A type of air velocity control system for the air supply profile in a fume hood based on a Venturi valve

Abstract

The present invention relates to the field of indoor ventilation, specifically to a ventilation hood air supply type surface wind speed control system based on a Venturi valve. The technical problem to be solved by the technical scheme is: including a variable air volume Venturi valve, a sensor system and a control system, wherein the variable Venturi valve is installed on the Venturi fume hood and connected to the ventilation duct; the sensor system includes a door height sensor, a pressure difference sensor, a temperature sensor, a humidity sensor, a displacement sensor, a Venturi flowmeter, and an area presence sensor; the control system includes a data acquisition module, an application scenario analysis module, an artificial parameter regulation and analysis module, an operation parameter monitoring module and an intelligent generation module; the sensor system is used to set sensors according to the size and layout of the room in the area where ventilation is required, and the control system calculates the minimum room ventilation volume at the current moment, the air supply volume of the ventilation system, the exhaust volume of the ventilation system, and the surface wind speed of the ventilation system according to the data measured by the sensor. The beneficial effects of the technical scheme are: effectively reducing energy efficiency and improving environmental safety.

Description

Technical Field

[0001] The present invention relates to the field of indoor ventilation, and in particular to a venturi valve-based internal air supply profile wind speed control system for a fume hood. Background Art

[0002] For built-in fume hoods, the air supply outlet is set inside the fume hood, and most of them are set on the forehead of the fume hood. In this way, the air flow flows from the inside to the bottom. The comfort of the experimenters is much better than that of the external air supply, but it is easy for toxic and harmful gases to be blown out of the window and escape. Therefore, although the internal air supply fume hood effectively reduces the energy efficiency, the safety requires better technical means to ensure. Therefore, the present invention proposes an internal wind type control system of the fume hood based on the Venturi valve, which can prevent the escape of toxic substances while effectively reducing the energy efficiency. Summary of the invention

[0003] The technical problem to be solved by this technical solution is: modern laboratories are becoming more and more standardized, and laboratory requirements are becoming higher and higher. As an important part of the laboratory, the energy consumption of fume hoods is also quite large, so the application of air supply fume hoods has also increased. The main advantages of the design of air supply fume hoods are: energy saving, greatly reducing the exhaust volume of the laboratory, thereby reducing the energy consumption of the laboratory; high comfort, and relatively easy control of laboratory temperature and humidity. However, when the relationship between air supply, window air intake and exhaust changes on one side, it is not easy to coordinate. This leads to the disorder of the negative pressure protection function of the core of the fume hood, and toxic and harmful gases are leaked instead. For example, when the front window of the fume hood is opened, the exhaust demand increases. At this time, if the increase in air supply is greater than the increase in exhaust, it is easy to cause negative pressure disorder, and positive pressure is formed in the fume hood, which exposes the operator to danger. The Venturi valve is an air volume regulation and control device manufactured based on the Venturi effect. It is a pressure-independent, high-adjustment ratio, fast-response, and high-precision air volume control valve. The application of Venturi valves in ventilation systems should be able to quickly and effectively remove pollutants that are harmful to health. The construction of ventilation control systems is also the most important part of ensuring the safety of risky environments.

[0004] To achieve the above objectives, the present invention creates a technical solution: a control system for the air supply profile velocity in a fume hood based on a Venturi valve, comprising a variable air volume Venturi valve, a sensor system and a control system, wherein the variable air volume Venturi valve is installed on the Venturi fume hood and connected to the ventilation duct; the variable air volume Venturi valve is provided with a control power supply, a guide plate, a window and an air supply wing on the Venturi fume hood, a grille is provided at the connection between the guide plate and the air supply wing and the guide plate, an outlet is provided on the guide plate, the air flow at the outlet is circulated with the variable air volume Venturi valve, so that the linear air flow entering the fume hood forms a circulation, which can effectively prevent the turbulence of the air flow, and a window regulating valve is provided on the fume hood for adjusting the height of the window to form an internal air circulation in the fume hood, thereby preventing harmful substances from overflowing from the fume hood.

[0005] The sensor system includes a door height sensor, a differential pressure sensor, a temperature sensor, a humidity sensor, a displacement sensor, a venturi flow meter, and an area presence sensor; the sensors can be used to set the sensors in the area that needs to be ventilated according to the size and layout of the room.

[0006] The control system includes a data acquisition module, an application scenario analysis module, a human parameter regulation and analysis module, an operation parameter monitoring module and an intelligent generation module; the sensor system is used to set sensors in areas that need ventilation according to the size and layout of the room, and the control system calculates the minimum air exchange volume of the room at the current moment, the air supply volume of the ventilation system, the exhaust volume of the ventilation system, and the surface wind speed of the ventilation system based on the data measured by the sensors.

