Intelligent control system for reflux extraction and rotary evaporation concentration device

Abstract

The invention relates to an intelligent control system for a reflux extraction and rotary evaporation concentration device, and belongs to the technical field of intelligent control of experimental equipment. The problems that in medical and chemical experiments, parameter setting depends on artificial experience, multi-parameter cooperative control precision is insufficient, and the automation degree of backflow and concentration mode switching is low are mainly solved. According to the technical scheme, an artificial intelligence control unit is adopted to connect various sensors and a cloud server, and automatic parameter calling, high-precision closed-loop control and real-time data recording are achieved by presetting an experiment template library containing temperature, vacuum degree, rotating speed, duration and continuous operation mode flag bits. And switching of the diversion structure and starting and stopping of the equipment are automatically controlled according to sequential logic. The system is mainly used for improving the automation level, control precision and result repeatability of reflux extraction and rotary evaporation concentration experiments.

Description

Technical Field

[0001] This invention belongs to the field of intelligent control technology for experimental equipment, specifically relating to an intelligent control system for a reflux extraction and rotary evaporation concentration apparatus. Background Technology

[0002] In experimental research in fields such as medicine and chemistry, reflux extraction and rotary evaporation concentration are common sample pretreatment procedures. Currently, the control of these devices largely relies on manual settings and real-time adjustments by operators. Because the experimental process involves multiple parameters such as temperature, vacuum, rotation speed, and time, their set values ​​and their interrelationships often depend on individual experience, lacking standardized and reproducible parameters. Parameter settings may vary significantly between different operators or between different experimental batches, directly affecting the repeatability and reliability of experimental results.

[0003] This over-reliance on human experience stems from the complexity and variability of process parameters and their close interrelationships. For example, temperature fluctuations can affect extraction efficiency, changes in vacuum are closely related to evaporation rates, and rotation speed affects film renewal and heat exchange. Operators often struggle to precisely coordinate the dynamic interplay of multiple parameters simultaneously, relying instead on rough estimates or past successes, lacking a systematic approach to optimization. Furthermore, manual settings are insufficient for achieving high-precision, stable parameter control; external environmental interference or equipment fluctuations can cause actual values ​​to deviate from the set range, thus impacting extraction or concentration results. For instance, the rationality of experimental design, the control of experimental conditions, and the standardization of data acquisition and processing are crucial for achieving stable and reliable experimental results.

[0004] To address these issues, attempts have been made to reduce human intervention by pre-setting fixed procedures or formulas. However, these methods still have significant limitations: first, the procedures are usually based on fixed parameters and cannot be adaptively adjusted according to different sample characteristics or experimental objectives; second, they lack real-time closed-loop control capabilities and are difficult to cope with disturbances that occur during the experiment; and third, the connection between different experimental steps (such as the conversion between reflux and concentration) still relies on manual judgment and operation, resulting in limited automation, low efficiency, and susceptibility to operational errors.

[0005] Therefore, existing technologies suffer from problems such as poor experimental repeatability, insufficient parameter control precision, and low degree of automation in multi-step collaboration due to over-reliance on human experience. These problems limit further improvements in experimental efficiency and result reliability, necessitating a control method capable of intelligent parameter management, high-precision control, and automatic switching between multiple modes. Summary of the Invention

[0006] One object of the present invention is to solve at least the above-mentioned problems and to provide at least the advantages that will be described later.

[0007] Another objective of this invention is to provide an intelligent control system for a reflux extraction and rotary evaporation concentration apparatus. This system can automatically call and optimize experimental process parameters through artificial intelligence technology, achieve high-precision closed-loop control of key parameters such as temperature, vacuum, and rotation speed, and switch between reflux extraction and evaporation concentration modes according to preset timing logic, thereby improving the repeatability, automation, and reliability of the experiment.

[0008] To achieve these objectives and other advantages of the present invention, an intelligent control system for a reflux extraction and rotary evaporation concentration apparatus is provided, comprising a rotating flask, a condenser, a heated water bath, and a vacuum pump. A flow guide structure for switching between reflux extraction and vacuum concentration functions is provided below the condenser. The system further includes: An artificial intelligence control unit connects to a temperature sensor, a vacuum sensor, a rotation speed sensor, a data storage unit, and a cloud server. The data storage unit contains a pre-set library of extraction and preparation experimental templates. Each template includes a temperature setpoint range of 30-90℃, a vacuum setpoint range of 5-100 kPa, a rotation speed setpoint range of 50-200 rpm, an experimental duration setpoint range of 10-300 minutes, and a continuous operation mode flag for defining the timing logic of reflux extraction and evaporation concentration. The artificial intelligence control unit integrates: The network interaction module supports remote access from both computer and mobile devices, and provides parameter settings, status monitoring and result visualization interfaces; The process optimization engine learns autonomously from literature databases and historical experimental data to generate process optimization schemes and dynamically update the template library. The processor, in response to template instructions, executes: 1) Automatically call template parameters; 2) Temperature fluctuation is maintained at ±0.5℃ and vacuum error at ±10KPa through closed-loop control; 3) Record sensor data to the data storage unit in real time; 4) When the template contains a continuous operation mode flag, the position of the flow guide structure is switched according to the preset timing. 5) The equipment will automatically stop when the entire experiment is completed.

[0009] This invention automatically calls preset experimental templates and achieves high-precision closed-loop control through an artificial intelligence control unit, which significantly reduces the dependence of the experimental process on the operator's personal experience, ensures the stability and control accuracy of key parameters such as temperature and vacuum, and improves the repeatability and reliability of experimental results. At the same time, by integrating a continuous operation mode flag, it realizes automatic timing switching between reflux extraction and evaporation concentration functions, improving the automation level and overall efficiency of the device.

[0010] Preferably, the process optimization engine is configured with a template cloning module, which creates an independent copy of the template in the working storage area in response to user commands. The copy inherits the parameter range of the original template. The processor is prohibited from modifying the original template and is only allowed to adjust the parameters of the copy. When the user modifies the copy, the process optimization engine generates parameter adjustment suggestions by comparing with the historical best solution and displays them on the operation interface.

[0011] This invention allows users to adjust parameters in an independent copy through a template cloning module, which protects the integrity and consistency of the original template while providing users with flexible and personalized settings. During the user's modification process, the system provides parameter suggestions based on historical best solutions in real time to assist users in making scientific adjustments, further optimizing experimental processes, and improving the success rate of experiments.

[0012] Preferably, the process optimization engine has a built-in parameter verification module that simultaneously initiates process suitability analysis when performing boundary verification: the input value is compared with the distribution of process parameters extracted from similar components in the literature database; if it deviates from the distribution center by ±2 standard deviations, an additional process risk warning is added; the boundary verification mechanism performs a two-stage operation: In the first stage, a warning window will pop up in real time through the display unit of the operation interface, which includes the name of the out-of-bounds parameter and its corresponding valid range value. The second phase prohibits the execution of experimental start commands containing out-of-bounds parameters until the user corrects the out-of-bounds parameters to within the valid range.

