation processes, namely, the sampling process of the raw material workstation, the ball milling process of the ball milling workstation, the low-temperature pre-calcination process and the high-temperature calcination process of the sintering workstation, and the detection process of the detection workstation. Except for the sampling process, the other processes require relatively long waiting times. For example, the sampling (weighing) time is generally about 0.5 hours, while ball milling, sintering, and detection only require sample delivery and retrieval, each of which takes about 10 minutes.

[0018] In a specific embodiment of this application, the specific optimization steps of the waiting time optimization unit are as follows:

[0019] Step 1: Number the multiple initial experimental schemes, with the smaller the number, the higher the priority of the optimization scheme;

[0020] Step 2: Decompose each initial experimental scheme into the following 5 sequential workstation processes: sampling, ball milling, low-temperature pre-calcination, high-temperature calcination, and detection;

[0021] Step 3: Determine if there is a next workstation process to be executed in the current initial experimental scheme. If yes, proceed to step 4. If no, it means that all workstation processes in the current initial experimental scheme have been completed, and execute the next priority initial process scheme.

[0022] Step 4: Execute the next workstation process to be executed in the current initial experimental plan. Determine whether the task to be executed is a loading task. If yes, proceed to step 5; otherwise, it means that the task to be executed is a retrieval task, so proceed to step 3.

[0023] Step 5: Determine the waiting time of the workstation process to be executed in the current initial experimental plan, and the execution time of the workstation process to be executed in the next priority initial experimental plan. If the waiting time is long, execute the workstation process to be executed in the next priority initial experimental plan; otherwise, wait in place.

[0024] Step 6: After each workstation process is completed, repeat steps 3 to 5.

[0025] Step 7: Once all workstation processes required in the initial experimental schemes have been completed, stop optimizing the process.

[0026] In step 1, the numbering rule is random, as long as all the initial process schemes are completed.

[0027] During the experimental operation of mobile robots, there is a lot of waiting time, especially in ball milling and sintering workstations. Traditional mobile robots have to wait for the previous workstation to finish before starting the next workstation experiment. The waiting time optimization module in this application is designed to solve this problem. After this optimization process, the experimental waiting time can be greatly shortened and the experimental efficiency can be improved.

[0028] In this application, the process optimization module further includes a preliminary optimization unit. Before the waiting time optimization unit, the preliminary optimization unit optimizes the initial experimental scheme to obtain a preliminary process optimization scheme. The waiting time optimization unit then optimizes the waiting time of the preliminary process optimization scheme to obtain the process optimization scheme.

[0029] Preferably, the preliminary optimization unit uses a Bayesian optimization algorithm to optimize the initial experimental scheme.

[0030] More preferably, the Bayesian optimization algorithm is specifically as follows:

[0031] The test results of the samples are taken as y, and the corresponding experimental parameters are taken as X. Different samples correspond to different y and X, thus forming a sample dataset. Then, through a machine learning model, the objective function y = f(X) is established.

[0032] Randomly select sample points (X, y), and use Gaussian process regression to calculate the posterior distribution of the sample points. If the termination condition is met, output the optimal X value, which is the best experimental parameter.

[0033] If the termination condition is not met, the acquisition function is used to select the next iteration point (X). i y i The acquisition function can be the UCB function, EI function or PI function, and Gaussian process regression is performed again until the optimal X is output.

[0034] In this application, the test result data y includes compaction density, specific surface area, and particle size D50.

[0035] In this application, the experimental parameter X includes the type and amount of raw materials, ball milling time and speed, and sintering temperature and time.

[0036] In this application, the data in the sample dataset is derived from a lithium iron phosphate material database.

[0037] In the Bayesian optimization algorithm described above, the termination condition is a compaction density ≥ 2.0 g / cm³. 3Specific surface area is 10-20 m² 2 The particle size (D50) is between 0.8 and 2.0 μm, with a density between 0.8 and 2.0 μm.