[0007] The calculation formula of the minimum room ventilation volume Vmin is: Vmin = (Fl × Fk × Fh) / hn, where Fl represents the length of the room, Fk represents the width of the room, Fh represents the height of the room, and hn represents the number of ventilation times;

[0008] The calculation formula of the air supply volume Ls of the ventilation system is: Ls = Vmin / wt, where Vmin represents the minimum air exchange volume of the room and wt represents the air exchange time;

[0009] The calculation formula of the exhaust volume Lp of the ventilation system is: Lp = Vmin × 85%, where Vmin represents the minimum air exchange volume of the room;

[0010] The calculation formula for the window surface wind speed of the ventilation system is: Where Ls represents the air supply of the ventilation system, Tl represents the length of the fume hood, and Tk represents the width of the fume hood.

[0011] Furthermore, the data acquisition module includes a sensor system interface module and an API interface module. The sensor system interface module is used to collect data measured by the receiving sensor, and the API interface module is used to receive local weather forecast related information; the application scenario analysis module is used to extract scenario variables from the preprocessed application scenario related information, calculate the target variables based on the preprocessed data, and analyze and evaluate the scenario variables and target variables; the artificial parameter control analysis module is used to generate different parameter combination schemes according to the parameters manually set and modified by the user, and select the parameter combination scheme with the best performance; the operation parameter monitoring module is used to monitor the operation parameters of the ventilation system in real time, calculate the system performance index, and provide real-time feedback.

[0012] Furthermore, the intelligent generation module includes a control strategy intelligent generation module and a system autonomous control processing module. The control strategy intelligent generation module divides the working mode of the ventilation system into energy-saving mode and normal mode based on the regional existence data, and automatically generates a control strategy based on the parameter combination scheme with the best performance; the system autonomous control processing module is used to receive the data transmitted by the control strategy intelligent generation module, perform autonomous control based on the control strategy, and realize intelligent control of the ventilation system.

[0013] Furthermore, the specific processing process of the application scenario analysis module is as follows:

[0014] A1. Extract scenario information from pre-processed application scenario related information through natural language processing technology and text mining technology;

[0015] A2. Further analyze the scene information to obtain scene variables;

[0016] A3. Classify and label the scene keywords to obtain scene categories Ni, i=1, 2, 3...n, and classify the scene variables containing the scene keywords into the corresponding scene categories according to the corresponding scene keywords;

[0017] A4. Calculate the minimum room ventilation volume, the air supply volume of the ventilation system, the air exhaust volume of the ventilation system, and the window surface wind speed of the ventilation system at the current moment based on the data measured by the sensor;

[0018] A5. Record the minimum room ventilation volume within 24 hours, the air supply volume of the ventilation system, the exhaust volume of the ventilation system, and the window surface wind speed of the ventilation system as target variables;

[0019] A6. For each scenario variable and each target variable, calculate the mean and vector representation of each scenario variable and each target variable;

[0020] A7. Calculate the covariance matrix between each scenario variable and each target variable;

[0021] A8. For each set of covariance matrices, calculate the partial correlation coefficient

[0022] A9. The partial correlation coefficient and the pre-set partial correlation coefficient threshold Make a judgment and comparison. If If , it means that there is a strong mutual influence between the scene variable and the target variable, the scene variable is retained. Otherwise, it means that the mutual influence between the scene variable and the target variable is weak, and the scene variable is discarded.

[0023] Furthermore, for each scene variable and each target variable, the mean and vector representation of each scene variable and each target variable are calculated; the processing process is as follows:

[0024] A61. Calculate the mean μx of each scenario variable under the scenario category and the mean бx of each target variable. Where μρ represents the data in each scene variable, represents the data in each target variable, ρ represents the amount of data in each scenario variable, Indicates the data volume of each target variable, μxgj, μxyj, μxzj, μxkj, μxwj, μxsj, μxqj, μxrj, μxpj, μxdj, μxtj, μxmj, and μxfj respectively represent the mean values ​​of light, noise, vibration, area, temperature range, humidity range, air flow velocity, number of people in the space, frequency of activities of people in the space, intensity of activities, size of fume hood, door height, and area to be covered; бa, бb, бc, and бd respectively represent the minimum air exchange volume of the room, the air supply volume of the ventilation system, the exhaust volume of the ventilation system, and the mean value of the window surface wind speed of the ventilation system;

[0025] A62. Take the mean of each scene variable as the vector representation of the scene variable The mean of each target variable is used as the vector representation of the target variable

[0026] Furthermore, the covariance matrix between each scene variable and each target variable is calculated; the calculation process is as follows:

[0027] A71. For each scene variable Each target variable Calculate the covariance between them Where E[] represents the expected value, and They represent the vector representation of the scene variable and the target variable, μx and бx represent the mean of the scene variable and the target variable, respectively;

[0028] A72. Fill the calculated covariance into the corresponding position of the covariance matrix to obtain the covariance matrix.