[0013] The parameter verification module of this invention performs process suitability analysis by combining literature database data while checking for out-of-range parameters. It also issues risk warnings for parameters that deviate significantly from the conventional process range, and provides phased prompts and mandatory interventions. This effectively prevents experimental failures or equipment damage caused by improper parameter settings, and enhances the safety and reliability of the system.

[0014] Preferably, the parameter verification module displays two types of baseline values ​​simultaneously for each parameter to be adjusted in the parameter modification interface of the operation interface: The first baseline value is the existing setting of this parameter in the current independent copy; The second baseline value is the dynamic optimal value recommended by the process optimization engine, which is calculated and generated based on the real-time updated historical experimental model. Two types of benchmark values ​​are displayed side by side in the parameter input box as numerical labels, with the source template name indicated for the preset recommended value; When a user inputs a new parameter value, the parameter verification module first calculates the difference between the input value and the second benchmark value before performing out-of-bounds verification. If the absolute difference between the input value and the dynamic optimal value exceeds the process sensitivity threshold of the corresponding parameter (temperature ±5℃, vacuum ±5KPa, rotation speed ±30rpm, duration ±15%), a deviation warning message will be added to the warning window. The message includes the specific deviation amount and direction of the current input value relative to the preset recommended value, as well as a process confidence label. The process confidence label displays the data volume and model fit degree on which the current recommended value is based.

[0015] The parameter setting interface synchronously displays the current value and the system's recommended optimal value, giving users a clear reference benchmark when adjusting parameters. When the input value deviates from the recommended value by more than the process sensitivity threshold, the system provides detailed deviation information and confidence level explanations to help users understand the scientific basis for parameter adjustment, make more reasonable decisions, and thus improve the scientific nature of the experimental process and the predictability of the results.

[0016] Preferably, the parameter modification interface of the operation interface is supplemented with a batch editing mode entry; When the user activates the batch editing mode, the interface generates a table view containing all adjustable parameters. The table displays the parameter name identifier, the current independent copy setting value, the original template preset recommended value, the user-editable input box, and the recommended range generated by the process optimization engine. After the user completes the batch parameter input and triggers the confirmation command, the process optimization engine performs multi-objective optimization simulation on the input values, and the simulation results are pushed to the operation interface with a delay.

[0017] The introduction of batch editing mode significantly improves the operational efficiency of multi-parameter collaborative adjustment; after batch modification, the system performs multi-objective optimization simulation, which can evaluate the overall effect of parameter combination and delay the push of simulation results to provide data support for users, which helps to obtain better composite process parameters and simplifies the configuration process of complex experiments.

[0018] Preferably, the artificial intelligence control unit adds a hardware status verification stage after responding to the template instruction and before starting the experiment: The actual water volume in the heating water bath is detected by a water level sensor. When the detected value is lower than the minimum working water level threshold, an insufficient water level warning is generated. The static vacuum level of the system is monitored by a sealed force sensor. A test pressure of 10-30 kPa is applied before the vacuum pump is started. If the pressure change exceeds 5 kPa within 10-30 seconds, a seal failure warning is generated. The configuration status of the artificial intelligence control unit - template matching rule base: The process optimization engine trains a risk prediction model based on historical fault data and dynamically adjusts the water level / sealing threshold. The fault handling guidelines are linked to a knowledge graph, and emergency response cases matching the template are pushed to users.

[0019] Adding a hardware status self-check step before the experiment starts can proactively identify common potential faults such as insufficient water volume in the water bath and poor system sealing. Combined with knowledge graphs, it provides emergency response plans, transforming fault handling from passive response to proactive prevention. This reduces the risk of experiment interruption or failure due to poor equipment condition and ensures the smooth progress of the experiment.

[0020] Preferably, the control unit adds a risk level identifier to the state-template matching rule base. The risk level identifier is dynamically calculated by the process optimization engine: based on the template parameter combination and real-time sensor data, the failure probability is predicted. A probability ≥30% is defined as a level 1 risk; a probability of 10%-29% is defined as a level 2 risk. When multiple concurrent faults are detected during the hardware status verification phase, a tiered handling process is executed: If a Level 1 risk warning is present, the experiment start function on the operation interface will be locked immediately, a warning bar of the first color will flash continuously at the top of the display interface, all risk types will be displayed simultaneously, and a cloud alarm will be triggered and pushed to the bound mobile phone until the Level 1 risk warning is lifted. If only a level 2 risk warning exists, it will be statically displayed in the middle of the display interface as a warning bar of the second color, and users will be allowed to manually skip the warning.

[0021] By dynamically calculating risk levels and implementing differentiated handling strategies for different levels of risk (such as locking and initiating for level 1 risks and allowing skipping for level 2 risks), operators are given a certain degree of flexibility while ensuring experimental safety. This optimizes the human-computer interaction experience and achieves a balance between safety and operational efficiency.

[0022] Preferably, the data storage unit is configured with a ring buffer storage structure, which defines a fixed capacity storage block with a capacity limit of 120% of the maximum data volume of a single experiment. The capacity is automatically expanded as the experiment duration progresses and is backed up to the cloud in real time. When the amount of real-time data written does not reach the capacity limit, it is stored linearly in timestamp order; When the real-time data write volume reaches the capacity limit, the first-in-first-out overwrite rule is activated: the earliest stored data from group 1 to group N is marked as overwriteable, where N = buffer capacity - free capacity; new data is written to the overwriteable data positions in the order of timestamps.

[0023] A circular buffer structure is used for data storage, which can be dynamically expanded according to the duration of the experiment. This ensures continuous data recording while making efficient use of storage space. The first-in-first-out overwrite rule ensures the cyclical use of the storage area, avoids data overflow, and ensures data security through real-time cloud backup.

[0024] Preferably, the control unit is equipped with a key node marking module, which responds to three triggering conditions: Condition 1: The key node marking module generates marking instructions in real time via the manual marking button on the operation interface; Condition 2: When the rate of change of the temperature sensor's monitored value exceeds 0.5℃ / second for 5 consecutive seconds, or the rate of change of the vacuum sensor's monitored value exceeds 10KPa / second for 3 consecutive seconds, the critical node marking module automatically generates a marking instruction. Condition 3: When the process optimization engine identifies the inflection point of the coupled change rate of temperature-vacuum-speed, the critical node marking module automatically generates marking instructions; The circular buffer allocates independent storage identifier bits for marked records; when executing overwrite rules, records containing identifier bits are preferentially retained.

