Intelligent control method and system for constant air volume of biological safety cabinet

Abstract

The application relates to the technical field of biosafety equipment control, and particularly provides a biosafety cabinet constant air volume intelligent control method and system, aiming to solve the technical problem of error control of an existing scheme under a shielding working condition. To this end, the biosafety cabinet constant air volume intelligent control method comprises the following steps: collecting an operation state signal of a biosafety cabinet; determining a multi-dimensional state feature vector based on the operation state signal; obtaining a working condition judgment result based on the state feature vector; and controlling the biosafety cabinet to maintain a constant working area air volume based on the working condition judgment result. In this way, the shielding working condition of the biosafety cabinet can be accurately identified without adding additional hardware sensors, so that low-cost constant air volume intelligent control is realized.

Description

Technical Field

[0001] This invention relates to the field of biosafety equipment control technology, specifically providing a method and system for intelligent control of constant airflow in a biosafety cabinet. Background Technology

[0002] Biosafety cabinets are critical equipment used in biomedical research and clinical testing to provide operational protection, and their core safety assurance relies on stable airflow organization. In conventional control schemes, the negative pressure value of the working area is usually monitored, and the fan speed is adjusted through a proportional-integral-derivative (PID) algorithm to maintain a constant airflow speed in the working area.

[0003] However, when the air outlet or grille is blocked, the increased airflow resistance leads to an abnormal rise in negative pressure. Traditional control systems may mistakenly interpret this as excessive airflow and execute a speed reduction command. This adjustment behavior is contrary to demand, further exacerbating insufficient airflow and creating a dangerous positive feedback loop that could compromise safety barriers and threaten the safety of operators and experimental samples.

[0004] Existing solutions, such as using constant airflow fans or adding wind speed sensors, can partially alleviate the problem, but they require hardware replacement or addition, leading to a significant increase in cost and complex installation and calibration, making them difficult to widely adopt. Therefore, existing solutions cannot accurately distinguish between changes in duct resistance and actual wind speed changes, resulting in erroneous control problems under obstructed conditions. Summary of the Invention

[0005] To overcome the aforementioned deficiencies, this application is proposed to provide a solution, or at least a partial solution, to the technical problem of erroneous control in existing solutions under obstructed conditions. This application provides a method and system for intelligent constant airflow control of a biosafety cabinet.

[0006] In a first aspect, this application provides a method for intelligent control of constant airflow in a biosafety cabinet, comprising:

[0007] The operating status signals of the biosafety cabinet are collected, including at least one of the negative pressure signal of the working area of ​​the biosafety cabinet, the fan drive current signal, and the fan speed signal;

[0008] Determine a multidimensional state feature vector based on the aforementioned operating state signal;

[0009] The working condition determination result is obtained based on the state feature vector, and the working condition determination result includes any one of normal working condition and occlusion working condition.

[0010] Based on the operating condition determination results, the biosafety cabinet is controlled to maintain a constant airflow in the working area.

[0011] In one embodiment of the constant airflow intelligent control method for biosafety cabinets in this application, determining the multidimensional state feature vector based on the operating state signal includes:

[0012] The negative pressure signal, the fan drive current signal, and the fan speed signal are preprocessed to obtain the preprocessed negative pressure value, fan current value, and fan speed value.

[0013] Based on the pre-processed negative pressure value, fan current value, and fan speed value, multi-dimensional state characteristics are determined, wherein the multi-dimensional state characteristics include at least the negative pressure change rate, pressure-speed ratio, current-speed ratio, and pressure integral deviation.

[0014] A multidimensional state feature vector is obtained based on the aforementioned multidimensional state features.

[0015] In one embodiment of the constant airflow intelligent control method for biosafety cabinets in this application, the step of obtaining the operating condition determination result based on the state feature vector includes:

[0016] The state feature vector is input into the machine learning classification model, and the working condition classification result and the confidence probability vector corresponding to each working condition category are output.

[0017] The probability vector is smoothed using a sliding window time-series smoothing process, and the category corresponding to the smoothed maximum probability value is used as the final working condition determination result.

[0018] In one embodiment of the constant airflow intelligent control method for biosafety cabinets in this application, the machine learning classification model includes any one of gradient decision tree model, support vector machine model, random forest model or neural network model; the machine learning classification model is pre-trained using a training dataset that includes at least normal operating condition samples, occlusion operating condition samples and filter aging operating condition samples.