[0029] Furthermore, the calculation formula of the partial correlation coefficient is: in Representing scene variables and the target variable The covariance between Representing scene variables The variance of Represents the target variable The variance of .

[0030] Furthermore, the specific processing process of the control strategy intelligent generation module is as follows:

[0031] C1. Based on the regional presence data, the working mode of the ventilation system is divided into energy-saving mode and normal mode; that is, when there are people, the working mode of the ventilation system is recorded as normal mode, and when there are no people, the working mode of the ventilation system is recorded as energy-saving mode;

[0032] C2. In energy-saving mode, according to the optimal parameter combination scheme, the lower limit of the value range of the scenario variable is used as the current operating parameter, and the Monte Carlo method based on reinforcement learning is used to generate an energy-saving control strategy;

[0033] C3. In the normal mode, according to the parameter combination scheme with the best performance, the location of the personnel is detected, and the density of personnel is calculated based on the location of the personnel. According to the density of personnel, the required wind speed of the ventilation system is calculated to generate a normal control strategy; the processing process is as follows:

[0034] C31. Use sensors to detect the location of people and count the number of people in a unit area;

[0035] C32. Calculate the population density RMD according to the number of people in the unit area and the area of ​​the region, where RSψ represents the number of people in the i-th unit area and MK represents the area of ​​the region;

[0036] C33. Calculate the required wind speed VF of the ventilation system according to the personnel density RMD and the air duct cross-sectional area SJ of the ventilation system, where W represents the correction factor;

[0037] C34. Combining the optimal performance parameter combination scheme and the required wind speed of the ventilation system, a Monte Carlo method based on reinforcement learning is used to generate a general control strategy.

[0038] The beneficial effects of the improved technical solution are:

[0039] The present invention combines the design of a variable venturi valve and a fume hood, and cooperates with the coordinated work of various modules of the sensor and control system to achieve intelligent control of the ventilation system, improve the accuracy, adaptability and flexibility of ventilation control, and meet the needs of different application scenarios.

[0040] Through the application scenario analysis module, scenario variables and target variables are analyzed and evaluated to provide data support, meet the needs of complex application scenarios, and improve the accuracy and reliability of data processing.

[0041] Different parameter combination schemes are generated according to the parameters manually set and modified by the user to achieve flexible ventilation control; through the control strategy intelligent generation module, the working mode is divided according to the regional existence data, and the control strategy is automatically generated in combination with the parameter combination scheme with the best performance, thereby improving the intelligence level of the ventilation system; through the system autonomous control processing module, autonomous control is carried out according to the control strategy to achieve intelligent control of the ventilation system. BRIEF DESCRIPTION OF THE DRAWINGS

[0042] Figure 1 It is a schematic diagram of the overall system of the present invention.

[0043] Figure 2 It is a block diagram of the control system of the present invention.

[0044] Figure 3 This is a flow chart of the steps and methods of the application scenario analysis module of the present invention.

[0045] Figure 4 It is a structural diagram of the Venturi fume hood of the present invention.

[0046] Figure 5 This is a structural diagram of the guide plate and wind supply wing of the present invention. DETAILED DESCRIPTION

[0047] In order to enable those skilled in the art to better understand the technical solution of the present invention, the present invention is described in detail below in conjunction with the accompanying drawings. The description in this part is only exemplary and explanatory and should not have any limiting effect on the protection scope of the present invention.

[0048] A venturi valve-based air supply profile velocity control system in a fume hood comprises a variable air volume venturi valve 1, a sensor system 3 and a control system 4. The variable air volume venturi valve 1 is installed on the venturi fume hood 2 and connected to the ventilation duct 10; the variable air volume venturi valve 1 is provided with a control power supply 11, the venturi fume hood 2 is provided with a guide plate 24, a window 21 and an air supply wing 23, the guide plate 24 is connected to the air supply wing 23, a grille is provided at the connection between the air supply wing 23 and the guide plate 24, an outlet is provided on the guide plate 24, the airflow at the outlet forms a flow with the variable air volume venturi valve 1, and the fume hood 2 is provided with a window regulating valve 22 for adjusting the height of the window 21. The design enables the linear airflow entering the fume hood 2 to form a cycle, which can effectively prevent the turbulence of the airflow, form an internal air circulation in the fume hood, and thus prevent harmful substances from overflowing the fume hood 2.