[0025] By automatically marking key node data in the experimental process under multiple conditions and implementing anti-coverage protection measures for these data, the retention and traceability of data in important experimental stages are ensured, providing valuable data support for subsequent process analysis, optimization, and experimental review.

[0026] Preferably, the continuous operation mode is executed according to the following process: After the user selects a template, they manually position the flow guide structure to the second working position of the reflux extraction and start closed-loop control. When the concentration of the extract reaches the set threshold or the predetermined time is reached, the artificial intelligence control unit pauses the equipment and issues a prompt sound, waiting for the user to manually remove the distilled extract. After completing filtration and loading the filtrate, the user needs to manually switch the flow guide structure to the first working position of evaporation and concentration, and click the "Continue" button on the operation interface to restore closed-loop control and complete the rotary evaporation and concentration process.

[0027] The specific method for monitoring the extract concentration to reach the set threshold is as follows: The optical probe of the near-infrared spectral sensor is integrated into the extract outlet of the reflux pipeline to scan the extract spectrum in real time. A database of characteristic absorption peaks for different solute-solvent combinations is pre-set in the concentration analysis module. The spectral data is converted into extract concentration values ​​using a partial least squares algorithm. After calibration with standard samples, the measurement error is ≤ ±8%. When the concentration change rate is ≤0.1% / min for 3 consecutive minutes, the extraction is considered to have reached equilibrium and the set threshold for extract concentration has been reached.

[0028] The detailed process of continuous operation mode was clarified. At the same time, by integrating an advanced near-infrared spectral sensor, the concentration of the extract can be monitored in real time, thereby enabling objective and quantitative judgment of the extraction endpoint. This replaces the traditional subjective judgment method that relies on human experience, improves the accuracy of the termination judgment in the extraction process, and ensures consistency between different experimental batches.

[0029] Other advantages, objectives and features of the present invention will become apparent in part from the following description, and in part from those skilled in the art through study and practice of the invention. Attached Figure Description

[0030] Figure 1 This is a schematic diagram of the structure of one technical solution of the present invention; Figure 2 This is a side view of the flow guide valve of the present invention. Figure 3 This is a schematic diagram of another technical solution of the present invention.

[0031] 1. Conical sleeve; 2. Snake-shaped condenser tube; 3. Evaporation flask; 4. Collection flask; 5. First port; 6. Second port; 7. Flow guide valve; 8. Feeding channel; 9. Evaporation inlet channel; 10. Feeding valve; 11. Glass shaft; 12. Clamping and fixing structure; 13. First condenser outlet channel; 14. Second condenser outlet channel; 15. Reflux outlet channel. Detailed Implementation

[0032] The present invention will now be described in further detail with reference to the accompanying drawings, so that those skilled in the art can implement it based on the description.

[0033] It should be understood that terms such as “having,” “comprising,” and “including” as used herein do not exclude the presence or addition of one or more other elements or combinations thereof.

[0034] It should be noted that, unless otherwise specified, the experimental methods described in the following embodiments are conventional methods, and the reagents and materials described are commercially available. In the description of this invention, it should be noted that, unless otherwise explicitly specified and limited, the terms "installation," "connection," and "setting" should be interpreted broadly. For example, they can refer to fixed connection or setting, detachable connection or setting, or integral connection or setting. Those skilled in the art can understand the specific meaning of the above terms in this invention based on the specific circumstances. The terms "lateral," "longitudinal," "upper," "lower," "front," "rear," "left," "right," "vertical," "horizontal," "top," "bottom," "inner," and "outer," etc., indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings, and are only for the convenience of describing this invention and simplifying the description. They do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation, and therefore should not be construed as a limitation of this invention.

[0035] like Figure 1-3 As shown, a rotary evaporator with reflux extraction function includes an evaporation system, a rotary drive system, a vacuum system, and a control system. The evaporation system includes a serpentine condenser 2, an evaporation flask 3, and a collection flask 4. The lower end of the serpentine condenser 2 is provided with a flow guiding structure, which includes: A conical sleeve 1 is inclinedly disposed below the serpentine condenser tube 2, and the diameter of the lower port is smaller than the diameter of the upper port. The conical sleeve 1 is provided with a first port 5 and a second port 6 in the middle. The first port 5 is connected to the lower end of the serpentine condenser tube 2, the second port 6 is connected to the inlet of the collection bottle 4, and the lower port of the conical sleeve 1 is connected to the evaporation bottle 3. The flow guide valve 7 has a conical structure, and its outer wall is fitted with the inner wall of the conical sleeve 1. The flow guide valve 7 includes a feeding channel 8, an evaporation inlet channel 9, a first condensation outlet channel 13, a second condensation outlet channel 14, and a reflux outlet channel 15. The feeding channel 8, the evaporation inlet channel 9, the first condensation outlet channel 13, the second condensation outlet channel 14, and the reflux outlet channel 15 are interconnected. The feeding channel 8 is provided with a feeding valve 10 to block the feeding channel 8. The flow guide valve 7 and the conical sleeve 1 are on the same axis, so that two working positions can be created by rotating the flow guide valve 7; When the flow guide valve 7 is in the first working position, the first condenser outlet channel 13 is opposite to the first port 5, the second condenser outlet channel 14 is opposite to the second port 6, and the return outlet channel 15 is opposite to the side wall of the conical sleeve 1 so that the evaporation flask 3, the serpentine condenser tube 2 and the collection bottle 4 are connected, thereby realizing the function of evaporation and concentration. When the flow guide valve 7 is in the second working position, the first condenser outlet channel 13 and the second condenser outlet channel 14 are opposite to the side wall of the conical sleeve 1, and the reflux outlet channel 15 is opposite to the first port 5 so that the evaporation flask 3 is connected to the serpentine condenser tube 2 through the flow guide valve 7, thereby realizing the function of reflux extraction.

[0036] In another technical solution, the length of the flow guide valve 7 is greater than that of the conical sleeve 1 so that the flow guide valve 7 extends from both ends of the conical sleeve 1. The lower end of the flow guide valve 7 is provided with a first groove, and a first annular sealing ring is fitted into the first groove. The conical sleeve 1 is fixedly connected to the connecting seat. The lower end of the flow guide valve 7 extends into the glass shaft 11 so that the first annular sealing ring abuts against the inner wall of the glass shaft 11 to achieve a rotational seal between the glass shaft 11 and the flow guide valve 7.

[0037] In another technical solution, a clamping and fixing structure 12 for clamping and fixing the flow guide valve 7 is also included, the clamping and fixing structure 12 comprising: The first anti-slip groove is provided on the conical sleeve 1, and the length direction of the first anti-slip groove is parallel to the length direction of the conical sleeve 1; The second anti-slip groove is installed on the flow guide valve 7, and the length direction of the second anti-slip groove is parallel to the length direction of the tapered sleeve 1. The locking ring has anti-slip protrusions on its inner side. The locking ring is partially fitted on the first anti-slip groove and partially fitted on the second anti-slip groove to prevent the flow guide valve 7 from rotating relative to the conical sleeve 1.