[0019] In one embodiment of the constant airflow intelligent control method for biosafety cabinets in this application, the step of obtaining the operating condition determination result based on the state feature vector further includes:

[0020] Using a Kalman filter, the fan speed command from the previous moment is used as the control input, and the measured pressure value of the working area collected at the current moment is used as the observation. The innovation quantity at the current moment is determined according to the pre-established linear state-space model.

[0021] Obtain the adaptive threshold;

[0022] When the innovation amount is greater than the adaptive threshold for multiple consecutive time periods, output a Kalman anomaly flag;

[0023] The working condition classification result output by the machine learning classification model is fused with the Kalman anomaly flag to generate the final working condition determination result.

[0024] In one embodiment of the constant airflow intelligent control method for biosafety cabinets in this application, the step of fusing the operating condition classification result output by the machine learning classification model with the Kalman anomaly flag includes:

[0025] When the probability of occlusion output by the machine learning classification model is greater than the first threshold and the Kalman anomaly flag is true, it is determined as occlusion confirmation.

[0026] When the probability of occlusion output by the machine learning classification model is greater than the first threshold and the Kalman anomaly flag is false, it is determined to be suspected occlusion.

[0027] When the probability of occlusion output by the machine learning classification model is less than or equal to the first threshold and the Kalman anomaly flag is true, it is determined to be a transient disturbance.

[0028] When the probability of occlusion output by the machine learning classification model is less than or equal to the first threshold and the Kalman anomaly flag is false, it is determined to be a normal operating condition.

[0029] In one embodiment of the intelligent airflow control method for biosafety cabinets in this application, controlling the biosafety cabinet to maintain a constant airflow in the working area based on the operating condition determination result includes:

[0030] When the operating condition classification result is an obstruction condition, freeze the fan speed command at the moment of obstruction; assess the obstruction level based on the deviation of the pressure-speed ratio from the calibration reference value; increase the fan speed according to the obstruction level until the fan current value reaches the preset safety upper limit; or

[0031] When the operating condition determination result is normal, closed-loop negative pressure control is executed.

[0032] In one embodiment of the constant air volume intelligent control method for biosafety cabinets in this application, the classification judgment result also includes the filter aging condition;

[0033] The method further includes: when the classification result is that the filter is in an aging condition, obtaining the filter resistance, and when the filter resistance is greater than a second threshold, increasing the current fan speed to compensate for the air volume reduction.

[0034] In one embodiment of the constant airflow intelligent control method for biosafety cabinets in this application, after the machine learning classification model determines that the obstruction condition has been lifted at multiple consecutive times, the method further includes: smoothly restoring the fan speed to the reference speed corresponding to the target negative pressure using a ramp function.

[0035] In a second aspect, a constant airflow intelligent control system for a biosafety cabinet is provided, comprising:

[0036] At least one processor;

[0037] And, a memory communicatively connected to the at least one processor;

[0038] The memory stores a computer program, which, when executed by the at least one processor, represents the aforementioned intelligent constant airflow control method for biosafety cabinets.

[0039] The above-described technical solutions of this application have at least one or more of the following beneficial effects:

[0040] The constant airflow intelligent control method for biosafety cabinets in this application includes: collecting operating status signals of the biosafety cabinet, which include at least one of the following: negative pressure signal of the working area, fan drive current signal, and fan speed signal; determining a multi-dimensional state feature vector based on the operating status signals; obtaining a working condition judgment result based on the state feature vector, which includes either a normal working condition or an obstruction working condition; and controlling the biosafety cabinet to maintain a constant airflow in the working area based on the working condition judgment result. Thus, without adding additional hardware sensors, the obstruction working condition of the biosafety cabinet can be accurately identified, and airflow compensation control can be performed for the obstruction working condition, thereby achieving low-cost constant airflow intelligent control. Attached Figure Description

[0041] The disclosure of this application will become more readily understood with reference to the accompanying drawings. It will be readily understood by those skilled in the art that these drawings are for illustrative purposes only and are not intended to limit the scope of protection of this application. Furthermore, similar numbers in the drawings are used to denote similar components, wherein:

[0042] Figure 1 This is a schematic diagram of the main process of the intelligent constant air volume control method for a biosafety cabinet in one embodiment of this application;

[0043] Figure 2 This is a schematic diagram of the main structure of the constant air volume intelligent control device for a biosafety cabinet in one embodiment of this application;

[0044] Figure 3 This is a schematic diagram of the structure of the constant air volume intelligent control system for a biosafety cabinet in one embodiment of this application. Detailed Implementation

[0045] Some embodiments of this application are described below with reference to the accompanying drawings. Those skilled in the art should understand that these embodiments are merely illustrative of the technical principles of this application and are not intended to limit the scope of protection of this application.

