Laboratory automation control system based on AI model

Abstract

The invention belongs to the technical field of temperature and humidity control, and particularly relates to a laboratory automatic control system based on an AI model. Aiming at the problems of difficulty in independent regulation and control of temperature and humidity, neglect of influence of personnel activity and the like in the traditional laboratory environment control technology, the system integrates a multi-mode environment perception module, a personnel behavior modeling module, an intelligent decision and control module and an anti-mistaken touch verification module. The multi-modal environment sensing module is provided with various sensors to collect data, and a three-dimensional environment thermodynamic diagram is constructed to monitor temperature and humidity in real time. The personnel behavior modeling module identifies personnel positions and activity intensities and quantifies the influence of the personnel positions and the activity intensities on temperature and humidity; the intelligent decision-making and control module optimizes regulation and control according to a manned-unmanned mode switching strategy; the mistaken touch prevention verification module recognizes the operation intention through LSTM, and mistaken operation is prevented. Accurate regulation and control of laboratory temperature and humidity are realized, the quality and safety of an experimental environment are improved, and the accuracy of an experimental result and the stability of equipment are guaranteed.

Description

Technical Field

[0001] The present invention belongs to the technical field of temperature and humidity control, and particularly relates to a laboratory automation control system based on an AI model. Background Art

[0002] In the field of laboratory environmental control, there are many problems in traditional technologies. First, it is difficult to achieve independent regulation of temperature and humidity. The requirements for temperature and humidity in laboratories are often relatively precise and complex. Different experimental projects may require specific temperature and humidity combinations, but the existing technologies cannot accurately and efficiently adjust the temperature and humidity independently and precisely. Often, one aspect is neglected while dealing with the other, affecting the accuracy of experimental results and the normal operation of experimental equipment. Second, the impact of personnel activities in the laboratory on temperature and humidity has been long neglected. Behaviors such as heat production from metabolism and moisture dispersion from breathing of personnel will change the temperature and humidity environment in the laboratory. However, during the operation of traditional control systems, the personnel factor is not taken into consideration, resulting in deviations in temperature and humidity regulation and making it difficult to maintain a stable experimental environment. Summary of the Invention

[0003] In view of the above technical problems in the background art, the present invention proposes a laboratory automation control system based on an AI model.

[0004] To achieve the above object, the technical solution adopted by the present invention is as follows: It includes a multi-modal environment perception module, a personnel behavior modeling module, an intelligent decision-making and control module, and an anti-misoperation verification module;

[0005] The multi-modal environment perception module includes sensor deployment and data fusion;

[0006] Sensor deployment: Distributed temperature and humidity sensors, airflow sensors, and human body perception sensors are deployed;

[0007] Data fusion: The flow field-coupled Kalman is used to fuse multi-source data to construct a three-dimensional environmental thermal map of the laboratory and real-time monitor the temperature and humidity conditions;

[0008] The personnel behavior modeling module includes behavior recognition and impact quantification;

[0009] Behavior recognition: The position of personnel is tracked in real time through the YOLO algorithm, and the activity intensity of personnel is identified in combination with pressure sensors;

[0010] Impact quantification: According to the activity intensity, functions of heat production from metabolism and moisture dispersion from breathing of personnel are constructed to dynamically calculate the contribution value of personnel to temperature and humidity;

[0011] The intelligent decision-making and control module performs dynamic strategy switching, including an unmanned model and a manned mode;

[0012] Unmanned mode: Model Predictive Control (MPC) is used to maintain constant temperature and humidity;

[0013] Occupied mode: Establish a multi-objective genetic algorithm with temperature and humidity accuracy, energy consumption, and personnel comfort as objectives for multi-objective optimization;

[0014] The anti-mis-touch verification module includes behavior intention recognition and mis-touch protection;

[0015] Behavior intention recognition: Use LSTM to learn and analyze multi-dimensional data such as the hand movement trajectory, touch time, and touch position of the operator to determine whether the operator's behavior intention is a normal operation or a mis-touch;

[0016] Mis-touch protection: When it is determined to be a mis-touch, prevent mis-operation through a voice confirmation mechanism and require secondary verification.