[0049] The sensor system includes a door height sensor, a differential pressure sensor, a temperature sensor, a humidity sensor, a displacement sensor, a venturi flow meter, and an area presence sensor; the sensors can be used to set the sensors in the area that needs to be ventilated according to the size and layout of the room.

[0050] The control system includes a data acquisition module, an application scenario analysis module, a human parameter regulation and analysis module, an operation parameter monitoring module and an intelligent generation module; the control system calculates the minimum air exchange volume of the room at the current moment, the air supply volume of the ventilation system, the exhaust volume of the ventilation system, and the surface wind speed of the ventilation system based on the data measured by the sensor.

[0051] The intelligent generation module includes a control strategy intelligent generation module and a system autonomous control processing module. The control strategy intelligent generation module divides the working mode of the ventilation system into energy-saving mode and normal mode based on the regional existence data, and automatically generates a control strategy based on the parameter combination scheme with the best performance; the system autonomous control processing module is used to receive the data transmitted by the control strategy intelligent generation module, perform autonomous control based on the control strategy, and realize intelligent control of the ventilation system.

[0052] The data acquisition module includes a sensor system interface module and an API interface module. The sensor system interface module is used to collect data measured by the receiving sensor, and the API interface module is used to receive local weather forecast related information; the application scenario analysis module is used to extract scenario variables from the preprocessed application scenario related information, calculate the target variables based on the preprocessed data, and analyze and evaluate the scenario variables and target variables; the artificial parameter control analysis module is used to generate different parameter combination schemes based on the parameters manually set and modified by the user, and select the parameter combination scheme with the best performance; the operation parameter monitoring module is used to monitor the operation parameters of the ventilation system in real time, calculate the system performance index, and provide real-time feedback.

[0053] Among them, the variable air volume Venturi valve is made using the Venturi principle. The external push-rod actuator drives the internal valve stem to move linearly, and the internal vertebral structure changes the valve opening area as the valve stem moves. The Venturi valve has mechanical pressure independence, and the spring in the vertebral structure can be automatically adjusted according to the pressure change in the pipeline to ensure that the valve air volume is stable. For its specific structural principle, please refer to our company's patent, a variable air volume Venturi airflow control valve, patent number: 201520385834.9.

[0054] Among them, the door height sensor is used to detect the height of the window so that the ventilation system can be automatically opened or closed when needed; the differential pressure sensor is used to monitor the pressure difference between different areas in the ventilation system to ensure normal air flow; the temperature sensor is used to measure the indoor temperature so that the operating speed of the ventilation system can be adjusted when necessary; the humidity sensor is used to measure the indoor humidity so that the operating speed of the ventilation system can be adjusted when necessary; the displacement sensor is used to detect the opening of the Venturi valve; the Venturi flowmeter is used to measure the pressure difference of the Venturi valve; the area presence sensor is used to detect whether there are people in a specific area.

[0055] Among them, the data acquisition module is used to collect data measured by sensors, and obtains application scenario-related information by crawling relevant websites and receiving external API data. The data includes room length, width, height, ventilation times, ventilation time, fume hood length, fume hood width, temperature data, humidity data, Venturi valve opening, Venturi valve pressure difference, and area presence data; the area presence data specifically refers to the presence of personnel in the area that needs to be ventilated, including the presence of people and the absence of people.

[0056] The application scenario analysis module is used to extract scenario variables from the preprocessed application scenario related information, calculate the target variables based on the preprocessed data, and analyze and evaluate the scenario variables and the target variables;

[0057] What needs to be specifically explained in this implementation is that the specific processing process of the application scenario analysis module is:

[0058] A1. Extract scenario information from the pre-processed application scenario related information through natural language processing technology and text mining technology; the scenario information includes scenario keywords, environmental conditions, spatial structure, temperature control, humidity adjustment, airflow management, number of personnel, activity intensity, fume hood specifications, door height, and the area to be covered; scenario application keywords include but are not limited to physical and chemical laboratories, biosafety laboratories, and animal breeding rooms, etc.;

[0059] A2. Further analyze the scene information to obtain scene variables; the scene variables include light, noise, vibration, area, temperature range, humidity range, air flow speed, number of people in the space, activity frequency and intensity of people in the space, size of fume hood, door height, and area to be covered;