[0038] In traditional experiments, optimizing experimental protocols relies heavily on the skill level of the experimenter, which significantly impacts experimental progress. To address this issue, the central control module proposed in this application possesses key information for experimental preparation in this field and derives preliminary process optimization schemes through a preliminary optimization unit. This avoids the shortcomings of relying solely on the experimenter's skill level to optimize experimental protocols. Furthermore, traditional experimental processes are entirely manual, leading to slow innovation. To address this, this application proposes a waiting time optimization unit, which not only enables unmanned laboratories and reduces labor costs but also allows the chemical robot to operate 24 / 7, achieving high-throughput preparation.

[0039] In a specific embodiment of this application, the workflow of the high-throughput lithium iron phosphate preparation system is as follows:

[0040] S1. The central control module extracts key experimental preparation information from the lithium iron phosphate material database and obtains the initial experimental plan based on artificial intelligence algorithms.

[0041] S2. The central control module sends the initial experimental plan to the process optimization module, which optimizes the waiting time to obtain the process optimization plan.

[0042] S3. The central control module sends instructions to the mobile robot to conduct a high-throughput experiment according to the process optimization plan.

[0043] S4. After the experiment is completed, the central control module selects the optimal experimental scheme based on the test data from the testing workstation.

[0044] Preferably, the workflow of the high-throughput lithium iron phosphate preparation system is as follows:

[0045] S1. The central control module extracts key experimental preparation information from the lithium iron phosphate material database and obtains the initial experimental plan based on artificial intelligence algorithms.

[0046] S2-1. The central control module sends the initial experimental plan to the process optimization module. The process optimization module uses the Bayesian optimization algorithm to optimize the initial experimental plan and obtain the preliminary process optimization plan.

[0047] S2-2, The waiting time optimization unit optimizes the waiting time of the preliminary process optimization plan to obtain the process optimization plan;

[0048] S3. The central control module sends instructions to the mobile robot to conduct a high-throughput experiment according to the process optimization plan.

[0049] S4. After the experiment is completed, the central control module selects the optimal experimental scheme based on the test data from the testing workstation.

[0050] In a specific embodiment of this application, the raw materials of the raw material workstation include: an iron source, a phosphorus source, and a lithium source, wherein:

[0051] The iron source is selected from at least one of ferrous oxalate FeC2O4·2H2O, ferrous acetate, and ferric phosphate.

[0052] The phosphorus source is selected from at least one of diammonium hydrogen phosphate (NH4)2HPO4, diammonium hydrogen phosphate, and iron phosphate.

[0053] The lithium source is selected from at least one of lithium carbonate (Li2CO3), lithium hydroxide, and lithium acetate.

[0054] In a specific embodiment of this application, the amount of raw materials used is controlled by controlling the atomic ratio of lithium (Li), iron (Fe), and phosphorus (P) in the iron source, phosphorus source, and lithium source, for example, Li:Fe:P = (0.95-1.05):1:1.

[0055] In a specific embodiment of this application, the specific process of the raw material workstation includes: the mobile robot receives instructions from the central control module, weighs the iron source, phosphorus source, and lithium source according to the proportions, and then transfers the weighed raw materials to the ball mill workstation.

[0056] In a specific embodiment of this application, the specific process of the ball milling workstation includes: a mobile robot receiving instructions from the central control module, loading the weighed raw materials into the ball mill, the central control module setting the ball milling time and speed, and after the ball milling is completed, the mobile robot removing the sample and transferring it to the sintering workstation. For example, the ball milling time is set to 6-12 hours, and the speed is set to 400-600 r / min to ensure that the raw materials are mixed evenly and refined.

[0057] The ball milling workstation contains multiple ball mills, enabling simultaneous ball milling of multiple samples to save waiting time and accelerate experimental speed. The ball milling time and speed are controlled by a central control module; the mobile robot only needs to load and unload samples from the ball mills.

[0058] In a specific embodiment of this application, the specific process of the sintering workstation includes: the mobile robot receiving instructions from the central control module, loading the ball-ground sample into the sintering furnace, the central control module setting the sintering program, and after sintering, the mobile robot taking out the sample and transferring it to the testing workstation.