[0017] Preferably, in the multi-modal environment perception module, the specific implementation of constructing a three-dimensional environmental heat map of the laboratory in data fusion includes state vector expansion and flow field model embedding;

[0018] State vector expansion: Introduce the air velocity vector (u, v, w) and the moisture content d of the air to construct a seven-dimensional state vector: where T t is the air temperature at the current moment, H t is the relative humidity at the current moment, is the temperature and humidity gradient vector at the current moment t;

[0019] Flow field model embedding: Predict the temperature and moisture content at the next moment through flow field model embedding, and the calculation method is: where ω t is random noise.

[0020] Preferably, in the personnel behavior modeling module, the YOLO algorithm for behavior recognition uses the YOLOv8 model architecture, and combines the improved Deformable DETR to improve the positioning accuracy in the scene of personnel occlusion. The improvement of Deformable DETR is to integrate spatio-temporal sequence memory, including a historical trajectory buffer with a length of 10 frames, which stores the three-dimensional coordinates and occlusion status of personnel; when the target is completely occluded, predict the position of the next frame through linear interpolation, and the prediction error ≤ 15 cm; at the same time, calculate the occlusion state transition probability, and assign 2 times the attention search weight to the area within 50 cm near the historical trajectory.

[0021] Preferably, the pressure sensors are deployed on the experimental table and the floor of the passage, collect pressure signals through a distributed array, calculate the change rate of the pressure data, that is, the change amount of the pressure per unit time, as the feature of the activity intensity; use random forest to take the extracted pressure features as input and output the activity intensity score of the personnel.

[0022] Preferably, in the quantification of the influence of the human behavior modeling module, constructing the human metabolic heat production and respiratory moisture dissipation functions according to the activity intensity and dynamically calculating the contribution value of the human to the temperature and humidity specifically includes:

[0023] The contribution value of human metabolic heat production to temperature, and the calculation formula is: Where K is the activity intensity score of the person, M is the heat production coefficient per unit activity intensity, t 人 is the duration of the person's activity, T proset is the target temperature preset in the laboratory, T range is the allowable fluctuation range of the laboratory temperature;

[0024] The contribution value of respiratory moisture dissipation to humidity, and the calculation method is: Where D is the moisture dissipation coefficient per unit activity intensity, H is the actual humidity of the current laboratory, H max is the maximum allowable value of the laboratory humidity.

[0025] Preferably, calculating the final laboratory temperature and humidity standards to be set according to the contribution values of temperature and humidity, and adding the influence factor of temperature on humidity when setting the standards to obtain the final laboratory temperature and humidity standards to be set, including the final temperature setting and the final humidity setting;

[0026] The calculation method of the final temperature setting is: Where C p is the specific heat capacity at constant pressure of air, V represents the volume of the laboratory, ρ is the air density, k T is the temperature regulation compensation coefficient;

[0027] The final humidity setting is achieved by coupling the influence of temperature, and the calculation method is: Where, H proset represents the target humidity preset in the laboratory, d sat (T set ) represents the saturated moisture content at temperature T set ; T sat_range is the compensation range of saturated humidity with temperature change.

[0028] Preferably, the specific steps of the multi-objective genetic algorithm optimization are:

[0029] Decision variables: Determine the operating parameters of the equipment for controlling temperature and humidity;

[0030] Coding: Encode the decision variables into chromosomes;

[0031] Initializing the population: Randomly generate a certain number of chromosomes to form an initial population, and the gene values of each chromosome are randomly generated within the value range of the decision variables;

[0032] Fitness evaluation: The non - dominated sorting is used to obtain the current optimal set of Pareto - optimal solutions, and then the selection, crossover and mutation operations are carried out to find the optimal Pareto - optimal solution set in each iteration.