[0038] In another technical solution, the locking ring is an elastic C-shaped ring with an opening, and its inner surface has multiple anti-slip protrusions evenly distributed circumferentially. Both the first and second anti-slip grooves have a V-shaped cross section, and the groove depth is greater than 1 / 2 of the height of the anti-slip protrusion. When the locking ring is in the locked state, its open end is closed by the locking bolt, so that the anti-slip protrusion is simultaneously engaged in the first anti-slip groove and the second anti-slip groove. When the locking ring is in the unlocked state, its open end expands, causing the anti-slip protrusion to disengage from the anti-slip groove.

[0039] In another technical solution, the open end of the locking ring is provided with an outwardly extending locking lug, and the locking lug has a coaxial elongated hole; the locking bolt passes through the elongated holes of the two locking lugs in sequence, and the axial preload is applied by the threaded locking nut to bring the open end of the locking ring closer together; a disc spring assembly is installed in the elongated hole, and the disc spring assembly is sleeved on the locking bolt. One end of the disc spring assembly abuts against the inner wall of the locking lug, and the other end abuts against the nut or bolt; when the locking nut is tightened, the disc spring assembly is compressed and generates a reverse force, forcing the two locking lugs on both sides to retract axially along the locking bolt, while driving the two ends of the locking ring to move closer together to achieve fixation.

[0040] In another technical solution, such as Figure 3 As shown, the lower end of the tapered sleeve 1 is provided with a second groove. The tapered sleeve 1 is fixedly connected to the connecting seat. The lower end of the tapered sleeve 1 extends into the glass shaft 11 so that the second annular sealing ring on the second groove abuts against the inner wall of the glass shaft 11 and seals it.

[0041] In another technical solution, the second annular sealing ring includes: The inner ring is a highly elastic solvent-resistant rubber ring, and its inner circumferential surface is fixed with the second groove by an interference fit. The outer ring is made of a low-friction coefficient material with self-lubricating properties. The inner circumferential surface of the outer ring is vulcanized and bonded to the outer circumferential surface of the inner ring. The outer circumferential surface of the outer ring is a sealing working surface that slides in contact with the inner wall of the glass shaft 11.

[0042] The present invention provides an intelligent control system for a reflux extraction and rotary evaporation concentration apparatus. It can be used in the aforementioned rotary evaporation apparatus with reflux extraction function, and of course, it can also be used in other apparatuses with similar structures that can switch between different working positions through a flow guiding structure to achieve the switching between reflux extraction and rotary evaporation concentration functions.

[0043] This invention relates to an intelligent control system for a reflux extraction and rotary evaporation concentration apparatus. Compared with existing technologies that commonly rely on manual parameter setting and switching of operating modes, this system introduces an artificial intelligence control unit, achieving intelligent management and high-precision control of the experimental process. This is similar to applying AI technology in the automated control of physical experiments to improve experimental accuracy and efficiency, reduce human error, and lower experimental costs.

[0044] In existing technologies, operators need to rely on personal experience to set multiple parameters, including temperature, vacuum, rotation speed, and time. They also need to manually switch between reflux extraction and evaporation concentration functions. This is not only inefficient, but also prone to large parameter fluctuations due to human error, which affects the repeatability and accuracy of the experiment.

[0045] This invention employs a pre-set extraction and preparation experimental template library. Each template contains a reasonable parameter range and continuous operation mode indicators that have been verified through extensive experiments. Users only need to select a template, and the artificial intelligence control unit can automatically call up the relevant parameters. Through closed-loop control, it precisely maintains temperature fluctuations within ±0.5 degrees Celsius and vacuum errors within ±10 kPa, and automatically controls the flow guiding structure to switch working positions according to a preset sequence, thereby seamlessly completing the two stages of reflux extraction and evaporation concentration.

[0046] In addition, the system integrates a network interaction module, supporting remote monitoring via computers and mobile phones. The process optimization engine can continuously learn and update the template library based on historical data and literature, significantly reducing reliance on operator experience and improving the level of automation and consistency of results.

[0047] In existing technologies, users typically face two dilemmas when they wish to adjust preset experimental parameters: First, the system completely prohibits modification of preset parameters, which limits the flexibility of experiments; Secondly, allowing arbitrary modification of the original template can disrupt baseline parameters, affecting the repeatability and consistency of subsequent experiments. Furthermore, even in systems that allow modification, there is a lack of intelligent guidance for parameter adjustment; users are forced to rely on personal experience and trial and error, often leading to experimental failures or efficiency losses due to improper parameter configuration.

[0048] This invention integrates a template cloning module into a process optimization engine, enabling users to create independent copies of the original template in the working storage area. These copies fully inherit the parameter ranges of the original template, but allow users to adjust them, while the original template remains unmodifiable. During user parameter modifications, the system compares the copied template with the optimal solutions in the historical experimental database in real time, references database parameter adjustment cases, generates parameter adjustment suggestions, and displays them instantly on the user interface, providing scientific reference for the user.

[0049] This effectively solves the problem of poor parameter adjustment flexibility and difficulty in maintaining template integrity in existing technologies. Users can freely experiment with personalized settings in independent copies without compromising the stability and reusability of the original template. The real-time suggestions provided by the system are based on a large amount of historical data and optimization algorithms, significantly reducing the risk of parameter settings due to insufficient human experience, improving the success rate of experiments and the reliability of processes, while maintaining the system's ease of use and scalability.

[0050] In existing technologies, system verification of experimental parameters is typically limited to checking whether input values ​​exceed the safety limits of the equipment hardware, such as whether the temperature exceeds the maximum withstand temperature of the heating mantle. This simple out-of-range check cannot identify parameter settings that, while within the equipment's permissible range, significantly deviate from standard process specifications. For example, a user might choose to set an extremely low vacuum for reflux extraction. Although such a setting itself will not damage the vacuum pump, it will drastically reduce extraction efficiency, ultimately leading to experimental failure. Problems caused by improper process parameter settings cannot be effectively prevented in existing systems.

[0051] The built-in parameter validation module in this process optimization engine not only performs traditional out-of-bounds validation but also simultaneously conducts process suitability analysis. By comparing the user-input parameter values ​​with the distribution of process parameters extracted from a large number of similar components in a literature database, the system can identify significant outliers—those input values ​​that deviate from the distribution center by more than two standard deviations. At this point, the system will initiate a two-stage intervention mechanism. In the first stage, a warning window will immediately pop up on the user interface, clearly indicating the name of the out-of-bounds parameter and its valid range. If the user ignores this warning, the system will enter the second stage, forcibly prohibiting the initiation of experimental commands containing this out-of-bounds parameter until the parameter is corrected to the valid range.