[0059] Preferably, after the mobile robot removes the sample from the sintering furnace, it grinds the sample and then transfers it to the testing workstation.

[0060] The sintering workstation contains multiple box furnaces, enabling simultaneous sintering of multiple samples to save waiting time and accelerate the experimental process. The box furnaces are in a nitrogen atmosphere. A mobile robot, following instructions from the central control module, loads and removes spheroidized samples from the sintering furnace.

[0061] The sintering process can be as follows: under an inert atmosphere, pre-fire at a low temperature of 300-350℃ for 5-10 hours, then cool to room temperature, and finally calcine at a high temperature of 600-800℃ for 10-20 hours.

[0062] In a specific embodiment of this application, the specific process of the testing workstation includes: the mobile robot receiving instructions from the central control module, performing performance testing on the sintered sample, and feeding back the test results to the central control module.

[0063] The testing workstation includes a powder compaction density meter, a specific surface area meter, and a particle size analyzer. The mobile robot, according to the instructions issued by the central control module, transports the sintered, cooled, and ground samples to the testing workstation. The testing workstation tests the compaction density, specific surface area, and particle size D50 of the samples and feeds the test data back to the central control module.

[0064] In a specific embodiment of this application, the chemical operations of the mobile robot are performed by a six-degree-of-freedom robotic arm equipped with a gripper with force feedback control, a depth camera, a laser sensor, and a vision-guided intelligent detection and positioning system. To meet the high positioning accuracy requirements of the gripper during experimental operations, each experimental workstation is affixed with a unique QR code label. The robot can identify the current workstation by visually recognizing the QR code label in its field of view, and accurately measure the pose of the workstation based on the relative pose between the QR code label and the target to be operated on.

[0065] To ensure precise and interactive chemical operations, the mobile robot is equipped with an integrated mapping and positioning system based on dual lidar to acquire its location information and laboratory map, thereby enabling navigation and autonomous obstacle avoidance.

[0066] Secondly, the present invention provides a high-throughput preparation method for lithium iron phosphate based on artificial intelligence, using the above-mentioned preparation system, the preparation method comprising:

[0067] (1) Place the iron source, phosphorus source and lithium source into the raw material workstation. The central control module sends the initial experimental plan to the process optimization module for waiting time optimization, and then feeds back the process optimization plan to the central control module.

[0068] (2) The mobile robot receives instructions from the central control module and conducts high-throughput experiments;

[0069] (3) The mobile robot weighs the raw materials at the raw material workstation in the synthesis module;

[0070] (4) After the raw materials are weighed, the mobile robot loads the raw materials into the ball mill, and the central control module controls the rotation speed and time of the ball mill.

[0071] (5) After ball milling, the mobile robot takes the sample out of the ball mill and puts it into the sintering furnace. The central control module controls the sintering temperature and time of the sintering furnace.

[0072] (6) After sintering, the mobile robot takes the sample out of the sintering furnace and finally sends it to the testing workstation.

[0073] (7) After the mobile robot completes all the initial experimental schemes, the central control module selects the optimal experimental scheme based on the test data of the detection workstation.

[0074] Steps (3) to (6) are repeated according to the number of initial experimental schemes, and the order of steps (3) to (6) is obtained by the process optimization module.

[0075] In a specific embodiment of this application, the parameters of the initial experimental scheme in step (1) include: the type and amount of raw materials, the ball milling speed and ball milling time, and the sintering time and temperature.

[0076] Beneficial effects:

[0077] The system of this invention enables unmanned laboratories, reduces labor costs, and allows the chemical robot to operate 24 hours a day, achieving the goal of high-throughput preparation.

[0078] The system of this invention improves upon the traditional optimization module design. In response to the problem of long waiting time in the preparation of lithium iron phosphate, a waiting time optimization module is adopted, which greatly reduces the waiting time of the chemical robot in place and further improves experimental efficiency. Attached Figure Description

[0079] Figure 1 This is a schematic diagram of the high-throughput lithium iron phosphate preparation system provided by the present invention;

[0080] Figure 2 A schematic diagram illustrating the structure of the synthesis module provided by this invention;

[0081] Figure 3 This is a flowchart of the Bayesian optimization algorithm used in this invention;

[0082] Figure 4This is a diagram illustrating the waiting time in the experimental process of a traditional mobile robot.