[0033] Termination condition judgment: When the fitness of the population is less than the fitness change threshold, it is regarded as the termination of iteration, and the optimal Pareto - optimal solution set is output.

[0034] Compared with the prior art, the advantages and positive effects of the present invention are as follows: The multi - modal environment perception module accurately monitors the temperature and humidity; the personnel behavior modeling module takes into account the personnel activities and dynamically calculates their impact on the temperature and humidity; the intelligent decision - making and control module optimizes the regulation strategies for the occupied and unoccupied modes respectively to balance the temperature and humidity, energy consumption and comfort; the anti - accidental touch verification module uses LSTM to identify the operation intention, combined with voice confirmation and secondary verification to avoid misoperation. Overall, it realizes the accurate and independent regulation of the temperature and humidity in the laboratory, improves the stability and safety of the experimental environment, and overcomes the deficiencies of the traditional technology. Detailed implementation manners

[0035] In order to more clearly understand the above - mentioned objects, features and advantages of the present invention, the present invention will be further described below with reference to the embodiments. It should be noted that, without conflict, the embodiments of the present application and the features in the embodiments can be combined with each other.

[0036] In the following description, many specific details are set forth in order to fully understand the present invention. However, the present invention can also be implemented in other ways different from those described herein. Therefore, the present invention is not limited by the specific embodiments disclosed in the following specification.

[0037] Embodiment: In the operation of modern laboratories, accurate environmental control is crucial for the accuracy of experimental results, the stability of experimental equipment, and the comfort of personnel. However, there are many problems in traditional laboratory environmental control technologies, such as difficulty in independent regulation of temperature and humidity, inability to effectively consider the impact of personnel activities on the environment, etc., which seriously restrict the efficient operation of laboratories. In order to achieve the effects of realizing accurate and independent regulation of laboratory temperature and humidity, fully considering the impact of personnel activities on the environment, improving the stability and safety of the experimental environment, and solving the deficiencies of traditional laboratory environmental control technologies, the present invention adopts a laboratory automation control system based on an AI model, specifically including a multi - modal environment perception module, a personnel behavior modeling module, an intelligent decision - making and control module, and an anti - accidental touch verification module.

[0038] The multi-modal environment perception module includes deployed sensors and data fusion. Deployed sensors: Distributed temperature and humidity sensors, airflow sensors, and human perception sensors are deployed. Specifically, the distributed temperature and humidity sensors are evenly distributed throughout the laboratory, installed near experimental benches, corners, and ventilation openings, etc., to comprehensively collect temperature and humidity data in different areas and accurately grasp the spatial differences in temperature and humidity. The airflow sensors are installed in the ventilation ducts and at key indoor airflow nodes. The sensors in the ventilation ducts can monitor parameters such as the temperature, humidity, and speed of the incoming and outgoing airflows, and the sensors at key indoor nodes can capture the changes in the airflow within the room, providing a basis for analyzing the impact of the airflow on temperature and humidity. The human perception sensors are arranged around the laboratory entrance, passageways, and experimental benches. Those at the entrance are used to detect the entry and exit of personnel, those in the passageways track the movement trajectories of personnel, and those around the experimental benches sense the activities of personnel during experimental operations, so as to timely detect potential changes in the environmental temperature and humidity caused by personnel activities.