[0060] A3. Classify and label the scene keywords to obtain scene categories Ni, i = 1, 2, 3...n, and classify the scene variables containing the scene keywords into the corresponding scene categories according to the corresponding scene keywords, that is, Ni = {xgj, xyj, xzj, xkj, xwj, xsj, xqj, xrj, xpj, xdj, xtj, xmj, xfj}; and xgj = {xg1, xg2, xg3, ..., xgm}, xyj = {xy1, xy2, xy3, ..., xym}, xzj = {xz1, xz 2, xz3,...,xzm},xkj={xk1,xk2,xk3,...,xkm},xwj={xw1,xw2,xw3,...,xwm}, xq1, xq2, xq3,..., xqm}, xrj={xr1, xr2, xr3,..., xrm}, xpj={xp1, xp2, xp3,..., xpm}, xdj={xd1, xd2, xd3,..., xdm} , xtj={xt1, xt2, xt3, …, xtm}, xfj={xf1, xf2, xf3, …, xfm}, where xgj represents the illumination data set contained in the scene category Ni, xyj represents the noise data set contained in the scene category Ni, xzj represents the vibration data set contained in the scene category Ni, xkj represents the area data set contained in the scene category Ni, xwj represents the temperature range data set contained in the scene category Ni, xsj represents the humidity range data set contained in the scene category Ni, xqj represents the airflow velocity data set contained in the scene category Ni, xrj represents the number of people in the space contained in the scene category Ni, xpj represents the activity frequency data set of the personnel in the space contained in the scene category Ni, xdj represents the activity intensity data set of the personnel in the space contained in the scene category Ni, xtj represents the fume hood size data set contained in the scene category Ni, xmj represents the door height data set contained in the scene category Ni, and xfj represents the area data set to be covered contained in the scene category Ni;

[0061] A4. Calculate the minimum room ventilation volume, the air supply volume of the ventilation system, the air exhaust volume of the ventilation system, and the window surface wind speed of the ventilation system at the current moment based on the data measured by the sensor;

[0062] The calculation formula of the minimum ventilation volume Vmin of the room is: Vmin = (Fl × Fk × Fh) / hn, where Fl represents the length of the room, Fk represents the width of the room, Fh represents the height of the room, and hn represents the number of ventilation times;

[0063] The calculation formula of the air supply volume Ls of the ventilation system is: Ls = Vmin / wt, where Vmin represents the minimum air exchange volume of the room, and wt represents the air exchange time;

[0064] The calculation formula of the exhaust volume Lp of the ventilation system is: Lp = Vmin × 85%, where Vmin represents the minimum air exchange volume of the room;

[0065] The calculation formula of the window surface wind speed of the ventilation system is: Where Ls represents the air supply volume of the ventilation system, Tl represents the length of the fume hood, and Tk represents the width of the fume hood;

[0066] A5. Record the minimum room ventilation volume within 24 hours, the air supply volume of the ventilation system, the exhaust volume of the ventilation system, and the window surface wind speed of the ventilation system as target variables;

[0067] A6. For each scenario variable and each target variable, calculate the mean and vector representation of each scenario variable and each target variable; the processing process is as follows:

[0068] A61. Calculate the mean μx of each scenario variable under the scenario category and the mean бx of each target variable. Where μρ represents the data in each scene variable, represents the data in each target variable, ρ represents the amount of data in each scenario variable, Indicates the data volume of each target variable, μxgj, μxyj, μxzj, μxkj, μxwj, μxsj, μxqj, μxrj, μxpj, μxdj, μxtj, μxmj, and μxfj respectively represent the mean values ​​of light, noise, vibration, area, temperature range, humidity range, air flow velocity, number of people in the space, frequency of activities of people in the space, intensity of activities, size of fume hood, door height, and area to be covered; бa, бb, бc, and бd respectively represent the minimum air exchange volume of the room, the air supply volume of the ventilation system, the exhaust volume of the ventilation system, and the mean value of the window surface wind speed of the ventilation system;

[0069] It should be noted that the mean of each scene variable refers to the situation that when obtaining application scenes from the Internet, there will be multiple scene keywords for multiple scene variables, and after classification by step A3, one scene keyword corresponds to multiple category scene variables, and one scene variable corresponds to multiple data, for example, the physical and chemical laboratory, the biosafety laboratory and the animal breeding room all contain lighting, and the lighting in the physical and chemical laboratory, the biosafety laboratory and the animal breeding room respectively contains multiple lighting data, therefore, the mean of each scene variable here refers to the mean of the corresponding data set; the mean of each target variable refers to the mean of each target variable data at each moment within 24 hours, for example, the minimum ventilation volume of a room at each moment within 24 hours is calculated as a room minimum ventilation volume set, and the mean of this room minimum ventilation volume set is calculated as the mean of one of the target variables; by calculating the mean of each scene variable and the target variable, the average level of each scene variable under the category and the average level of each target variable in a day can be obtained, which is helpful to understand the performance of different scene variables in the category and the performance of different target variables in a day, and provide a reference for subsequent analysis;