[0046] In the description of this application, "module" and "processor" can include hardware, software, or a combination of both. A module can include hardware circuitry, various suitable sensors, communication ports, memory, and may also include software components, such as program code, or a combination of software and hardware. A processor can be a central processing unit, microprocessor, image processor, digital signal processor, or any other suitable processor. The processor has data and / or signal processing capabilities. The processor can be implemented in software, in hardware, or a combination of both. Non-transitory computer-readable storage media includes any suitable medium capable of storing program code, such as magnetic disks, hard disks, optical disks, flash memory, read-only memory, random access memory, etc. The term "A and / or B" means all possible combinations of A and B, such as only A, only B, or A and B. The terms "at least one A or B" or "at least one of A and B" have a similar meaning to "A and / or B" and can include only A, only B, or A and B. The singular terms "a" or "this" can also include plural forms.

[0047] See appendix Figure 1 , Figure 1 This is a schematic diagram of the main flow of a biosafety cabinet constant airflow intelligent control method according to an embodiment of this application.

[0048] like Figure 1 As shown, the constant air volume intelligent control method for biosafety cabinets in this application embodiment mainly includes the following steps S10-S40.

[0049] Step S10: Collect the operating status signal of the biosafety cabinet, the operating status signal including at least one of the negative pressure signal of the working area of ​​the biosafety cabinet, the fan drive current signal and the fan speed signal.

[0050] Operating status signals are raw sensor data that characterize the current working status of the biosafety cabinet, including at least one of the following: negative pressure signal, current signal, and rotation speed signal.

[0051] Negative pressure signal is a simulated signal that reflects the pressure difference between the working area of ​​a biosafety cabinet and the atmospheric environment.

[0052] The fan drive current signal is an analog signal that reflects the magnitude of the current driving the fan.

[0053] The fan speed signal is an analog signal that reflects how fast the fan impeller rotates.

[0054] Step S20: Determine the multidimensional state feature vector based on the running state signal.

[0055] A multidimensional state feature vector is a multidimensional feature vector that is obtained by performing calculations or transformations based on the operating state signal and is used to describe the gas path load state of a biosafety cabinet.

[0056] Step S30: Obtain the working condition determination result based on the state feature vector. The working condition determination result includes any one of the normal working condition and the occlusion working condition.

[0057] Normal operating conditions refer to the operating state of a biosafety cabinet where the air outlet, grille, and internal air duct are not abnormally blocked, and the airflow organization is within the allowable range.

[0058] The obstruction condition refers to the operating state in which the airflow channel of the air vent, grille or internal air duct is blocked due to the stacking of experimental equipment, foreign objects covering it, etc., and the air duct resistance increases abnormally.

[0059] Step S40: Based on the operating condition determination results, control the biosafety cabinet to maintain a constant airflow in the working area.

[0060] Based on steps S10-S40 above, the operating status signals of the biosafety cabinet are first collected. These signals include at least one of the following: negative pressure signal in the working area, fan drive current signal, and fan speed signal. A multi-dimensional state feature vector is then determined based on the operating status signals. The operating condition judgment result is obtained based on the state feature vector, and this result includes either normal operating condition or obstruction operating condition. Based on the operating condition judgment result, the biosafety cabinet is controlled to maintain a constant airflow in the working area. In this way, without adding additional hardware sensors, obstruction conditions of the biosafety cabinet can be accurately identified, and airflow compensation control can be performed for obstruction conditions, thereby achieving low-cost constant airflow intelligent control.

[0061] The following provides further explanation of steps S10 to S40.