[0039] Data fusion: The flow field coupled Kalman filter is used to fuse multi-source data to construct a three-dimensional environmental thermal map of the laboratory and real-time monitor the temperature and humidity conditions. Specifically, in the data fusion of the multi-modal environment perception module, the specific implementation of constructing the three-dimensional environmental thermal map of the laboratory includes state vector expansion and flow field model embedding. State vector expansion: The air velocity vector (u, v, w) and the moisture content d of the air are introduced to construct a seven-dimensional state vector: where T t is the air temperature at the current moment, H t is the relative humidity at the current moment, is the temperature and humidity gradient vector at the current moment t. Flow field model embedding: The temperature and moisture content at the next moment are predicted through the embedding of the flow field model, and the calculation method is: where ω t is the random noise. When constructing the three-dimensional environmental thermal map of the laboratory, the flow field coupled Kalman filter algorithm is used to fuse and process the multi-source data containing the seven-dimensional state vector. By continuously updating the data in the state vector, the information such as temperature and humidity, airflow velocity, and moisture content of the air collected by the sensors is effectively integrated. Combining with the predicted values of the temperature and moisture content at the next moment obtained by the embedding of the flow field model, it is transformed into an intuitive three-dimensional thermal map. The thermal map shows the temperature and humidity conditions of each location in the laboratory in different colors and brightness, realizing real-time and visual monitoring of the temperature and humidity in the laboratory.

[0040] The personnel behavior modeling module includes behavior recognition and impact quantification; Behavior recognition: The position of personnel is tracked in real time through the YOLO algorithm, and the activity intensity of personnel is recognized in combination with pressure sensors. The YOLOv8 model architecture is adopted, and the improved Deformable DETR is combined to improve the positioning accuracy in the personnel occlusion scenario. The improvement of the Deformable DETR is to integrate spatio-temporal sequence memory, including a historical trajectory buffer with a length of 10 frames, which stores the three-dimensional coordinates and occlusion status of personnel; when the target is completely occluded, the next frame position is predicted through linear interpolation, and the prediction error ≤ 15 cm; at the same time, the occlusion state transition probability is calculated, and a 2-fold attention search weight is assigned to the area within 50 cm near the historical trajectory. Among them, the pressure sensors are deployed on the experimental bench and the ground of the passage, and the pressure signals are collected through a distributed array, and the change rate of the pressure data is calculated, that is, the change amount of the pressure per unit time, as the feature of the activity intensity; the extracted pressure features are used as the input of the random forest, and the activity intensity score of the personnel is output. Specifically, in the laboratory scenario, to accurately identify the activity intensity of personnel, a pressure sensor array is deployed on the experimental bench and the ground of the passage. The sensors collect pressure signals in real time, and the change rate of the pressure data is obtained by calculating the change amount of the pressure per unit time, which is used as the feature of the activity intensity of personnel. Subsequently, these pressure features are input into the random forest model. The random forest outputs the activity intensity score of the personnel based on the comprehensive judgment of multiple decision trees inside it. This process ensures the comprehensiveness of data collection through distributed deployment, effectively reflects the activity intensity using the pressure change rate, and the random forest model realizes the accurate assessment of the activity intensity of personnel by virtue of its strong learning ability and stability.

[0041] Impact quantification: Build functions of personnel metabolic heat production and respiratory moisture dissipation according to the activity intensity, and dynamically calculate the contribution values of personnel to temperature and humidity. The specific implementation includes: The contribution value of personnel metabolic heat production to temperature, and the calculation formula is: where K is the activity intensity score of personnel, M is the heat production coefficient per unit activity intensity, t 人 is the duration of personnel activity, T proset is the preset target temperature of the laboratory, T range is the allowable fluctuation range of the laboratory temperature. The contribution value of respiratory moisture dissipation to humidity is calculated as: where D is the moisture dissipation coefficient per unit activity intensity, H is the actual humidity of the current laboratory, H max is the maximum allowable value of the laboratory humidity. Calculate the final required laboratory temperature and humidity standards according to the contribution values of temperature and humidity. When setting the standards, the influence factor of temperature on humidity is also added to obtain the final required laboratory temperature and humidity standards, including the final temperature setting and the final humidity setting; the calculation method of the final temperature setting is: where C pis the specific heat capacity of air at constant pressure, V represents the volume of the laboratory, ρ is the air density, k T is the temperature control compensation coefficient; the final humidity setting is achieved by coupling the influence of temperature, and the calculation method is: Among them, H proset represents the target humidity set in the laboratory, d sat (T set ) represents the temperature T set Saturated moisture content under T sat_range It is the compensation range of saturated humidity changing with temperature. By constructing metabolic heat production and respiratory moisture dissipation functions, the contribution value of temperature and humidity can be dynamically and accurately calculated according to the intensity of personnel activities, saying goodbye to the traditional extensive temperature and humidity control. It not only considers the impact of personnel activities, but also adds the impact of temperature on humidity, so that the set standards are more in line with the actual environmental needs and the temperature and humidity standards can be adjusted according to the real-time activities of personnel, so as to achieve dynamic and accurate control of laboratory temperature and humidity and create a more stable and comfortable experimental environment.