[0052] This implementation method significantly enhances the intelligence level of parameter verification. It not only prevents equipment damage but also mitigates the risk of failure from the perspective of experimental process rationality. Combining innovative statistical methods and optimized experimental design, the system provides users with a scientifically sound guarantee. By dynamically adjusting sample size, optimizing randomization schemes, and employing adaptive experimental design, it significantly improves the reliability and success rate of experiments.

[0053] In existing technologies, even with parameter suggestions and out-of-range warnings, users still face a core dilemma when adjusting parameters: a lack of intuitive and reliable reference benchmarks. Users typically only see their current settings, while the system's recommended optimal values ​​are often hidden in secondary menus or complex reports, failing to provide immediate and direct comparisons during modifications. This makes it difficult for users to grasp the accuracy of the direction and the rationality of the adjustment, essentially remaining limited to trial and error based on personal experience. For example, a user might want to increase the temperature to accelerate the extraction process but lacks quick access to information on the optimal temperature range and expected effects recommended by the system based on historical data. This can lead to suboptimal extraction efficiency; for instance, overly conservative temperature adjustments may fail to achieve the desired extraction results, while overly aggressive temperature adjustments may introduce risks such as thermal degradation.

[0054] This invention features a deeply optimized parameter modification interface. When a user adjusts any parameter, two clear numerical labels are displayed side-by-side next to the input box: one is the current setting value in the user's current copy, and the other is the dynamically optimal value calculated and recommended in real time by the process optimization engine. This recommended value is not fixed but is derived from continuously updated historical experimental models, ensuring its scientific validity and timeliness. Crucially, the system immediately compares the new value with the recommended optimal value the moment the user enters it. If the system determines that the deviation exceeds the sensitive threshold for this type of process parameter, it will immediately add a detailed deviation prompt in the alert window, indicating not only the specific value and direction of the deviation but also providing a process confidence label, concisely informing the user of the amount of data and the reliability of the model upon which the recommended value is based.

[0055] This implementation method fundamentally changes the human-computer interaction model, transforming parameter adjustment from subjective guesswork to data-driven scientific decision-making. Users have a clear, reliable, and real-time updated reference coordinate system, enabling them to deeply understand the scientific basis and potential impact of each adjustment, thereby significantly improving the rationality of parameter settings and the predictability of experimental results.

[0056] Within the existing technological framework, when researchers need to coordinate adjustments to multiple process parameters, they often face tedious and inefficient operations. Users must locate and modify each parameter individually, repeatedly switching between different interfaces or tabs. This discrete approach is not only time-consuming and labor-intensive, but more importantly, it severs the intrinsic connections between parameters, making it difficult for users to have a holistic view and effectively assess how adjusting one parameter will affect other parameters, and ultimately what kind of combined impact it will have on the overall process effect. For example, if a user wants to simultaneously optimize extraction efficiency and energy consumption, it needs to coordinate multiple variables such as temperature, vacuum level, and rotation speed. Existing systems lack integrated support for such multi-objective optimization problems, forcing users to rely on guesswork and make multiple independent attempts, resulting in a blind process and unpredictable outcomes.

[0057] Building upon its intelligent parameter management system, this invention further introduces a batch editing mode. Users can activate this mode with a single click through a dedicated entry point on the parameter modification interface, and the system immediately generates a clearly structured table view. This table centrally displays all adjustable parameters, including temperature, vacuum level, rotation speed, duration, etc., along with their key information. Each row corresponds to a parameter, displaying its name, the current setpoint in the copy, the preset recommended value from the original template, an edit box allowing users to directly input new values, and a dynamically recommended range generated by the process optimization engine.

[0058] Users can efficiently input and modify all parameters in batches within this form. After editing, the user triggers a confirmation command. The system does not immediately start the experiment; instead, it first submits this new set of parameters to the process optimization engine. The engine uses its built-in multi-objective optimization algorithm to perform rapid simulation calculations on the new parameter set, evaluating its comprehensive performance across multiple objectives such as extraction efficiency, product quality, and energy consumption. The simulation results will be pushed to the user interface with a slight delay, providing data support and visual feedback on the expected effects of this set of parameters.

[0059] This implementation fundamentally changes the operational logic of multi-parameter optimization, transforming the originally isolated, serial parameter adjustment process into an integrated, parallel decision-making process. It significantly improves the efficiency of complex process configurations and, through overall simulation of parameter combinations, provides users with scientific data foresight, significantly reducing the configuration difficulty and failure risk of multi-parameter experiments.

[0060] In existing technologies, operators typically need to manually check the equipment status before starting reflux extraction or rotary evaporation concentration experiments, such as visually checking the water bath level and judging the system's airtightness based on experience. This method is highly dependent on the operator's attention and experience, and human negligence or judgment errors may lead to experimental malfunctions such as water bath dry burning or vacuum leakage, thereby interrupting the experimental process, causing sample damage, or even equipment damage. Existing systems lack the ability to automatically sense and warn of hardware status, cannot proactively identify potential risks before the experiment begins, and fault handling is often a reactive measure, reducing the overall reliability and efficiency of the experiment.

[0061] This invention adds a hardware status verification stage to the artificial intelligence control unit. This stage is automatically executed after the user selects a template and before the experiment officially starts. The system uses a water level sensor to monitor the actual water volume in the heating water bath in real time. If the water level is lower than the preset minimum working water level threshold, a low water level warning is immediately generated to prevent the risk of dry burning. Simultaneously, a high-precision sealing pressure sensor applies a test pressure of 10 to 30 kPa to the system before the vacuum pump starts. If the pressure change exceeds 5000 Pa within 10 to 30 seconds, a sealing failure is determined, and a corresponding warning is generated. Furthermore, the system constructs a status-template matching rule base. The process optimization engine continuously trains a risk prediction model based on historical fault data, dynamically adjusting the thresholds for water level and sealing criteria to ensure better alignment with practical application needs. A fault handling guidance module supported by a knowledge graph can automatically push emergency handling cases matching the experimental template, providing users with real-time operational guidance.

[0062] The knowledge graph can be constructed based on a domain knowledge base and historical fault data. First, equipment components (such as vacuum pumps, sealing rings, and heating water baths), fault types (such as seal failure and insufficient water level), and handling solutions (such as replacing sealing rings and replenishing water to the threshold) are extracted as entity nodes. Then, various semantic relationships between nodes are constructed through directed edges, such as "component-fault" relationships (vacuum pump - occurrence - vacuum leakage), "fault-solution" relationships (vacuum leakage - corresponding handling - pressure test and tightening of interface), and "solution-component" relationships (replacing sealing ring - acting on - condenser interface). At the same time, attributes such as fault probability and solution effectiveness confidence can be added to nodes. The graph database is used for storage and association querying, thereby realizing intelligent recommendation and reasoning of fault diagnosis and handling solutions.