[0070] A62. Take the mean of each scene variable as the vector representation of the scene variable The mean of each target variable is used as the vector representation of the target variable For example, the mean of illumination data is μxgj, which can be represented as a vector of length T [μxg1, μxg2, …, μxgT]; the mean vector is a vector containing multiple means, each element of which corresponds to the mean of a variable, describing the central position of each scene variable in the data set;

[0071] A7. Calculate the covariance matrix between each scene variable and each target variable. The covariance matrix is ​​a symmetric matrix, in which each element represents the covariance between the scene variable and the target variable. The calculation process is as follows:

[0072] A71. For each scene variable Each target variable Calculate the covariance between them Where E[] represents the expected value, and They represent the vector representation of the scene variable and the target variable, μx and бx represent the mean of the scene variable and the target variable, respectively;

[0073] A72, filling the calculated covariance into the corresponding position of the covariance matrix to obtain the covariance matrix;

[0074] A8. For each set of covariance matrices, calculate the partial correlation coefficient The calculation formula of the partial correlation coefficient is: in Representing scene variables and the target variable The covariance between Representing scene variables The variance of Represents the target variable The partial correlation coefficient refers to the correlation measure between the scenario variable and the target variable under the control of other variables. By calculating the partial correlation coefficient, the interference of other variables can be eliminated and the degree of association between the two variables can be more accurately evaluated.

[0075] A9. The partial correlation coefficient and the pre-set partial correlation coefficient threshold Make a judgment and comparison. If If , it means that there is a strong mutual influence between the scene variable and the target variable, the scene variable is retained; otherwise, it means that the mutual influence between the scene variable and the target variable is weak, and the scene variable is discarded; the pre-set partial correlation coefficient threshold It can be set according to the specific situation, and this embodiment does not specifically limit the specific data;

[0076] The artificial parameter control analysis module is used to generate different parameter combination schemes according to the parameters manually set and modified by the user, calculate the target prediction value according to the parameter combination in the parameter combination scheme, and select the parameter combination scheme with the best performance according to the target variable prediction value of each scheme and the user's target variable expected value; the parameter combination scheme includes different value ranges and weights of different scenario variables; the target prediction value refers to the value of the target variable at a certain moment in the future;

[0077] It should be specifically explained in this embodiment that the specific processing process of the artificial parameter control analysis module is as follows:

[0078] B1. Obtain the parameters manually set and modified by the user, specifically the value range, weight, target variable expected value, and target prediction value time range of the scenario variables set and retained by the user; the target prediction value time range specifically refers to a reference time range set by the user to predict the value of the target variable at a certain moment in the future, which can help determine which time period of historical data needs to be used for prediction; the target variable expected value specifically refers to an expected value set by the user for the target variable;

[0079] B2. Generate cs different parameter combination schemes using a parameter optimization algorithm according to the parameters provided by the user; each scheme includes a set of specific parameter combinations; the parameter optimization algorithm belongs to the prior art means, so it is not specifically described in this embodiment;

[0080] B3. For each parameter combination scheme, the value of the target variable at a certain moment in the future is calculated according to the parameter combination in the scheme, that is, the target predicted value; the processing process is as follows:

[0081] B31. Collect historical data, including the parameter values ​​in each parameter combination scheme and the corresponding target variable values, and divide the data set into a training set and a test set;

[0082] B32. Select a numerical feature representation method to convert the parameter values ​​in the parameter combination scheme into feature vectors that can be processed by the machine learning model;

[0083] B33. Select a regression model, train the model using the training set, and adjust the model parameters to make the model fit the data better;

[0084] B34. Use the test set to evaluate the trained model, calculate the error between the predicted result and the true value, and optimize the model based on the evaluation results, including adjusting hyperparameters and increasing the amount of training data;

[0085] B35. Use the trained model to predict the target variable at a certain time in the future to obtain the predicted value φ of the target variable for the parameter combination scheme;

[0086] B4. Calculate the difference ε between the target variable prediction value and the target variable expected value φ for each parameter combination scheme. Where φI represents the predicted value of the target variable of the Ith parameter combination method, and cs represents the number of parameter combination schemes to evaluate the performance of each scheme;

[0087] B5. Sort the difference values ​​ε in ascending order, and select the parameter combination with the smallest difference value as the parameter combination with the best performance;

[0088] The control strategy intelligent generation module divides the working mode of the ventilation system into energy-saving mode and normal mode based on the regional existence data, and automatically generates a control strategy based on the optimal performance parameter combination scheme and transmits it to the system autonomous control processing module;