[0062] Specifically, for step S10, the operating status signals of the biosafety cabinet are collected, including real-time monitoring of multiple physical quantities. For example, the negative pressure signal of the work area can be acquired by a negative pressure sensor installed in the work area. The 0-5V analog voltage signal output by the sensor can be sampled by the analog-to-digital converter (ADC) built into the microcontroller (MCU), and the sampling frequency can be set to 100Hz. To suppress high-frequency noise introduced by factors such as fan vibration, the sampled raw pressure signal can be digitally filtered, for example, using a second-order Butterworth low-pass filter, the formula of which can be expressed as:

[0063] in, This is the filtered pressure value at the current moment. This represents the raw sampling pressure value collected by the sensor at the current moment. This represents the original sampling pressure value collected by the sensor at the previous moment. These are the raw sampling pressure values ​​collected by the sensor at the first two time points. This is the filtered pressure output value from the previous moment. These are the filtered pressure output values ​​from the first two time points. , , These are all filter forward coefficients. , All of these are filter feedback coefficients.

[0064] The wind turbine drive current signal can be acquired by a sampling resistor connected in series in the wind turbine drive circuit, and then sent to the analog-to-digital converter after being amplified by an operational amplifier. Similarly, the interference of pulse width modulation (PWM) switching noise can be eliminated by methods such as moving average filtering.

[0065] The method of obtaining the fan speed signal is related to the fan type. For example, for a brushless DC fan (BLDC) without an encoder, the speed can be estimated by the back electromotive force (BEMF) zero-crossing detection method; for a fan with a Hall sensor, the Hall pulse count can be read directly to obtain an accurate speed signal.

[0066] The above is a further explanation of step S10. Step S20 will be further explained below.

[0067] Specifically, step S20 can be implemented through steps S201 to S203.

[0068] Step S201: Preprocess the negative pressure signal, fan drive current signal and fan speed signal to obtain the preprocessed negative pressure value, fan current value and fan speed value.

[0069] Specifically, stable and reliable physical quantity values ​​can be extracted from the original, noisy sensor signals through operations such as digital filtering and moving average filtering, providing a high-quality data foundation for subsequent feature construction.

[0070] Step S202: Determine multidimensional state characteristics based on the pre-processed negative pressure value, fan current value and fan speed value, wherein the multidimensional state characteristics include at least the negative pressure change rate, pressure-speed ratio, current-speed ratio and pressure integral deviation.

[0071] Specifically, features such as negative pressure change rate, pressure-speed ratio, current-speed ratio, and pressure integral deviation are key indicators characterizing the operational status of biosafety cabinets from different dimensions. For example, the negative pressure change rate can be measured through... Obtain, among which This represents the current negative pressure value in the work area. The negative pressure value of the working area at the previous moment is used to capture the dynamic rate of pressure change and is sensitive to transient disturbances.

[0072] The formula for calculating the pressure-speed ratio is: ,in This refers to the actual operating speed of the fan (RPM). The pressure-speed ratio represents the negative pressure generated per unit speed and is a key characteristic distinguishing normal operating conditions from shutdown conditions. Under normal operating conditions, the pressure... With rotational speed They are approximately proportional, and this ratio It remains near a stable baseline value; however, when shading occurs, abnormally high It may decrease, leading to A significant jump.

[0073] The formula for calculating the current-speed ratio is: ,in This represents the current current of the fan. The current-speed ratio reflects the energy consumption per unit speed and is closely related to the airflow resistance. When the fan is blocked, the resistance increases, and this ratio will also increase.

[0074] The formula for calculating the pressure integral deviation is as follows: ,in The desired pressure is used to capture the deviation caused by the actual pressure continuously deviating from the desired pressure after occlusion occurs.

[0075] In addition, multidimensional state characteristics can also include negative pressure values. Fan current value and fan speed value .

[0076] Step S203: Obtain multidimensional state feature vectors based on multidimensional state features.

[0077] Specifically, the multiple state features in step S202 can be concatenated into a single vector, which is a multidimensional state feature vector, used to comprehensively characterize the current operating state of the biosafety cabinet. For example, the multidimensional state feature vector can be represented as follows: ,in This is the negative pressure value of the working area. The rate of change of negative pressure. For the fan current, This refers to the fan speed. The pressure-speed ratio, For pressure integral deviation, This is the current-speed ratio.

[0078] The above is a further explanation of step S20. Step S30 will be further explained below.