[0042] The intelligent decision-making and control module performs dynamic strategy switching, including unmanned model and manned mode. Unmanned mode: Model predictive control MPC is used to maintain constant temperature and humidity. Specifically, in the unmanned mode of the laboratory intelligent decision-making and control module, the model predictive control MPC is used to maintain constant temperature and humidity. First, the temperature and humidity environmental data of the laboratory are collected to build a mathematical model that accurately describes the dynamic changes of the environment. Based on this model, the target value of constant temperature and humidity and the allowable fluctuation range of temperature and humidity are set. After that, the model is combined with the current environmental data to predict the change trend of the laboratory temperature and humidity in the future. Based on the prediction results, considering the goals and constraints at the same time, the optimization algorithm is used to calculate the optimal control strategy, such as adjusting the operating parameters of air conditioning and ventilation equipment. Then the control strategy is executed and the environmental data is continuously monitored. The actual monitoring results are compared with the predicted situation. If there is a deviation, the control strategy is adjusted in time. Through such continuous prediction, control, feedback and adjustment, the laboratory environment is always guaranteed to be stable within the set temperature and humidity range.

[0043] Occupied mode: Establish a multi-objective genetic algorithm targeting temperature and humidity accuracy, energy consumption, and personnel comfort, and conduct multi-objective optimization. Specifically, in the occupied mode of the laboratory intelligent decision-making and control module, the process of using the multi-objective genetic algorithm to achieve multi-objective optimization for temperature and humidity control and finally obtain the control parameters is as follows: First, determine the operating parameters of the equipment for controlling temperature and humidity as decision variables. Then encode these decision variables into chromosomes. A chromosome consists of genes, and each gene corresponds to a value of a decision variable. Subsequently, initialize the population, randomly generate a certain number of chromosomes to form the initial population. The gene values of each chromosome are randomly generated within the range of decision variable values, which can ensure the diversity of the population. When performing fitness evaluation, use non-dominated sorting to find the current optimal set of Pareto optimal solutions. Non-dominated sorting can distinguish the superiority and inferiority relationships of different solutions and divide the solutions into different levels. Then perform selection, crossover, and mutation operations on these solutions. Selection is to select individuals with high fitness, crossover is to exchange some genes to generate new individuals, and mutation is to randomly change some gene values, so as to continuously search for a better set of Pareto optimal solutions in each iteration. After each iteration, judge the termination condition. When the fitness of the population is less than the fitness change threshold, it means that the iteration has converged, and at this time the iteration terminates, and the optimal set of Pareto optimal solutions is output. Finally, from the optimal set of Pareto optimal solutions, select the most suitable solution according to actual needs and preferences, and decode the gene values in the solution into the actual values of the decision variables. These actual values are the final control parameters for controlling the operation of the temperature and humidity equipment.