[0089] It should be specifically explained in this implementation that the specific processing process of the control strategy intelligent generation module is as follows:

[0090] C1. Based on the regional presence data, the working mode of the ventilation system is divided into energy-saving mode and normal mode; that is, when there are people, the working mode of the ventilation system is recorded as normal mode, and when there are no people, the working mode of the ventilation system is recorded as energy-saving mode;

[0091] C2. In energy-saving mode, according to the optimal parameter combination scheme, the lower limit of the value range of the scenario variable is used as the current operating parameter, and the Monte Carlo method based on reinforcement learning is used to generate an energy-saving control strategy to achieve the purpose of energy saving;

[0092] C3. In the normal mode, according to the optimal parameter combination scheme, the location of the personnel is detected, and the density of personnel is calculated based on the location of the personnel. According to the density of personnel, the required wind speed of the ventilation system is calculated, and a normal control strategy is generated to ensure the comfort of the personnel. The processing process is as follows:

[0093] C31. Use sensors to detect the location of people and count the number of people in a unit area;

[0094] C32. Calculate the population density RMD based on the number of people per unit area and the area. Where RSψ represents the number of people in the i-th unit area, and MK represents the area of ​​the region;

[0095] C33. Calculate the required wind speed VF of the ventilation system based on the population density RMD and the cross-sectional area SJ of the ventilation system. Where W represents the correction coefficient. The specific value of the correction coefficient can be adjusted according to actual needs and design requirements. Generally speaking, when the density of people is high, the wind speed of the ventilation system needs to be increased to improve air flow and fresh air supply. Therefore, the value of the correction coefficient can be appropriately increased in the calculation;

[0096] C34. Combining the optimal performance parameter combination scheme and the required wind speed of the ventilation system, a conventional control strategy is generated by using a Monte Carlo method based on reinforcement learning; the Monte Carlo method based on reinforcement learning belongs to the existing technical means, so this embodiment does not make a specific description;

[0097] The system autonomous control processing module is used to receive data transmitted by the control strategy intelligent generation module, and perform autonomous control based on the control strategy to achieve intelligent control of the ventilation system;

[0098] The autonomous regulation based on the control strategy specifically utilizes the Venturi valve for rapid response and rapid adjustment of the valve exhaust volume to ensure a constant wind speed on the fume hood surface with a control accuracy within ±5%;

[0099] The operating parameter monitoring module is used to monitor the operating parameters of the ventilation system in real time, calculate the system performance index, and provide real-time feedback, so that users can understand the system status at any time;

[0100] It should be specifically explained in this implementation that the processing process of the operating parameter monitoring module is:

[0101] D1. Real-time monitoring of the ventilation system to obtain the operating parameters of the ventilation system, including the system control response time, system adjustment rate, system control actual value and target value, number of successful system control, and total number of control times;

[0102] D2. Calculate the system parameter control accuracy and system control processing capacity based on the operating parameters;

[0103] The calculation formula of the system parameter control accuracy CTa is specifically: Where ctθ represents the θth system control target value, ctsθ represents the θth system control actual value, and wx represents sample data, that is, the corresponding number of system control target values ​​and actual values;

[0104] The calculation formula of the system control processing capacity CTb is specifically:

[0105] Among them, TX represents the system control response time, Cv represents the system adjustment rate, Cg represents the number of successful system control, Zc represents the total number of control, β1, β2, β3 represent the proportional coefficients of each item, and the size of the proportional coefficient is to quantify each parameter to obtain a specific value for subsequent comparison. The size of the proportional coefficient can be determined as long as it does not affect the proportional relationship between the parameter and the quantified value;

[0106] D3. Calculate the system performance index Q according to the system parameter control accuracy CTa and the system control processing capacity CTb, Q = γ × CTa + (1-γ) × CTb, where γ represents the weight coefficient, which is used to balance the influence of the system parameter control accuracy and the system control processing capacity on the system performance;

[0107] D4. Compare and judge the system performance index Q with the pre-set system performance index Q threshold. If Q≥Q threshold, it means that the system performance meets expectations, and the control strategy implemented in the current application scenario is recorded. Otherwise, it means that the system performance does not meet expectations, and the control strategy implemented in the current application scenario is marked, and the control strategy intelligent generation module is re-executed to generate a new control strategy; the pre-set system performance index Q threshold can be set according to the specific situation, and this embodiment does not specifically limit the specific data.

[0108] Finally: The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the protection scope of the present invention.