[0063] By implementing a data-driven preventative maintenance program, laboratory equipment condition management has shifted from reactive response to proactive prevention. This program can identify most common hardware malfunctions before experiments begin, significantly reducing the risk of experimental failure due to poor equipment condition. Simultaneously, utilizing a dynamically updated rule base and knowledge graph, the system not only implements basic verification functions but also continuously optimizes judgment criteria and handling strategies with accumulated user experience, demonstrating continuously improving intelligent characteristics. This allows users to initiate experiments with higher confidence in equipment conditions, improving the smoothness of the entire process and the reproducibility of results.

[0064] In current practices of experimental equipment control systems, hardware anomalies are typically presented to operators in a uniform manner, such as pop-up dialog boxes or beeping sounds. This approach has significant drawbacks: it fails to differentiate the severity of risks, potentially leading to insufficient awareness of high-risk faults or overreaction to low-risk alerts. For example, a slightly below-standard water level in a water bath and a severe vacuum leak in the system might trigger similar warning sounds, making it difficult for operators to immediately identify the priority of action. This could easily delay addressing critical faults or lead to frequent interruptions of the experiment to deal with secondary issues. This lack of a tiered warning mechanism reduces human-computer interaction efficiency and makes it difficult to truly mitigate experimental risks.

[0065] Building upon basic hardware status verification, this invention further introduces a dynamic risk level identification and tiered handling mechanism. Utilizing machine learning technology, the process optimization engine analyzes the parameter combinations and sensor data of the selected experimental template in real time to predict the probability of potential failures. The system classifies risks into two distinct levels: failures with a predicted probability of occurrence of 30% or higher are defined as Level 1 risks, while those between 10% and 29% fall under Level 2 risks. When the system detects multiple anomalies simultaneously during the self-check phase, it initiates the tiered handling process. For Level 1 risks, such as a highly predicted probability of vacuum leakage, the system immediately locks the experiment start function, interrupting the process. A prominent red warning bar flashes continuously at the top of the interface, clearly listing all identified Level 1 risk types, and simultaneously pushes alarm information to the user's linked mobile phone, creating multiple alerts until the risk is resolved. For Level 2 risks, such as water levels slightly below the ideal value but still within the safety margin, the system statically displays a yellow warning bar in the middle of the interface, clearly indicating the potential problem, but allows the user to manually skip the warning and continue the experiment based on their actual judgment.

[0066] This tiered response strategy based on dynamic risk prediction significantly improves the intelligence and operational efficiency of human-computer interaction. It frees operators from monotonous alarms, allowing them to quickly focus on the most pressing dangers while retaining flexibility in handling non-emergency situations. The system no longer mechanically blocks all experiments under abnormal conditions; instead, it grants experienced operators a degree of autonomy in decision-making, ensuring a safety baseline and achieving an organic balance between safety and experimental efficiency.

[0067] In existing technologies, data storage systems for experimental equipment often employ a simple linear recording method. This method has two significant drawbacks: First, when the experiment lasts for a long time and the amount of data is huge, the data recording may be interrupted or lost due to insufficient storage space. Second, all data are treated equally, and once the overwrite mechanism is activated, the oldest data is often mechanically deleted. This may result in the permanent deletion of valuable data reflecting key transitions or abnormal states during the experiment, while ordinary data from the later stable operation phase is retained. This makes subsequent data analysis and process review lack sufficient basis and makes it difficult to accurately locate the key factors affecting the experimental results.

[0068] This invention features an innovative design for the data storage unit structure, employing a circular buffer combined with intelligent overwrite rules. This buffer is not fixed; its initial capacity is set to 120% of the maximum expected data volume for a single experiment, and it can intelligently expand according to the actual duration of the experiment, ensuring sufficient space to handle experiments of varying lengths. All data is synchronized to the cloud for backup in real time during writing, ensuring data security. Its core mechanism lies in intelligent data circulation and overwrite rules: when the real-time data volume has not reached the buffer capacity limit, the system linearly stores data according to timestamp order, ensuring the integrity of the data sequence. Once the data volume reaches the limit, the system does not simply clear the buffer, but instead initiates a first-in, first-out (FIFO) overwrite rule. It automatically calculates the earliest stored, overwriteable data segments and sequentially overwrites these older data positions, thereby achieving cyclical use of storage space and ensuring uninterrupted recording.

[0069] Even more intelligently, the system adds a critical node marking function. Important data segments can be marked through various methods, including manual marking, sensor-monitored value mutation identification, or intelligent inflection point identification by the process optimization engine. A circular buffer will assign independent storage identifiers to these marked records. When executing data overwrite rules, the system will prioritize retaining these marked critical data segments, while choosing to overwrite unmarked ordinary data segments.

[0070] This implementation method completely changes the passive situation of data storage. It not only achieves seamless and continuous recording of experimental data within a limited space, avoiding data overflow and interruption, but more importantly, it endows the system with the ability to judge the value of data. Through intelligent coverage and key data protection mechanisms, it ensures that the core data that best reflects the changing patterns of the experimental process and key process nodes are preserved, providing a high-quality and high-value data foundation for subsequent in-depth process analysis, optimization iteration, and experimental review, greatly improving the efficiency and intelligence level of data utilization.

[0071] A long-standing technical challenge in the acquisition and storage of experimental data is that systems often treat all data equally. Existing technologies are mostly limited to simple linear recording or basic cyclic overwrite storage, making it difficult to automatically distinguish between brief moments containing critical changes and stable, unchanging periods during the experiment. If operators need to investigate specific stages, such as moments of rapid temperature increases or sudden drops in vacuum, they must afterwards review massive amounts of uniformly timed data records one by one—a process that is inefficient and prone to overlooking crucial details. More importantly, when storage space reaches its limit and the overwrite mechanism is triggered, these valuable data fragments recording process turning points or abnormal states face the same risk of being erased as ordinary data. This means that the most analytically valuable information may be the first to be deleted, leaving subsequent process diagnostics and optimization work without a precise data foundation.