[0044] The anti-misoperation verification module includes behavior intention recognition and misoperation prevention; Behavior intention recognition: Use LSTM to learn and analyze multi-dimensional data such as the hand movement trajectory, touch time, and touch position of the operator to determine whether the operator's behavior intention is a normal operation or a misoperation; Misoperation prevention: When it is determined to be a misoperation, prevent misoperation through a voice confirmation mechanism and require secondary verification. Specifically, in the anti-misoperation verification module for temperature and humidity adjustment based on a large-screen touch screen, behavior intention recognition is the core link. First, integrate high-precision sensing components in the touch layer of the large screen to collect multi-dimensional data such as the hand movement trajectory, touch time, and touch position of the operator in real time. Transmit this data to the trained long short-term memory network (LSTM) model. During the training phase, the model uses a large amount of operation data in real scenarios, including normal temperature and humidity adjustment operations and various possible misoperation situations, to learn the differences in hand movement patterns between normal operations and misoperations. When the operator performs a touch operation on the large screen, the LSTM model will quickly analyze the real-time data. For example, normal temperature and humidity adjustment operations usually have coherent and purposeful hand movements, such as precise sliding or clicking in the temperature and humidity value area; while misoperations often manifest as short and random touches. The model judges whether this operation is an intention to normally adjust the temperature and humidity or a misoperation based on these characteristics.

[0045] If the LSTM model determines that it is a mis-touch, the system immediately activates the mis-touch protection mechanism. A voice interaction window will pop up on the large screen, and at the same time, a clear voice prompt will be issued: "It is detected that this operation may be a mis-touch. Please confirm whether you want to adjust the temperature and humidity." If the operator replies "No" through voice, the system will not make any changes to the temperature and humidity settings and will maintain the current temperature and humidity control status. If the operator replies "Yes" through voice, indicating their willingness to adjust the temperature and humidity, the system will trigger a secondary verification. The secondary verification uses a dynamic password. A randomly generated four-digit password will be displayed on the large screen, and at the same time, the operator is required to speak out the password through voice. Only when the operator accurately speaks out the password will the system recognize this operation as a valid temperature and humidity adjustment instruction, and then adjust the temperature and humidity accordingly according to the operator's operations on the large screen to ensure the stable and precise control of the laboratory temperature and humidity environment.

[0046] The above are only the preferred embodiments of the present invention, and are not intended to limit the present invention in other forms. Any person skilled in the art may use the technical content disclosed above to make changes or modifications into equivalent embodiments with equivalent changes and apply them to other fields. However, any simple modifications, equivalent changes and modifications made to the above embodiments based on the technical essence of the present invention without departing from the technical solution content of the present invention still fall within the protection scope of the technical solution of the present invention.

Claims

1. A laboratory automation control system based on an AI model, characterized in that, It includes a multi-modal environment perception module, a personnel behavior modeling module, an intelligent decision-making and control module, and an anti-mis-touch verification module; The multi-modal environment perception module includes deployed sensors and data fusion; Deployed sensors: Deploy distributed temperature and humidity sensors, airflow sensors, and human perception sensors; Data fusion: Use flow field coupled Kalman to fuse multi-source data, construct a 3D environmental thermal map of the laboratory, and monitor the temperature and humidity in real time; The personnel behavior modeling module includes behavior recognition and impact quantification; Behavior recognition: Real-time track the position of personnel through the YOLO algorithm, and combine with pressure sensors to identify the activity intensity of personnel; Impact quantification: Construct functions of personnel metabolic heat production and respiratory moisture dissipation according to the activity intensity, and dynamically calculate the contribution value of personnel to temperature and humidity; The intelligent decision-making and control module performs dynamic strategy switching, including an unmanned model and a manned mode; Unmanned mode: Use model predictive control (MPC) to maintain constant temperature and humidity; Manned mode: Establish a multi-objective genetic algorithm with temperature and humidity accuracy, energy consumption, and personnel comfort as objectives for multi-objective optimization; The anti-mis-touch verification module includes behavior intention recognition and mis-touch protection; Behavior intention recognition: Use LSTM to learn and analyze multi-dimensional data such as the hand movement trajectory, touch time, and touch position of the operator, and judge whether the behavior intention of the operator is a normal operation or a mis-touch; Mis-touch protection: When it is determined to be a mis-touch, prevent mis-operation through a voice confirmation mechanism and require secondary verification.