[0109] The above is only a specific implementation of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art who is familiar with the present technical field can easily think of changes or substitutions within the technical scope disclosed in the present application, which should be included in the protection scope of the present application. Therefore, the protection scope of the present application should be based on the protection scope of the claims.

Claims

1. A venturi valve-based air velocity control system for the air supply profile in a fume hood. Features: It includes a variable air volume venturi valve, a sensor system and a control system. The control system includes an application scenario analysis module, a human parameter regulation and analysis module, and an operation parameter monitoring module. The control system calculates the minimum air exchange volume of the room at the current moment, the air supply volume of the ventilation system, the exhaust volume of the ventilation system, and the surface wind speed of the ventilation system according to the data measured by the sensor. The specific processing process of the application scenario analysis module is as follows: A1. Extract scenario information from pre-processed application scenario related information through natural language processing technology and text mining technology; A2. Further analyze the scene information to obtain scene variables; A3. Classify and label the scene keywords to obtain scene categories Ni, i=1, 2, 3...n, and classify the scene variables containing the scene keywords into the corresponding scene categories according to the corresponding scene keywords; A4. Calculate the minimum room ventilation volume, the air supply volume of the ventilation system, the air exhaust volume of the ventilation system, and the window surface wind speed of the ventilation system at the current moment based on the data measured by the sensor; A5. Record the minimum room ventilation volume within 24 hours, the air supply volume of the ventilation system, the exhaust volume of the ventilation system, and the window surface wind speed of the ventilation system as target variables; A6. For each scenario variable and each target variable, calculate the mean and vector representation of each scenario variable and each target variable; A7. Calculate the covariance matrix between each scenario variable and each target variable; A8. For each set of covariance matrices, calculate the partial correlation coefficient A9. The partial correlation coefficient and the pre-set partial correlation coefficient threshold Make a judgment and comparison. If If , it means that there is a strong mutual influence between the scene variable and the target variable, the scene variable is retained. Otherwise, it means that the mutual influence between the scene variable and the target variable is weak, and the scene variable is discarded.

2. According to claim 1, a venturi valve-based air supply profile wind speed control system in a fume hood, Features: For each scene variable and each target variable, the mean and vector representation of each scene variable and each target variable are calculated; the processing process is as follows: A61. Calculate the mean μx of each scenario variable under the scenario category and the mean бx of each target variable. Where μρ represents the data in each scene variable, represents the data in each target variable, ρ represents the amount of data in each scenario variable, Indicates the data volume of each target variable, μxgj, μxyj, μxzj, μxkj, μxwj, μxsj, μxqj, μxrj, μxpj, μxdj, μxtj, μxmj, and μxfj respectively represent the mean values ​​of light, noise, vibration, area, temperature range, humidity range, air flow velocity, number of people in the space, frequency of activities of people in the space, intensity of activities, size of fume hood, door height, and area to be covered; бa, бb, бc, and бd respectively represent the minimum air exchange volume of the room, the air supply volume of the ventilation system, the exhaust volume of the ventilation system, and the mean value of the window surface wind speed of the ventilation system; A62. Take the mean of each scene variable as the vector representation of the scene variable The mean of each target variable is represented as a vector of the target variable 3. According to claim 1, a venturi valve-based air supply profile wind speed control system in a fume hood, Features: The covariance matrix between each scene variable and each target variable is calculated; the calculation process is as follows: A71. For each scene variable Each target variable X θ , calculate the covariance between them Where E[] represents the expected value, and They represent the vector representation of the scene variable and the target variable, μx and бx represent the mean of the scene variable and the target variable, respectively; A72, filling the calculated covariance into the corresponding position of the covariance matrix to obtain the covariance matrix; The calculation formula of the partial correlation coefficient is: in Representing scene variables and the target variable X θ The covariance between Representing scene variables The variance of Cov(X θ ,X θ ) represents the target variable X θ The variance of .

4. According to claim 1, a venturi valve-based air supply profile wind speed control system in a fume hood, Features: The calculation formula of the minimum ventilation volume Vmin of the room is: Vmin = (Fl × Fk × Fh) / hn, where Fl represents the length of the room, Fk represents the width of the room, Fh represents the height of the room, and hn represents the number of ventilation times; The calculation formula of the air supply volume Ls of the ventilation system is: Ls = Vmin / wt, where Vmin represents the minimum air exchange volume of the room, and wt represents the air exchange time; The calculation formula of the exhaust volume Lp of the ventilation system is: Lp = Vmin × 85%, where Vmin represents the minimum air exchange volume of the room; The calculation formula of the window surface wind speed of the ventilation system is: Where Ls represents the air supply of the ventilation system, Tl represents the length of the fume hood, and Tk represents the width of the fume hood.

Images