[0079] Regarding step S30, in a preferred embodiment, firstly, the multi-dimensional state feature vector obtained in step S20 is input into a pre-trained machine learning classification model, which outputs the work condition classification result and the confidence probability vector corresponding to each work condition category; the probability vector is then smoothed using a sliding window time series, and the category corresponding to the smoothed maximum probability value is taken as the final work condition determination result. This model can be any one of the following: Gradient Decision Tree (GBDT) model, Support Vector Machine (SVM) model, Random Forest model, or Neural Network model.

[0080] Specifically, taking the GBDT model as an example, this model is trained using a large amount of sample data collected during factory calibration, covering normal operating conditions, obstructed operating conditions, and filter aging conditions. After receiving the state feature vector, the model outputs a probability vector containing the confidence levels corresponding to each type of operating condition. For example, a vector containing normal operating conditions, obstructed operating conditions, and filter aging conditions can be represented as follows: To avoid misjudgments caused by noise in single-frame data, a sliding window can be used to smooth the output probability vector over time. The smoothing formula can be expressed as follows: ,in The sliding window size (e.g., W=3), and The value ranges from 0 to W-1, and the category with the highest probability value in the smoothed probability vector is taken as the final working condition determination result.

[0081] Furthermore, to improve the response speed and detection robustness to rapid occlusion events, in another embodiment, a Kalman filter is used. The fan speed command of the previous moment is used as the control input, and the measured pressure value of the working area collected at the current moment is used as the observation. The innovation quantity at the current moment is determined according to the pre-established linear state-space model. An adaptive threshold is obtained. When the innovation quantity is greater than the adaptive threshold for multiple consecutive moments, a Kalman anomaly flag is output. The working condition classification result output by the machine learning classification model is fused with the Kalman anomaly flag to generate the final working condition judgment result.

[0082] Specifically, using a Kalman filter, a linear state-space model can be established with the fan speed command at the previous moment as the control input and the measured pressure value in the working area as the observation. The state equation of this model can be expressed as: The observation equation can be expressed as: ,in ~N(0,R_noise), where, For the predicted pressure value, The pressure inertia coefficient, For speed-pressure gain, This refers to the fan speed command from the previous moment. This is the actual observed pressure value. R_noise represents the observation noise, and R_noise represents the variance of the observation noise.

[0083] The innovation amount at the current moment can be obtained through Kalman recursion calculation. The calculation formula is as follows: Under normal operating conditions, the amount of innovation It follows a zero-mean Gaussian distribution. When shading occurs, the actual pressure... It will deviate significantly from the model's predicted values, resulting in a decrease in the amount of innovation. A sustained positive offset was observed.

[0084] The adaptive threshold calculation formula is: ,in This represents the historical average of innovations. This is the sensitivity coefficient. Let be the sliding standard deviation of the innovation quantity. For example, 3 can be used as an example of a sensitivity coefficient.

[0085] Specifically, when the innovation quantity is satisfied at multiple consecutive time points (e.g., 3 time points) Greater than the adaptive threshold If this occurs, output the Kalman anomaly flag. Finally, the work condition classification result output by the machine learning classification model is fused with the Kalman anomaly flag to generate the final work condition determination result. For example, the following fusion rule can be set: when the occlusion work condition probability output by the machine learning model is greater than the first threshold and the Kalman anomaly flag is true ( When the probability of occlusion output by the machine learning classification model is greater than the first threshold and the Kalman anomaly flag is false, it is determined as occlusion confirmation (high confidence), meaning the final operating condition is an occlusion condition; when the probability of occlusion output by the machine learning classification model is greater than the first threshold and the Kalman anomaly flag is false ( When the system detects a suspected occlusion (with moderate confidence), it enters an observation and waiting mode, and performs occlusion processing only after multiple confirmations. The system will only proceed if the occlusion probability output by the machine learning classification model is less than or equal to the first threshold and the Kalman anomaly flag is true. When the probability of occlusion output by the machine learning classification model is less than or equal to the first threshold and the Kalman anomaly flag is false, it is judged as a transient disturbance (low confidence), only logs are recorded, and the control mode is not switched; when the probability of occlusion output by the machine learning classification model is less than or equal to the first threshold and the Kalman anomaly flag is false, When the threshold value is 0.5, 0.6, 0.7, etc., it is determined to be a normal operating condition and normal control is maintained. For example, 0.5, 0.6, 0.7, etc. can be used as examples of the first threshold value.

[0086] The above is a further explanation of step S30. Step S40 will be further explained below.