2. The laboratory automation control system based on the AI model according to claim 1, wherein, In the multi-modal environment perception module, the specific implementation of constructing a 3D environmental thermal map of the laboratory in data fusion includes state vector expansion and flow field model embedding; State vector expansion: Introduce the air velocity vector (u, v, w) and the moisture content d of the air to construct a seven-dimensional state vector: where T t is the air temperature at the current moment, H t is the relative humidity at the current moment, is the temperature and humidity gradient vector at the current moment t; Flow field model embedding: Predict the temperature and moisture content at the next moment through flow field model embedding. The calculation method is as follows: where ω t is random noise.

3. The laboratory automation control system based on the AI model according to claim 1, characterized in that, In the behavior recognition of the personnel behavior modeling module, the YOLO algorithm uses the YOLOv8 model architecture, and combines with the improved Deformable DETR to improve the positioning accuracy in the personnel occlusion scenario. The improvement of Deformable DETR is to integrate spatio-temporal sequence memory, including a historical trajectory buffer with a length of 10 frames, which stores the 3D coordinates and occlusion status of personnel; when the target is completely occluded, predict the next frame position through linear interpolation, and the prediction error ≤ 15 cm; at the same time, calculate the occlusion state transition probability, and assign 2 times the attention search weight to the area within 50 cm near the historical trajectory.

4. The laboratory automation control system based on the AI model according to claim 1, characterized in that, The pressure sensors are deployed on the experimental table and the channel ground, collect pressure signals through a distributed array, and calculate the change rate of pressure data, that is, the change amount of pressure per unit time, as the feature of activity intensity; Use random forest to take the extracted pressure features as input and output the activity intensity score of personnel.

5. The laboratory automation control system based on the AI model according to claim 1, wherein In the impact quantification of the personnel behavior modeling module, the specific steps of constructing functions of personnel metabolic heat production and respiratory moisture dissipation according to the activity intensity and dynamically calculating the contribution value of personnel to temperature and humidity include: The contribution value of human metabolic heat production to temperature, the calculation formula is: Where K is the activity intensity score of the person, M is the heat production coefficient per unit activity intensity, t 人 is the duration of the person's activity, T proset is the target temperature preset in the laboratory, T range is the allowable fluctuation range of the laboratory temperature; The contribution value of breathing moisture dispersion to humidity is calculated as follows: where D is the moisture dispersion coefficient per unit activity intensity, H is the actual humidity of the current laboratory, and H max is the maximum allowable value of the laboratory humidity.

6. The laboratory automation control system based on the AI model according to claim 5, wherein Calculate the final laboratory temperature and humidity standards that need to be set according to the contribution value of temperature and humidity. When setting the standards, also add the influence factor of temperature on humidity to obtain the final laboratory temperature and humidity standards that need to be set, including the final temperature setting and the final humidity setting; The calculation method of the final temperature setting is as follows: where C p is the specific heat capacity of air at constant pressure, V represents the volume of the laboratory, ρ is the air density, and k T is the temperature control compensation coefficient; The final humidity is set by taking into account the influence of the coupling temperature and is calculated as follows: where H proset represents the target humidity preset in the laboratory, and d sat (T set ) represents the saturated moisture content at temperature T set ; and T sat_range is the compensation range of the saturated humidity with respect to the temperature change.

7. The laboratory automation control system based on the AI model according to claim 1, wherein The specific steps of optimizing the multi-objective genetic algorithm are: Decision variables: Determine the device operating parameters for controlling temperature and humidity; Coding: Encode the decision variables into chromosomes; Initializing the population: Randomly generate a certain number of chromosomes to form an initial population, and the gene values of each chromosome are randomly generated within the value range of the decision variables; Fitness evaluation: Use non-dominated sorting to obtain the current optimal set of Pareto optimal solutions, and then perform selection, crossover, and mutation operations to find the optimal Pareto optimal solutions in each iteration; Judgment of termination condition: When the fitness of the population is less than the fitness change threshold, it is regarded as the termination of iteration, and the optimal Pareto optimal solutions are output.

Images