[0083] Figure 5 A schematic diagram illustrating the waiting time of the experimental process for the mobile robot provided in this application;

[0084] Figure 6 A schematic diagram of the workflow of a high-throughput lithium iron phosphate preparation system provided in this application;

[0085] Figure 7 A schematic diagram of the workflow of another high-throughput lithium iron phosphate preparation system provided in this application;

[0086] Figure 8 A flowchart illustrating the waiting time optimization unit provided in this application. Detailed Implementation

[0087] To make the technical problems solved, the technical solutions, and the beneficial effects of this invention clearer, the invention will be further described in detail below with reference to embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention.

[0088] The endpoints and any values ​​of the ranges disclosed herein are not limited to the precise ranges or values, and these ranges or values ​​should be understood to include values ​​close to these ranges or values. For numerical ranges, the endpoint values ​​of the various ranges, the endpoint values ​​of the various ranges and individual point values, and individual point values ​​can be combined with each other to obtain one or more new numerical ranges, which should be considered as specifically disclosed herein.

[0089] In the description of this invention, it should be noted that the terms "comprising," "including," or any other variations thereof are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.

[0090] The present invention will be further described in detail below through specific embodiments.

[0091] Example 1

[0092] This embodiment provides an artificial intelligence-based high-throughput lithium iron phosphate preparation system, including: a synthesis module, a mobile robot, a process optimization module, and a central control module, wherein:

[0093] The synthesis module includes: a raw material workstation, a ball milling workstation, a sintering workstation, and a testing workstation;

[0094] The mobile robot is used to transfer materials between workstations in the synthesis module.

[0095] The process optimization module includes a waiting time optimization unit;

[0096] The central control module is used for data storage and analysis, receiving feedback information from the testing workstation and the process optimization module, and controlling the movement of the mobile robot. The central control module is connected to the synthesis module, the mobile robot, and the process optimization module. The central control module sends the initial experimental plan to the process optimization module, which then feeds back the optimized plan to the central control module. The mobile robot completes the experimental operation according to the instructions of the central control module, and the central control module then obtains the optimal experimental plan based on the test data from the testing workstation.

[0097] This embodiment provides a method for preparing lithium iron phosphate using the above-described preparation system, including:

[0098] (1) Place the iron source, phosphorus source and lithium source into the raw material workstation. The central control module provides two initial experimental schemes:

[0099] Initial experimental scheme 1 (sample 1):

[0100] The raw materials were 73.89g of lithium carbonate and 271.8g of iron phosphate; the ball milling speed was 500r / min and the ball milling time was 8h; the sintering procedure was to preheat at a low temperature of 350℃ for 6h, cool and then calcine at a high temperature of 650℃ for 12h.

[0101] Initial experimental scheme 2 (sample 2):

[0102] The raw materials were 73.89g of lithium carbonate and 271.8g of iron phosphate; the ball milling speed was 600r / min and the ball milling time was 8h; the sintering procedure was to preheat at a low temperature of 350℃ for 6h, cool and then calcine at a high temperature of 700℃ for 15h.

[0103] (2) The mobile robot receives the instructions from the central control module, completes the process optimization scheme given by the process optimization module, and begins to weigh the raw materials of sample 1 at the raw material workstation in the synthesis module.

[0104] (3) After the raw materials for sample 1 are weighed, the mobile robot loads the raw materials into the ball mill, and the central control module controls the rotation speed and time of the ball mill.

[0105] (4) After the ball mill starts ball milling, the mobile robot receives the instruction from the central control module and returns to the raw material workstation to weigh the raw material for sample No. 2.

[0106] (5) After the raw materials for sample No. 2 are weighed, the mobile robot loads the raw materials into the ball mill, and the central control module controls the rotation speed and time of the ball mill.