[0087] For step S40, based on the working condition determination result, the biosafety cabinet is controlled to maintain a constant working area airflow, including: when the working condition classification result is an obstruction condition, freezing the fan speed command at the moment of obstruction; assessing the obstruction level based on the deviation of the pressure-speed ratio relative to the calibration reference value; increasing the fan speed according to the obstruction level until the fan current value reaches the preset safety upper limit; or when the working condition determination result is a normal working condition, executing closed-loop negative pressure control.

[0088] Specifically, when the operating condition is determined to be normal, the system executes conventional closed-loop negative pressure control, such as an incremental PID algorithm, with the following control deviation: The incremental output of the PID algorithm is: The final control parameters for the wind turbine are: ,in, For the current control deviation, To calibrate the target negative pressure, To create negative pressure in the work area, , , These are PID parameters.

[0089] When the operating condition determination result is an obstruction condition (e.g., obstruction confirmed), the system immediately switches to obstruction compensation control mode. First, the fan speed command at the moment of obstruction is frozen, and PID integral calculation is paused to prevent the system from continuing to erroneously reduce the speed. Then, based on the pressure-speed ratio... relative calibration reference value deviation To assess the level of occlusion, for example When, the occlusion level is mild, when When, the occlusion level is moderate, when If the shading level is high, the shading level is considered severe. Then, based on the assessed shading level, the fan speed is gradually increased in increments to perform adaptive speed-up compensation. For example, the fan speed is gradually increased in increments of ΔN_step (e.g., 1% of rated speed for mild shading, 2% for moderate shading, and 3% for severe shading), with a two-cycle (e.g., 100ms) observation period after each speed-up. Change, if If the current safety limit I_max is not exceeded (e.g., 110% of the rated current), the speed will continue to increase; if the limit is exceeded, the speed will stop until the fan current reaches the preset safety limit. ,For example .

[0090] The GBDT classifier continues to run during the compensation period; if the maximum speed-up is completed... If the speed is still greater than the preset threshold (e.g., 0.6, 0.7, etc.), the system locks the current speed and activates an audible and visual alarm, such as a buzzer alarm and flashing indicator lights, to prompt the operator to remove the obstruction; at the same time, the duration of obstruction, the compensation speed, the current measured air volume, and other information are recorded in the log.

[0091] In one specific implementation, after the machine learning classification model determines that the shading condition has been lifted at multiple consecutive time points, the method further includes: smoothly restoring the fan speed to the reference speed corresponding to the target negative pressure using a ramp function.

[0092] Specifically, during the compensation period, the system continuously monitors the operating conditions. When it detects that the occlusion has been lifted (for example, the machine learning model determines that the occlusion condition has been lifted at multiple consecutive time points, and the Kalman innovation is high), the system will respond accordingly. After that, the system will smoothly restore the fan speed to the reference speed corresponding to the target negative pressure using a ramp function to avoid airflow impact, and then switch back to the normal closed-loop negative pressure control mode.

[0093] Furthermore, when the operating condition is determined to be filter aging, if the calculated filter resistance exceeds a preset second threshold, the system can proactively increase the current fan speed to compensate for the airflow reduction caused by filter aging and send a maintenance or replacement reminder to the user. The second threshold can be a pre-set threshold, which can be adaptively adjusted according to the actual application scenario, and is not specifically limited thereto.

[0094] It should be noted that although the steps in the above embodiments are described in a specific order, those skilled in the art will understand that in order to achieve the effect of this application, different steps do not necessarily have to be executed in such an order. They can be executed simultaneously (in parallel) or in other orders, and these variations are all within the scope of protection of this application.

[0095] Furthermore, this application also provides a constant airflow intelligent control device for biosafety cabinets.

[0096] See appendix Figure 2 , Figure 2 This is a main structural block diagram of a biosafety cabinet constant airflow intelligent control device according to an embodiment of this application. Figure 2 As shown, the constant airflow intelligent control device for biosafety cabinets in this embodiment mainly includes a data acquisition module 11, a determination module 12, an acquisition module 13, and a control module 14. In some embodiments, one or more of the data acquisition module 11, the determination module 12, the acquisition module 13, and the control module 14 can be combined into one module.