[0072] This invention introduces an intelligent key node marking module, significantly improving the value judgment capability of data management. This module proactively identifies and marks key moments in the experimental process in three ways: First, users can manually add real-time tags at any time based on experimental phenomena using the manual marking button on the operation interface; second, the system automatically monitors the rate of change of sensor data, automatically identifying abnormal abrupt changes and generating a mark when the temperature change rate exceeds 0.5 degrees Celsius per second for five consecutive seconds or the vacuum change rate exceeds 10 kilopascals per second for three consecutive seconds; third, the process optimization engine can comprehensively analyze the coupled change trends of temperature, vacuum, and rotation speed, intelligently identifying inflection points characterizing process transitions or equilibrium states. All these marked data records are recorded in a circular buffer and assigned an independent storage identifier. This identifier serves as the "identity card" for the data's importance. When the storage space reaches its limit and triggers a first-in-first-out overwrite rule, the system prioritizes retaining these key data with identifiers, while overwriting unmarked regular data segments.

[0073] This approach fundamentally transforms the passive model of data storage, turning it into a proactive management strategy based on data value judgment. It ensures that what is recorded is not only a continuous data stream, but also a wealth of informative data essence. Key segments in experiments that best reflect process mechanisms, anomalies, and transition points are intelligently identified and securely preserved. This provides researchers with invaluable data anchors for in-depth post-experimental process analysis, tracing the causes of anomalies, and optimizing experimental protocols, greatly improving the efficiency and relevance of data backtracking analysis, and enabling data storage to truly serve knowledge discovery and value creation.

[0074] In existing technologies, achieving continuous operation of reflux extraction and rotary evaporation concentration is a critical problem that urgently needs to be solved. Traditional methods rely entirely on manual control and judgment based on operator experience. First, the user must manually switch the flow guide to the reflux position and set parameters to begin extraction. Determining the extraction endpoint lacks objective standards, typically relying on a preset fixed time or judging concentration by observing the color of the extract and taking offline samples based on experience. This method is highly subjective, inefficient, and prone to errors. After reaching the extraction endpoint, the equipment must be manually paused, the extract manually removed, filtered, and the filtrate transferred to a rotary flask, among other operations. Finally, the flow guide must be manually switched back to the concentration position, parameters reset, and the evaporation process restarted. The entire process is cumbersome, frequently interrupted, and highly dependent on the operator's skill and concentration. Consistency between different batches is difficult to guarantee, resulting in a very low level of automation.

[0075] This invention provides a highly intelligent continuous operation mode implementation. At the start, the user only needs to select a preset continuous operation template and manually position the flow guide structure to the reflux extraction position before starting the system. The artificial intelligence control unit then takes over, entering a high-precision closed-loop control state. For determining the extraction endpoint, the system abandons traditional subjective experience methods and instead uses a near-infrared fiber optic sensor integrated at the extract outlet of the reflux pipeline. The sensor's optical probe performs real-time in-situ scanning of the flowing extract to acquire its spectral information. The concentration analysis module pre-stores a database of characteristic absorption peaks for different solute-solvent combinations and uses a partial least squares algorithm to quickly convert real-time spectral data into accurate extract concentration values. The system is calibrated with standard samples to ensure measurement reliability.

[0076] During the extraction process, the system monitors concentration changes in real time. It not only monitors whether the absolute concentration value reaches the set threshold, but also intelligently monitors the rate of change. When the system detects that the rate of change in extract concentration does not exceed 0.1% per minute for three consecutive minutes, it determines that the extraction process has reached equilibrium and is automatically considered to have reached the extraction endpoint. At this point, the system automatically pauses the equipment and issues a clear audible prompt to notify the user to proceed to the next step. The user then manually removes the distilled extract, completes the filtration, loads the resulting filtrate into a rotary flask, manually switches the flow guide structure to the evaporation and concentration position, and finally triggers the continue command on the operating interface. The system then automatically resumes closed-loop control, seamlessly completing the subsequent rotary evaporation and concentration stages.

[0077] The core value of this implementation lies in extracting the endpoint determination from key aspects of the continuous operation process, transforming it from a subjective, manual process into an objective, quantifiable scientific monitoring system. Continuous production processes significantly reduce the possibility of human intervention and error through automation and integration, ensuring consistent and predictable output. The application of digital technologies, such as real-time monitoring system design and AI-driven quality prediction models, further ensures high consistency across different experimental batches. Generally speaking, improvements in automation, such as the intelligent upgrading of laboratory equipment, not only reduce the complexity of manual operations and improve the accuracy of experimental data but also significantly enhance laboratory safety, thereby significantly improving the automation level, reliability, and repeatability of the entire continuous operation process.

[0078] Although embodiments of the present invention have been disclosed above, they are not limited to the applications listed in the specification and embodiments. They can be applied to various fields suitable for the present invention. For those skilled in the art, other modifications can be easily made. Therefore, without departing from the general concept defined by the claims and their equivalents, the present invention is not limited to the specific details and illustrations shown and described herein.

Claims

1. An intelligent control system for a reflux extraction and rotary evaporation concentration apparatus, comprising a rotating flask, a condenser, a heated water bath, and a vacuum pump, characterized in that, The condenser is provided with a flow guide structure below it for switching between reflux extraction and vacuum concentration. The intelligent control system also includes: An artificial intelligence control unit connects to a temperature sensor, a vacuum sensor, a rotation speed sensor, a data storage unit, and a cloud server. The data storage unit contains a pre-set library of extraction and preparation experimental templates. Each template includes a temperature setpoint range of 30-90℃, a vacuum setpoint range of 5-100 kPa, a rotation speed setpoint range of 50-200 rpm, an experimental duration setpoint range of 10-300 minutes, and a continuous operation mode flag for defining the timing logic of reflux extraction and evaporation concentration. The artificial intelligence control unit integrates: The network interaction module supports remote access from both computer and mobile devices, and provides parameter settings, status monitoring and result visualization interfaces; The process optimization engine learns autonomously from literature databases and historical experimental data to generate process optimization schemes and dynamically update the template library. The processor, in response to template instructions, executes: 1) Automatically call template parameters; 2) Temperature fluctuation is maintained at ±0.5℃ and vacuum error at ±10KPa through closed-loop control; 3) Record sensor data to the data storage unit in real time; 4) When the template contains a continuous operation mode flag, the position of the flow guide structure is switched according to the preset timing. 5) The equipment will automatically stop when the entire experiment is completed.

2. The intelligent control system for the reflux extraction and rotary evaporation concentration apparatus according to claim 1, characterized in that, The process optimization engine is configured with a template cloning module, which creates an independent copy of the template in the working storage area in response to user commands. The copy inherits the parameter range of the original template. The processor is prohibited from modifying the original template and is only allowed to adjust the parameters of the copy. When the user modifies the copy, the process optimization engine generates parameter adjustment suggestions by comparing with the historical best solution and displays them on the operation interface.