[0107] (6) After the ball milling of sample No. 1 is completed, the mobile robot takes sample No. 1 out of the ball mill and puts it into the sintering furnace. The central control module controls the sintering temperature and time of the sintering furnace.

[0108] (7) After the ball milling of sample No. 2 is completed, the mobile robot takes sample No. 2 out of the ball mill and puts it into the sintering furnace. The central control module controls the sintering temperature and time of the sintering furnace.

[0109] (8) After the sintering of sample No. 1 is completed, the mobile robot takes the sample out of the sintering furnace and finally sends it to the testing workstation.

[0110] (9) After the sintering of sample No. 2 is completed, the mobile robot takes the sample out of the sintering furnace and finally sends it to the testing workstation.

[0111] (10) The central control module obtains the optimal experimental scheme based on the test data of the two samples, which is the initial experimental scheme 2.

[0112] like Figure 4 As shown, traditional mobile robots, without optimized process design, can only prepare the next sample after the previous one is completed. During ball milling and sintering, the mobile robot can only wait in place. Typically, sampling (weighing) takes about 0.5 hours, ball milling about 6 hours, low-temperature pre-calcination about 6 hours, high-temperature calcination about 12 hours, and testing about 3 hours. Therefore, if two samples are prepared, the waiting time in place is approximately 54 hours.

[0113] like Figure 5 As shown, after optimization by the process optimization module, while sample 1 is being ball-milled, the mobile robot weighs sample 2 and transfers it to another ball mill for ball milling. After sample 1 is finished ball-milling, it is removed and transferred to a low-temperature pre-calcination stage. During the pre-calcination waiting period, sample 2 is then removed from the ball mill. Therefore, the time required for preparing two samples in place is approximately 25 hours. Due to the relatively long waiting time in this project, the advantages of the technical solution provided in this application become more pronounced if there are many initial experimental schemes and an increased number of samples.

[0114] The specific embodiments of the present invention have been described in detail above. However, the present invention is not limited to the specific details in the above embodiments. Within the scope of the technical concept of the present invention, various simple modifications can be made to the technical solution of the present invention, and these simple modifications all fall within the protection scope of the present invention.

[0115] It should also be noted that the various specific technical features described in the above specific embodiments can be combined in any suitable manner without contradiction. In order to avoid unnecessary repetition, the present invention will not describe the various possible combinations separately.

[0116] Furthermore, various different embodiments of the present invention can be combined in any way, as long as they do not violate the spirit of the present invention, they should also be regarded as the content disclosed by the present invention.

Claims

1. A high-throughput lithium iron phosphate preparation system based on artificial intelligence, characterized in that, include: Synthesis module, mobile robot, process optimization module, and central control module; The synthesis module includes: a raw material workstation, a ball milling workstation, a sintering workstation, and a testing workstation; The mobile robot is used to transfer materials between workstations in the synthesis module. The process optimization module includes a preliminary optimization unit and a waiting time optimization unit; The central control module is used for data storage and analysis, receiving feedback information from the detection workstation and process optimization module, and controlling the movement of the mobile robot. The specific optimization steps of the waiting time optimization unit are as follows: Step 1: Number the multiple initial experimental schemes, with the smaller the number, the higher the priority of the optimization scheme; Step 2: Decompose each initial experimental scheme into the following 5 sequential workstation processes: sampling, ball milling, low-temperature pre-calcination, high-temperature calcination, and detection; Step 3: Determine if there is a next workstation process to be executed in the current initial experimental scheme. If yes, proceed to step 4. If no, it means that all workstation processes in the current initial experimental scheme have been completed, and execute the next priority initial process scheme. Step 4: Execute the next workstation process to be executed in the current initial experimental scheme. Determine whether the workstation process to be executed is a loading task. If yes, proceed to step 5; otherwise, it means that the workstation process to be executed is a retrieval task, so proceed to step 3. Step 5: Determine the waiting time of the workstation process to be executed in the current initial experimental plan, and the execution time of the workstation process to be executed in the next priority initial experimental plan. If the waiting time is long, execute the workstation process to be executed in the next priority initial experimental plan; otherwise, wait in place. Step 6: After each workstation process is completed, repeat steps 3 to 5. Step 7: Once all workstation processes required in the initial experimental schemes have been completed, stop optimizing the process; The central control module extracts key experimental preparation information from the lithium iron phosphate material database and obtains an initial experimental plan based on artificial intelligence algorithms. Before the waiting time optimization unit, the preliminary optimization unit optimizes the initial experimental scheme to obtain a preliminary process optimization scheme. The waiting time optimization unit then optimizes the waiting time of the preliminary process optimization scheme to obtain the process optimization scheme.