[0097] In some embodiments, the acquisition module 11 can be configured to acquire the operating status signals of the biosafety cabinet, which include at least the negative pressure signal of the working area of ​​the biosafety cabinet, the fan drive current signal, and the fan speed signal.

[0098] The determination module 12 can be configured to determine a multi-dimensional state feature vector based on the running state signal.

[0099] The acquisition module 13 can be configured to acquire the working condition determination result based on the state feature vector. The working condition determination result includes either the normal working condition or the occlusion working condition.

[0100] Control module 14 can be configured to control the biosafety cabinet to maintain a constant airflow in the working area based on the operating condition determination results.

[0101] In one implementation, the specific function is as described in steps S10-S40.

[0102] The aforementioned biosafety cabinet constant airflow intelligent control device is used to perform... Figure 1 The embodiments of the constant air volume intelligent control method for biosafety cabinets shown are similar in technical principle, the technical problems solved, and the technical effects produced. Those skilled in the art can clearly understand that, for the sake of convenience and brevity, the specific working process and related descriptions of the constant air volume intelligent control device for biosafety cabinets can be found in the embodiments of the constant air volume intelligent control method for biosafety cabinets, and will not be repeated here.

[0103] Furthermore, it should be understood that since the various modules are only provided to illustrate the functional units of the device described in this application, the physical devices corresponding to these modules may be the processor itself, or a part of the processor's software, hardware, or a combination of both. Therefore, the number of modules shown in the figures is merely illustrative.

[0104] Those skilled in the art will understand that the various modules in the device can be adaptively split or combined. Such splitting or combining of specific modules will not cause the technical solution to deviate from the principles of this application; therefore, the technical solutions after splitting or combining will fall within the protection scope of this application.

[0105] Those skilled in the art will understand that all or part of the processes in the method of the above-described embodiment can also be implemented by a computer program instructing related hardware. The computer program can be stored in a computer-readable storage medium, and when executed by a processor, it can implement the steps of the various method embodiments described above. The computer program includes computer program code, which can be in the form of source code, object code, executable file, or some intermediate form. The computer-readable storage medium can include any entity or device capable of carrying the computer program code, a medium, a USB flash drive, a portable hard drive, a magnetic disk, an optical disk, a computer memory, a read-only memory, a random access memory, an electrical carrier signal, a telecommunication signal, and a software distribution medium, etc.

[0106] Furthermore, this application also provides a constant airflow intelligent control system for a biosafety cabinet. The constant airflow intelligent control system may include at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores a computer program, and when the computer program is executed by the at least one processor, it implements the constant airflow intelligent control method for the biosafety cabinet described in any of the above embodiments. See also Figure 3 As shown, Figure 3 The structure of a constant airflow intelligent control system for a biosafety cabinet is illustrated by way of example, which includes a processor 100 and a memory 200.

[0107] Furthermore, this application also provides a computer-readable storage medium. In one embodiment of the computer-readable storage medium according to this application, the computer-readable storage medium can be configured to store a program that executes the biosafety cabinet constant airflow intelligent control method of the above-described method embodiments. This program can be loaded and run by a processor to implement the above-described biosafety cabinet constant airflow intelligent control method. For ease of explanation, only the parts related to the embodiments of this application are shown; for specific technical details not disclosed, please refer to the method section of the embodiments of this application. The computer-readable storage medium can be a memory device comprising various electronic devices. Optionally, in the embodiments of this application, the computer-readable storage medium is a non-transitory computer-readable storage medium.

[0108] The technical solution of this application has been described in conjunction with the specific embodiments shown in the accompanying drawings. However, it will be readily understood by those skilled in the art that the scope of protection of this application is obviously not limited to these specific embodiments. Without departing from the principles of this application, those skilled in the art can make equivalent changes or substitutions to the relevant technical features, and the technical solutions after these changes or substitutions will all fall within the scope of protection of this application.

Claims

1. A method for intelligent control of constant airflow in a biosafety cabinet, characterized in that, include: The operating status signals of the biosafety cabinet are collected, including at least one of the negative pressure signal of the working area of ​​the biosafety cabinet, the fan drive current signal, and the fan speed signal; Determine a multidimensional state feature vector based on the aforementioned operating state signal; The working condition determination result is obtained based on the multidimensional state feature vector, and the working condition determination result includes any one of normal working condition and occlusion working condition. Based on the operating condition determination results, the biosafety cabinet is controlled to maintain a constant airflow in the working area.