3. The intelligent control system for the reflux extraction and rotary evaporation concentration apparatus according to claim 2, characterized in that, The process optimization engine has a built-in parameter verification module. When performing out-of-bounds verification, it simultaneously starts process suitability analysis: it compares the input value with the distribution of process parameters extracted from the same components in the literature library. If it deviates from the distribution center by ±2 standard deviations, it adds a process risk warning. The boundary verification mechanism performs a two-phase operation: In the first stage, a warning window will pop up in real time through the display unit of the operation interface, which includes the name of the out-of-bounds parameter and its corresponding valid range value. The second phase prohibits the execution of experimental start commands containing out-of-bounds parameters until the user corrects the out-of-bounds parameters to within the valid range.

4. The intelligent control system for the reflux extraction and rotary evaporation concentration apparatus according to claim 3, characterized in that, In the parameter modification interface of the operation interface, the parameter verification module simultaneously displays two types of benchmark values ​​for each parameter to be adjusted: The first baseline value is the existing setting of this parameter in the current independent copy; The second baseline value is the dynamic optimal value recommended by the process optimization engine, which is calculated and generated based on the real-time updated historical experimental model. Two types of benchmark values ​​are displayed side by side in the parameter input box as numerical labels, with the source template name indicated for the preset recommended value; When a user inputs a new parameter value, the parameter verification module first calculates the difference between the input value and the second benchmark value before performing out-of-bounds verification. If the absolute difference between the input value and the dynamic optimal value exceeds the process sensitivity threshold of the corresponding parameter (temperature ±5℃, vacuum ±5KPa, rotation speed ±30rpm, duration ±15%), a deviation warning message will be added to the warning window. The message includes the specific deviation amount and direction of the current input value relative to the preset recommended value, as well as a process confidence label. The process confidence label displays the data volume and model fit degree on which the current recommended value is based.

5. The intelligent control system for the reflux extraction and rotary evaporation concentration apparatus according to claim 4, characterized in that, The parameter modification interface of the operation interface has been expanded to include a batch editing mode entry. When the user activates the batch editing mode, the interface generates a table view containing all adjustable parameters. The table displays the parameter name identifier, the current independent copy setting value, the original template preset recommended value, the user-editable input box, and the recommended range generated by the process optimization engine in parallel. After the user completes batch parameter input and triggers the confirmation command, the process optimization engine performs multi-objective optimization simulation on the input values, and the simulation results are pushed to the operation interface with a delay.

6. The intelligent control system for the reflux extraction and rotary evaporation concentration apparatus according to claim 1, characterized in that, The AI ​​control unit adds a hardware status verification phase after responding to the template command and before starting the experiment: The actual water volume in the heating water bath is detected by a water level sensor. When the detected value is lower than the minimum working water level threshold, an insufficient water level warning is generated. The static vacuum level of the system is monitored by a sealed pressure sensor. A test pressure of 10-30 kPa is applied before the vacuum pump is started. If the pressure change exceeds 5 kPa within 10-30 seconds, a seal failure warning is generated. The configuration status of the artificial intelligence control unit - board matching rule base: The process optimization engine trains a risk prediction model based on historical fault data and dynamically adjusts the water level / sealing threshold. The fault handling guidelines are linked to a knowledge graph, and emergency response cases matching the template are pushed to users.

7. The intelligent control system for the reflux extraction and rotary evaporation concentration apparatus according to claim 6, characterized in that, The control unit adds a risk level identifier to the state-template matching rule base. The risk level identifier is dynamically calculated by the process optimization engine: based on the template parameter combination and real-time sensor data, the failure probability is predicted. A probability ≥30% is defined as a level 1 risk; a probability of 10%-29% is defined as a level 2 risk. When multiple concurrent faults are detected during the hardware status verification phase, a tiered handling process is executed: If a Level 1 risk warning is present, the experiment start function on the operation interface will be locked immediately, a warning bar of the first color will flash continuously at the top of the display interface, all risk types will be displayed simultaneously, and a cloud alarm will be triggered and pushed to the bound mobile phone until the Level 1 risk warning is lifted. If only a level 2 risk warning exists, it will be statically displayed in the middle of the display interface as a warning bar of the second color, and users will be allowed to manually skip the warning.

8. The intelligent control system for the reflux extraction and rotary evaporation concentration apparatus according to claim 1, characterized in that, The data storage unit is configured with a ring buffer storage structure, which defines a fixed capacity storage block with a capacity limit of 120% of the maximum data volume of a single experiment. The capacity is automatically expanded as the experiment duration is extended and the cloud backup is synchronized in real time. When the amount of real-time data written does not reach the capacity limit, it is stored linearly in timestamp order; When the real-time data write volume reaches the capacity limit, the first-in-first-out overwrite rule is activated: the earliest stored data from the 1st to the Nth group is marked as overwriteable, where N = buffer capacity - free capacity; New data is written to the overwriteable state data location in timestamp order.

9. The intelligent control system for the reflux extraction and rotary evaporation concentration apparatus according to claim 8, characterized in that, The artificial intelligence control unit is equipped with a key node marking module, which responds to three triggering conditions: Condition 1: The key node marking module generates marking instructions in real time via the manual marking button on the operation interface; Condition 2: When the rate of change of the temperature sensor's monitored value exceeds 0.5℃ / second for 5 consecutive seconds, or the rate of change of the vacuum sensor's monitored value exceeds 10KPa / second for 3 consecutive seconds, the critical node marking module automatically generates a marking instruction. Condition 3: When the process optimization engine identifies the inflection point of the coupled change rate of temperature-vacuum-speed, the critical node marking module automatically generates marking instructions; The circular buffer allocates independent storage identifier bits for marked records; when executing overwrite rules, records containing identifier bits are preferentially retained.

10. The intelligent control system for a reflux extraction and rotary evaporation concentration apparatus according to claim 1, characterized in that, The continuous operation mode is executed according to the following procedure: After the user selects a template, they manually position the flow guide structure to the second working position of the reflux extraction and start closed-loop control. When the concentration of the extract reaches the set threshold or the predetermined time is reached, the artificial intelligence control unit pauses the equipment and issues a prompt sound, waiting for the user to manually remove the distilled extract. After the user completes filtration and loads the filtrate, they can manually switch the flow guiding structure to the first working position of evaporation and concentration, and trigger the "continue" command through the operation interface to restore closed-loop control and complete the rotary evaporation and concentration process. The specific method for monitoring the extract concentration to reach the set threshold is as follows: The optical probe of the near-infrared spectral sensor is integrated into the extract outlet of the reflux pipeline to scan the extract spectrum in real time. A database of characteristic absorption peaks for different solute-solvent combinations is pre-set in the concentration analysis module. The spectral data is converted into extract concentration values ​​using a partial least squares algorithm. After calibration with standard samples, the measurement error is ≤ ±8%. When the concentration change rate is ≤0.1% / min for 3 consecutive minutes, the extraction is considered to have reached equilibrium and the set threshold for extract concentration has been reached.

Images