2. The preparation system according to claim 1, characterized in that, The central control module is connected to the synthesis module, the mobile robot, and the process optimization module. The central control module sends the initial experimental plan to the process optimization module, which then feeds back the optimized plan to the central control module. The mobile robot completes the experimental operation according to the instructions of the central control module, and the central control module then obtains the optimal experimental plan based on the test data from the analysis and testing workstation.

3. The preparation system according to claim 1, characterized in that, The central control module retrieves relevant literature based on internet big data, performs data extraction and cleaning based on natural language learning methods, and forms a lithium iron phosphate material database.

4. The preparation system according to claim 1, characterized in that, The raw materials in the raw material workstation include: iron source, phosphorus source and lithium source, wherein: The iron source is selected from at least one of ferrous oxalate FeC2O4·2H2O, ferrous acetate, and ferric phosphate. The phosphorus source is selected from at least one of diammonium hydrogen phosphate (NH4)2HPO4, diammonium hydrogen phosphate, and iron phosphate. The lithium source is selected from at least one of lithium carbonate (Li2CO3), lithium hydroxide, and lithium acetate.

5. The preparation system according to claim 4, characterized in that, The specific process of the raw material workstation includes: the mobile robot receives instructions from the central control module, weighs the iron source, phosphorus source and lithium source according to the proportion, and then transfers the weighed raw materials to the ball mill workstation.

6. The preparation system according to claim 1, characterized in that, The specific process of the ball milling workstation includes: the mobile robot receives instructions from the central control module, loads the weighed raw materials into the ball mill, the central control module sets the ball milling time and speed, and after the ball milling is completed, the mobile robot takes out the sample and transfers it to the sintering workstation.

7. The preparation system according to claim 1, characterized in that, The specific process of the sintering workstation includes: the mobile robot receives instructions from the central control module, loads the ball-ground sample into the sintering furnace, the central control module sets the sintering program, and after sintering, the mobile robot takes out the sample and transfers it to the testing workstation.

8. The preparation system according to claim 1, characterized in that, The specific process of the testing workstation includes: the mobile robot receiving instructions from the central control module, performing performance testing on the sintered sample, and feeding back the test data to the central control module.

9. The preparation system according to claim 8, characterized in that, The performance tests include compaction density, specific surface area, and particle size D50.

10. A high-throughput preparation method for lithium iron phosphate based on artificial intelligence, comprising preparation using the preparation system described in any one of claims 1 to 9, characterized in that, The preparation method includes: (1) Place the iron source, phosphorus source and lithium source into the raw material workstation. The central control module sends the initial experimental plan to the process optimization module for waiting time optimization, and then feeds back the process optimization plan to the central control module. (2) The mobile robot receives instructions from the central control module and conducts high-throughput experiments; (3) The mobile robot weighs the raw materials at the raw material workstation in the synthesis module; (4) After the raw materials are weighed, the mobile robot loads the raw materials into the ball mill, and the central control module controls the rotation speed and time of the ball mill. (5) After ball milling, the mobile robot takes the sample out of the ball mill and puts it into the sintering furnace. The central control module controls the sintering temperature and time of the sintering furnace. (6) After sintering, the mobile robot takes the sample out of the sintering furnace and finally sends it to the testing workstation. (7) After the mobile robot completes all the experimental operations, the central control module selects the optimal experimental scheme based on the test data of the detection workstation. Steps (3) to (6) are repeated according to the number of initial experimental schemes, and the order of steps (3) to (6) is obtained by the process optimization module.

Images