2. The intelligent constant airflow control method for biosafety cabinets according to claim 1, characterized in that, The determination of the multidimensional state feature vector based on the operating state signal includes: The negative pressure signal, the fan drive current signal, and the fan speed signal are preprocessed to obtain the preprocessed negative pressure value, fan current value, and fan speed value. Based on the pre-processed negative pressure value, fan current value, and fan speed value, multi-dimensional state characteristics are determined, wherein the multi-dimensional state characteristics include at least the negative pressure change rate, pressure-speed ratio, current-speed ratio, and pressure integral deviation. A multidimensional state feature vector is obtained based on the aforementioned multidimensional state features.

3. The intelligent constant airflow control method for biosafety cabinets according to claim 2, characterized in that, The step of obtaining the working condition determination result based on the multidimensional state feature vector includes: The state feature vector is input into the machine learning classification model, and the working condition classification result and the confidence probability vector corresponding to each working condition category are output. The probability vector is smoothed using a sliding window, and the working condition category corresponding to the maximum smoothed probability value is taken as the final working condition determination result.

4. The intelligent constant airflow control method for biosafety cabinets according to claim 3, characterized in that, The machine learning classification model includes any one of the following: gradient decision tree model, support vector machine model, random forest model, or neural network model; the machine learning classification model is pre-trained using a training dataset that includes at least normal operating condition samples, occluded operating condition samples, and filter aging operating condition samples.

5. The intelligent constant airflow control method for biosafety cabinets according to claim 3, characterized in that, The process of obtaining the working condition determination result based on the multidimensional state feature vector also includes: Using a Kalman filter, the fan speed command from the previous moment is used as the control input, and the measured pressure value of the working area collected at the current moment is used as the observation. The innovation quantity at the current moment is determined according to the pre-established linear state-space model. Obtain the adaptive threshold; When the innovation amount is greater than the adaptive threshold for multiple consecutive time periods, output a Kalman anomaly flag; The working condition classification results output by the machine learning classification model are fused with the Kalman anomaly flag to generate the final working condition determination result.

6. The intelligent constant airflow control method for biosafety cabinets according to claim 5, characterized in that, The step of fusing the working condition classification result output by the machine learning classification model with the Kalman anomaly flag includes: When the probability of occlusion output by the machine learning classification model is greater than the first threshold and the Kalman anomaly flag is true, it is determined as occlusion confirmation. When the probability of occlusion output by the machine learning classification model is greater than the first threshold and the Kalman anomaly flag is false, it is determined to be suspected occlusion. When the probability of occlusion output by the machine learning classification model is less than or equal to the first threshold and the Kalman anomaly flag is true, it is determined to be a transient disturbance. When the probability of occlusion output by the machine learning classification model is less than or equal to the first threshold and the Kalman anomaly flag is false, it is determined to be a normal operating condition.

7. The intelligent constant airflow control method for biosafety cabinets according to claim 1, characterized in that, The method of controlling the biosafety cabinet to maintain a constant working area airflow based on the operating condition determination result includes: When the operating condition determination result is an obstruction condition, freeze the fan speed command at the moment of obstruction; assess the obstruction level based on the deviation of the pressure-speed ratio from the calibration reference value; increase the fan speed according to the obstruction level until the fan current value reaches the preset safety upper limit; or When the operating condition determination result is normal operating condition, closed-loop negative pressure control is executed.

8. The intelligent constant airflow control method for biosafety cabinets according to claim 7, characterized in that, The operating condition determination results also include the filter's aging operating condition; The method further includes: when the working condition determination result is that the filter is in an aging condition, obtaining the filter resistance, and when the filter resistance is greater than a second threshold, increasing the current fan speed to compensate for the air volume reduction.

9. The intelligent constant airflow control method for biosafety cabinets according to claim 7, characterized in that, After the machine learning classification model determines that the shading condition has been lifted at multiple consecutive time points, the method further includes: smoothly restoring the fan speed to the reference speed corresponding to the target negative pressure using a ramp function.

10. A constant airflow intelligent control system for a biosafety cabinet, characterized in that, include: At least one processor; And, a memory communicatively connected to the at least one processor; The memory stores a computer program, which, when executed by the at least one processor, implements the constant airflow intelligent control method for the safety cabinet according to any one of claims 1 to 9.

Images